U.S. ENVIRONMENTAL PROTECTION AGENCY (EPA) SCIENTIFIC ADVISORY COMMITTEE on CHEMICALS (SACC)
                                PUBLIC MEETING

                      TOXIC SUBSTANCES CONTROL ACT (TSCA)
                   PEER REVIEW FOR EPA DRAFT RISK EVALUATION
                           OF 1-Bromopropane (1-BP)
                                       
                                       
                     DOCKET NUMBER: EPA-HQ-OPPT-2019-0235
             TSCA SACC WEBSITE http://www.epa.gov/tsca-peer-review
                                       
             HYATT REGENCY CRYSTAL CITY AT REAGAN NATIONAL AIRPORT
                         2799 JEFFERSON DAVIS HIGHWAY
                              ARLINGTON, VA 22202
                              SEPTEMBER 10, 2019

EPA STAFF


TAMUE L. GIBSON, M.S.
DFO EPA/OSCP

MARK HARTMAN
EPA/OPPT
NHAN NGUYEN, PhD
EPA/OPPT
KATHERINE ANITOLE, PhD
EPA/OPPT
STAN BARONE JR., PhD
EPA/OPPT
EVA WONG
EPA/OPPT
SACC PANEL MEMBERS

KENNETH PORTIER, PhD
AMERICAN CANCER SOCIETY (RETIRED) 
JOHN KISSEL, PhD
UNIVERSITY OF WASHINGTON (RETIRED)
CHARLES BARTON, PhD
INDEPENDENT CONSULTANT
KATHLEEN GILBERT, PhD
UNIVERSITY OF ARKANSAS
STEVEN BENNETT, PhD
HOUSEHOLD & COMMERCIAL PRODUCTS ASSOCIATION
MARK JOHNSON, PhD
US ARMY PUBLIC HEALTH CENTER
HENRY ANDERSON, MD
UNIVERSITY OF WISCONSIN-MADISON
HOLLY DAVIES, PhD
WASHINGTON STATE DEPARTMENT OF HEALTH
ALAN KAUFMAN, M.S.
TOY INDUSTRY ASSOCIATION	
DEBORAH CORY-SLECHTA, PhD
UNIVERSITY OF ROCHESTER
DANIEL SCHLENK, PhD
UNIVERSITY OF CALIFORNIA
SHERI BLYSTONE, PhD
SNF HOLDING COMPANY
CATHERINE WILLETT, PhD
HUMANE SOCIETY 
WILLIAM DOUCETTE, PhD
UTAH STATE UNIVERSITY
RUTHANN RUDEL, M.S.
SILENT SPRING INSTITUTE
SHEELA SATHYANARAYANA, MD
UNIVERSITY OF WASHINGTON
TSCA SACC ad hoc PEER REVIEWERS

MICHAEL R. TWISS, PhD
CLARKE UNIVERSITY
JAMES D. BLANDO, PhD
OLD DOMINION UNIVERSITY
GEORGE P. COBB, PhD
BAYLOR UNIVERSITY
DAVID A. EASTMOND, PhD
UNIVERSITY OF CALIFORNIA
ISAAC N. PESSAH, PhD
UNIVERSITY OF CALIFORNIA
CHARLES V. VORHEES, PhD
UNIVERSITY OF CINCINNATI
EDWARD J. PERKINS, PhD
U.S. ARMY CORPS OF ENGINEERS
PUBLIC COMMENTERS

RICHARD DENISON, PhD
ENVIRONMENTAL DEFENSE FUND
JONATHAN KALMUSS-KATZ, ESQ
EARTHJUSTICE
LINDSAY MCCORMICK
ENVIRONMENTAL DEFENSE FUND
STEPHANIE SCHWARZ
ENVIRONMENTAL DEFENSE FUND
ROBERT SUSSMAN
SAFER CHEMICALS HEALTHY FAMILIES
TRACEY WOODRUFF, PhD
UNIVERSITY OF CALIFORNIA
SUZANNE HARTIGAN, PhD
AMERICAN CHEMISTRY COUNCIL
ROBERT MILLER, JR. 
ALBEMARLE CORPORATION
BEN GANN
AMERICAN CHEMISTRY COUNCIL
RICHARD MORFORD
ENVIRO TECH INTERNATIONAL, INC.

                                       

TABLE OF CONTENTS


OPENING OF MEETING	5
INTRODUCTION OF PANEL MEMBERS	10
OPPT TECHNICAL PRESENTATION OF 1-BROMOPROPANE	14
QUESTIONS FROM THE PANEL	79
PUBLIC COMMENTS	113
CHARGE QUESTION 1	154
CHARGE QUESTION 2	181
CHARGE QUESTION 3	232
CHARGE QUESTION 4	258
OPENING OF THE MEETING - DAY 2	300
FOLLOW UP FROM THE PREVIOUS DAY	300
CHARGE QUESTION 5	317
CHARGE QUESTION 6	428
OPENING OF THE MEETING - DAY 3	509
FOLLOW UP FROM THE PREVIOUS DAY	509
CHARGE QUESTION 7	510
CHARGE QUESTION 8	578





                              OPENING OF MEETING 
                  
                  MS. TAMUE GIBSON:  Good morning and thank you for coming to today's meeting.  I am Tamue Gibson and I will be serving as the Designated Federal Official to the US Toxic Substance Control Act, Science Advisory Committee on Chemicals (TSCA SACC) for this meeting.  And I want to thank Dr. Portier for agreeing to serve as the chair of the SACC for this meeting.  
                  I also want to thank both the members of the committee and the public for attending this important meeting.  We appreciate your time and effort of the committee members in preparing for this meeting, taking into account their busy schedules.  In addition, I want to thank EPA's Office of Pollution Prevention and Toxics and my colleagues on the TSCA SACC staff for their hard work in preparing for this important review of EPA's draft risk evaluation of 1-Bromopropane.  
                  By way of background, the TSCA SACC is a federal advisory committee that provides independent scientific peer review and advice to the Agency on chemical related issues regarding impact of proposed regulatory actions on human health and the environment.  The TSCA SACC only provides advice and recommendations to EPA.  Decision making and implementation authority remains with the Agency.  
                  The TSCA SACC consists of 21 charter members.  The expertise of these members is augmented by a pool of potential ad hoc nominees that are considered and selected on an as needed basis.  For the present meeting, seven ad hoc expert reviewers have been added, and 16 of the 21 established SACC members are contributing to the peer review of 1-Bromopropane.  
                  As the DFO for this meeting, I serve as the liaison between the TSCA SACC and the Agency.  I am responsible for ensuring provisions of the Federal Advisory Committee Act are met.  The Federal Advisory Committee Act of 1972 established a system that governs the creation, operation, and termination of executive branch advisory committees.  TSCA SACC meetings are subject to all FACA requirements.  These include open meetings, timely public notice of meetings, and document availability, which is provided via the Office of Pollution Prevention and Toxic's public docket at regulations.gov.  
                  As a Designated Federal Official for this meeting, a critical responsibility is to work with appropriate officials to ensure that all appropriate ethics regulations are satisfied, and in that capacity, the committee members receive training on the Federal Conflict of Interest laws.  
                  In addition, each participant has filed a standard government financial disclosure report.  Along with our Deputy Ethics official with the Office of Science Coordination and Policy and in consultation with the Office of General Counsel, we have reviewed those reports to ensure that all ethics requirements are met.  A sample copy of this form is available on the TSCA SACC website.  The address for this website is noted on the meeting agenda.  
                  The TSCA SACC will review challenging scientific issues over the next few days.  We have a full agenda, and meeting times are approximate.  Thus, we may not keep to exact times as noted due to committee discussions and public comments.  We strive to ensure adequate time for Agency presentations, public comments, and committee deliberations.  
                  For all presenters, committee members, and public commenters, please identify yourselves and speak directly into the microphones provided since this meeting is being webcasted, recorded, and transcribed.  Copies of all EPA presentation materials and written public comments are available in the public docket at regulations.gov.  Any copies of presentations given during this meeting by public commenters will be available in the docket within the next week.  
                  Members of the committee are encouraged to fully consider all written an oral public comments submitted for this meeting.  For members of the public that have not pre-registered for public comments, please notify either myself or another member of the TSCA SACC staff.  If you are interested in making a comment, at this time, the agenda is full.  However, as we move through the proceedings, if time allows, we may be able to accommodate additional brief comments of five minutes or less.  
                  As I mentioned previously, there is a public docket for this meeting.  All background materials, questions posed to the committee by the Agency, and other documents relating to this meeting are available in the docket.  Some documents are also available on the EPA SACC website.  
                  The docket number and website are noted on the meeting agenda.  For members of the press, EPA media relation staff are available to answer your questions at this meeting.  Please address all questions to Robert Daguillard.  He's our EPA media contact, and he can be reached at 202-564-6618.  
                  At the conclusion of the meeting, the TSCA SACC will prepare a report as a response to questions posed by the Agency, the background materials, presentations, and public comments.  The final report also serves as meeting minutes.  We anticipate the final report and meeting minutes will be completed in approximately 90 days after this meeting.  
                  Again, I wish to thank the committee for your avid participation, the program office for your diligent work, my colleagues in the Office of Science Coordination and Policy, and the public for your attendance.  I'm looking forward to both the Office of Pollution Prevention and Toxic's speakers, their presentation on the 1-BP overview and, later, the SACC's consideration and discussions regarding the risk evaluation for 1-BP.  At this time, I now turn the meeting over to our chair, Dr. Ken Portier. 
INTRODUCTION OF PANEL MEMBERS
                  DR. KENNETH PORTIER:  Thank you, Ms. Gibson or Tamue.  I'm going to refer to her as Tamue, the DFO.  I'm Ken Portier, chair of the SACC, and I welcome this large panel for a review of 1-Bromopropane.  I welcome the EPA staff and the public to this public meeting.  My background is in biostatistics and agriculture, environment, and even in public health.  And I'm looking forward to this second discussion of 1-Bromopropane in the last three years.  I'll start with Mark. 
                  DR. MARK JOHNSON:  Good morning.  I'm Mark Johnson, Director for Toxicology at the Army's Public Health Center.  My background is in environmental toxicology and risk assessment. 
                  DR. EDWARD PERKINS:  Hi, my name is Ed Perkins.  I'm from the U.S. Army Corps of Engineers.  I'm Senior Scientist in Genetic Toxicology and Environmental Networks.  And my expertise is in toxicogenomics and environmental toxicology. 
                  DR. CHARLES VORHEES:  I'm Professor of Pediatrics and Environmental Health at the University of Cincinnati and Cincinnati Children's Hospital.  My area is neuroscience and neurotoxicology. 
                  DR. KATHLEEN GILBERT:  I'm Kate Gilbert.  I recently retired as a professor from the University of Arkansas for Medical Sciences, and my background is immunotoxicology. 
                  DR. DANIEL SCHLENK:  Good morning.  I'm Dan Schlenk.  I'm a professor of aquatic ecotoxicology at the University of California at Riverside, and my expertise is in fate and effects of environmental contaminants on aquatic species. 
                  DR. STEVEN BENNETT:  I'm Steven with the Household Commercial Products Association.  I'm a chemist by training, and I bring expertise on consumer use and exposure to products. 
                  MR. ALAN KAUFMAN:  Hi, Al Kaufman with the Toy Association, biologist and organic chemist by training and expertise is in manufacturing processes and downstream uses. 
                  DR. MICHAEL TWISS:  My name's Michael Twiss.  I'm a professor of biology at Clarke University.  I'm a Great Lakes limnologist.  My expertise is in aquatic ecosystems. 
                  DR. CATHERINE WILLETT:  My name is Kate Willett.  I'm with Humane Society International.  My background is in biochemistry and genetics and zebra fish biology, but the past -- most recently, I've become an expert in the application of new approach methodologies for chemical hazard and risk assessment. 
                  DR. GEORGE COBB:  I'm George Cobb, chair of the Environmental Science Department at Baylor University, environmental chemist by training. 
                  DR. JAMES BLANDO:  Jim Blando, I'm an associate professor at Old Dominion University.  My background is industrial hygiene and occupational safety. 
                  DR. SHERI BLYSTONE:  Hi, Sheri Blystone, organic chemist by training.  I've been working as a product safety and compliance professional in the chemical industry for about 25 years, currently with SNF Holding Company. 
                  DR. WILLIAM DOUCETTE:  Bill Doucette, professor in environmental engineering at Utah State University.  I'm an environmental fate chemist. 
                  DR. SHEELA SATHYANARAYANA:  I'm Sheela Sathyanarayana, and I'm a pediatrician and associate professor of pediatrics and adjunct in the Department of Environmental and Occupational Health Sciences.  And my area of focus is pediatric environmental epidemiology. 
                  DR. HOLLY DAVIES:  I'm Holly Davies.  I'm a toxicologist at the Washington State Department of Health.  My background is in development and reproduction. 
                  DR. DAVID EASTMOND:  I'm Dave Eastmond from the University of California at Riverside.  I'm a professor there.  My expertise is in genetic toxicology and environmental carcinogenesis. 
                  MS. RUTHANN RUDEL:  My name's Ruthann Rudel.  I'm Research Director at Silent Spring Institute in Boston, and my background's in toxicology, chemistry, exposure assessments, and risk assessment. 
                  DR. CHARLES BARTON:  My name is Chuck Barton.  I'm an independent consultant.  My specialty is toxicology, specifically hazard identification, exposure estimation, and risk assessment. 
                  DR. JOHN KISSEL:  I'm John Kissel.  I'm Professor Emeritus of Environmental and Occupational Health Sciences at the University of Washington in Seattle.  I'm an environmental engineer by training and a human exposure scientist by practice. 
                  DR. KENNETH PORTIER:  And we have three panel members joining us on the phone.  Dr. Anderson? 
                  DR. HENRY ANDERSON:  I'm Henry Anderson.  I'm a specialist in occupational environmental medicine and public health, along with epidemiology.  I'm at the University of Wisconsin School of Medicine in Public Health. 
                  DR. KENNETH PORTIER:  Thank you.  Dr. Cory-Slechta?  She may be joining us a little later.  Dr. Pessah? 
                  DR. ISAAC PESSAH:  Good morning.  Isaac Pessah, Professor of Toxicology with an expertise in neurotoxicology, University of California Davis. 
                  DR. KENNETH PORTIER:  Thank you.  At this point, we're going to go on ahead, and we're turning it over to the EPA Office of Pollution Prevention and Toxics who are going to introduce the presenters and make a short technical presentation this morning.  Dr. Hartman? 
                  
OPPT TECHNICAL PRESENTATION OF 1-BROMOPROPANE
                  
                  MR. MARK HARTMAN:  Thank you.  So I am the Immediate Office of Pollution Prevention and Toxics.  I'm actually the Deputy Director.  My name is Mark Hartman.  I'm the Deputy Director of the Office of Pollution Prevention and Toxics, and it's a pleasure to be here with you today.  I want to add to the welcoming to the committee.  I see many familiar faces that are now veterans of the first two SACC committees, so we do appreciate very much that you are coming back for more of the fun.  And for those of you who are joining us for the first time, welcome to the pool.  The water is warm.  
                  So implementing the ambitious requirements of the 2016 TSCA amendments has been and continues to be a major priority for the EPA.  And this meeting is an important milestone in the continuing efforts we are making for the first ten chemicals going through the risk evaluation process, specifically, moving forward on the 1-Bromopropane assessment.  Transparency and robust and rigorous scientific discussion dialogue are foundational elements of the program that we established for evaluating existing chemicals.  And the proceedings of this committee over the next two and a half days is to provide independent peer review of our draft risk evaluation are central to our efforts in this regard.  
                  I heard one of the panel members -- I overheard one of them talking about what are the impacts of the work that's been done so far on these assessments and looking forward to seeing how the comments that have been made on the first three chemicals that have gone through this process are reflected in the final risk evaluation.  
                  So I can tell you that every day at EPA, I'm witnessing the impacts of those discussions on the development of the ongoing chemicals in the first ten.  So in addition to the impacts you all see, I'm sure when you see the final risk evaluations go out for the chemicals you've already reviewed, just know that we are continually learning from these engagements and applying those learnings to the evaluations that are currently under way.  And hopefully you will be able to see evidence of that as you continue to read our documents.  
                  In addition to the work that we've continued to do on the first ten, since the last time we've met, a couple of important milestones have been met as well just in general in the existing chemicals review program. On August 23, we released the 20 dossiers for what we're calling the high priority chemicals.  We affectionately call them the next 20.  And we anticipate following the end of that comment period and looking at comments that we receive through that process to begin the risk evaluation process for those 20 chemicals in late December.  
                  And we've also received and made publicly available the first two requests by manufacturers for a manufacturer requested risk evaluation as well, which are also currently in the docket for public review.  So it's safe to say that if you're buying stock in the EPA existing chemicals program, you can be reassured that we're going to have a lot of work to be done and that we're looking forward to in the future.  
                  I'd also like to thank the team, many of whom are here today, the career staff at the EPA, who have worked very hard and shown a great deal of commitment and diligence in bringing the risk evaluation to the point it is today.  One of the best parts of my job is getting to see the commitment and dedication of the career team and how hard they've worked to try to make the vision of the 2016 TSCA amendments a reality.  
                  So with that, I'm going to go ahead and pass it along to Dr. Anitole who's the team lead for the group, and she'll do the introductions.  Thank you. 
                  DR. KATHERINE ANITOLE:  Good morning to the Scientific Advisory Committee on Chemicals, public commenters and stakeholders.  My name is Katherine Anitole, and I'm the team lead for the 1-Bromopropane, or 1-BP team.  Our management lead for this assessment is Nhan Nguyen.  Today, I will be presenting an overview of the 1-BP draft risk evaluation.  I am fortunate to be part of a very dedicated and talented team whose names are listed here and whose work is reflected in this presentation.  
                  Over the next hour or so, I will walk you through an overview of the draft risk evaluation for 1-BP.  Based on suggestions received from the SACC committee during earlier draft risk evaluation presentations, I will diverge slightly from the organization of the draft risk evaluation document for this presentation.  I will begin by providing an overview of general topics.  Following this, I will describe the conditions of use, chemistry, environmental fate and transport and releases to the environment for 1-BP.  From here, I will diverge from the draft risk evaluation by describing the characterization of environmental exposure, hazard, and risk together.  This will be followed by the characterization of human exposure, hazard, and risk.  
                  To provide some context and background for the risk evaluation, this slide summarizes some of the refinements EPA has made to the 2016 draft risk evaluation that underwent peer review with the Chemical Science Advisory Committee, or CSAC, in June of 2016.  These refinements include application of the TSCA systematic review process; review, consideration and incorporation of new information which provided additional weight of evidence information to assess the mode of action for cancer; review, consideration, and evaluation of the dermal exposure route for both occupational and consumer scenarios; assessment of additional conditions of use; inclusion of nested modeling for developmental toxicity; and application of three modeling approaches for cancer: multistage and two model averaging approaches.  
                  This slide depicts an overview of the 1-BP risk evaluation roadmap showing the inputs and outputs during various phases of the risk evaluation process.  I will very briefly go over the roadmap, but the details of the draft risk evaluation, which includes information in the output boxes, will be described as we proceed with the presentation.  The risk evaluation process begins with a gathering of information from various sources to develop a scope document, which describes the conditions of use, or COUs, for industrial, commercial uses, consumer uses, environmental releases, exposures, including potentially exposed susceptible subpopulations, or PESS, hazards, and finally development of the initial conceptual models and analysis plan.  
                  Based on comments received on the scope document, a problem formulation document is developed with a further refinement of the initial conceptual models and analysis plan.  Based on comments received on the problem formulation document, the risk evaluation process begins.  This is where the analysis plan is executed with development of the final draft risk evaluation that characterizes the risk.  I should mention that the scope documents for the first ten chemicals will differ from those that EPA anticipates publishing in the future.  There will not be separate scope and problem formulation documents after the first ten chemicals.  
                  One note of orientation as I begin, on each slide in the upper right hand corner you will find a callout to the relevant sections of the draft risk evaluation and, where appropriate, an additional reference to the associated charge question.  We hope that you will find this helpful as we move through the overview presentation.  This slide provides an overview of the systematic review process that we use for the TSCA risk evaluations, which includes data screening, data evaluation, data integration, and data extraction.  The systematic review process is described in various parts of the draft risk evaluation, such as EPA's application of systematic review in TSCA risk evaluations document, in Section 1.5 of the draft 1-BP risk evaluation, and also in the supplementary documentation of study evaluations.  
                  This slide, on the left hand side, provides a summary of results of the systematic review process for the various subject matter scientific disciplines.  On the righthand side is an example literature diagram illustrating a systematic review process and results for human health corresponding to the boxes shown on the previous slide.   The literature flow diagrams are included for each of the disciplines and are provided in Section 1.5 of the draft risk evaluation document.  This information relates to Charge Question 1.  
                  This slide summarizes the physical-chemical properties of 1-BP.  The general structure of 1-BP is presented on the left.  A few basic physical-chemical properties are provided on the right side of the slide.  1-BP is slightly soluble in water with a high vapor pressure at close to room temperature.  The vapor pressure provides an indication of the relative tendency of a substance to volatilize.  
                  To understand conditions of use relative to one another and associated potential exposures under those conditions of use, this diagram depicts each stage of the 1-BP lifecycle as reported in the 2016 Chemical Data Reporting, or CDR.  Further details and description of this diagram can be found in Section 1.4.1 of the draft risk evaluation.  The primary use of 1-BP is as a solvent for cleaning and degreasing in various industrial, commercial, and consumer uses and activities, including vapor degreasing, cold cleaning, aerosol degreasing, spray adhesives, dry cleaning, and in cleaning and furniture care products, as well as a variety of other uses.  The EPA's CDR database indicates that there were two manufacturers and eight importers producing or importing 25.9 million pounds of 1-BP in the U.S. in 2015.  
                  The table on the next two slides identifies the conditions of use in corresponding exposure scenarios assessed in the risk evaluation.  The first column has the general description of the category and subcategory for each condition of use included in the risk evaluation.  The categories and subcategories are based on functional use codes and product category codes as defined in EPA's CDR rule.  The second column identifies specific exposure scenarios used in the draft risk evaluation to assess the corresponding condition of use on the left hand side.  Additionally information regarding the conditions of use for 1-BP are discussed in Section 1.4 of the draft risk evaluation.  Table 2-2 of the draft risk evaluation provides a crosswalk of the conditions of use and the sections of the draft risk evaluation where the occupational exposure scenarios are discussed.  
                  In building the risk evaluation, the EPA developed a series of conceptual diagrams to illustrate the flow of 1-BP from conditions of use to receptors via exposure pathways and routes.  The EPA also identified the hazards relevant to each receptor.  Here, I present one example of a conceptual model diagram for occupational settings.  The conceptual models for 1-BP can be found in Section 1.4.2 of the draft risk evaluation.  
                  This slide provides an overview of the ecological receptors and human populations assessed in the draft risk evaluation.  For workers, we assessed both workers and occupational nonusers, or ONUs.  For purpose of the risk evaluation, workers refer to employees who directly handle 1-BP as part of their job function.  Occupational nonusers are employees who do not directly handle 1-BP but who perform work in an area where 1-BP is present.  Similarly for consumers, we assessed both consumers and bystanders.  Consumers are direct users of a product.  Bystanders a nonproduct users who are incidentally exposed to the product or article in residential settings.  More details on the receptors and populations are described in Section 1.4.2 of the draft risk evaluation.  
                  Next, I will discuss the environmental fate and transport properties of 1-BP.  This information is described in Section 2.1 of the draft risk evaluation.  This slide provides a summary of the environmental fate and transport properties of 1-BP.  1-BP's Henry's Law constant, which is a ratio of vapor pressure to water solubility, indicates its volatilization from water will be important.  A fit for purpose model used in 1-BP scoping estimated volatilization half-lives of hours for a model stream and days for a model lake.  1-BP's soil to sediment organic carbon partition coefficient, or KOC, indicates 1-BP partitioning to soil and sediment will be low.  1-BP is not persistent or bioaccumulative based on its hydrolysis anerobic biodegradation half-lives and its estimated bioconcentration factor.  
                  Moving to the righthand side of the table, data quality for environmental fate studies was evaluated against criteria grouped into eight broad evaluation domains containing 18 metrics more specifically related to study design.  While measured data from high quality studies are preferred, data gaps were filled with estimated endpoint data from EPI Suite 4.1.  The study quality scores for endpoints estimated by EPI Suite were based on a single metric for structure activity models.  The models in EPI Suite have defined endpoints, known chemical domains, and performance statistics.  EPI Suite was peer reviewed by the EPA Science Advisory Board in 2007.  
                  At this point, I would like to describe the environmental risk assessment, starting with the releases and exposures.  This information is described in Section 2.2 and Appendix G of the draft risk evaluation and relates to Charge Question 4.2.  The manufacturing, processing, use and disposal of 1-BP can result in releases to the environment.  To examine whether near facility surface water concentrations could approach aquatic concentrations of concern for 1-BP, EPA employed a conservative approach using a conservative modeling approach in combination with the Toxic Release Inventory, or TRI, data to estimate near facility surface water concentrations.  Reported releases to water were limited to five pounds from one facility for the TRI reporting year 2016.  This facility, located in Corpus Christi, Texas, has reported one pound of 1-BP released to the Nueces River with 100 percent from storm water on an annual basis.  They also reported four pounds of 1-BP released to an unnamed water body with 83 percent from storm water on an annual basis.  Subsequent reporting to the TRI has indicated that water releases continue to be of the same order of magnitude or less.  
                  There was one facility reporting water releases.  In modeling surface water concentrations, EPA used a peer-reviewed tool, exposure and fate assessment screening tool, or EFAST, and presumed these direct releases to be untreated.  The total release period was varied over a period of 120 or 100 days.  High end surface water concentrations, that is those obtained assuming low receiving water body streams from all EFAST rungs, range from 0.08 to 77.9 micrograms per liter.  As mentioned in the previous slide, post problem formulation, EPA determined that subsequent reporting of the TRI has indicated that water releases continue to be on the same order of one pound per year.  Therefore, aquatic exposures are expected to be of the same order of magnitude as described in the problem formulation.  
                  At this point, I would like to describe the hazards and risk characterization associated with the environmental assessment of 1-BP.  This section is discussed primarily in Section 3.1 and Appendix F of the draft risk evaluation and pertains to Charge Question 4.  This diagram presents all environmental exposure pathways and receptors that are within the scope of the assessment, which includes all aquatic and terrestrial pathways.  Exposure pathways to terrestrial and sediment dwelling aquatic species are greyed out because EPA determined that no further quantitative analysis was necessary as part of this assessment.  
                  This determination was based on a qualitative consideration of the physical-chemical and environmental fate characteristics, as well as the conditions of use for 1-BP, which indicates limited presence in terrestrial environments and aquatic sediments.  As a result, exposures and risks to these receptors are not expected.  Available TRI release information indicated that aquatic releases are expected for 1-BP.  As a result, EPA carried out a quantitative screening level risk analysis comparing available environmental hazard data for aquatic species and estimated aquatic exposure concentrations.  
                  On this slide is an overview of the available environmental hazard data that were identified for 1-BP under the conditions of use of the assessment.  1-BP is reported to present a moderate hazard to fish, aquatic invertebrates, and algae.  Based on the available data, fish, or rainbow trout, are the most sensitive aquatic species.  As part of the systematic review process, EPA identified and reviewed one acute fish toxicity study for 1-BP, Geiger et al. 1988.  This study was determined to be of high quality following data evaluation and indicated that 1-BP presents a moderate hazard to aquatic receptors.  
                  Five aquatic toxicity data summaries were identified in a European Chemicals Agency, or ECHA, database.  No U.S. sponsor was identified, and the full study reports were not obtained by EPA.  As a result, the data quality of these studies is uncertain.  In an attempt to utilize all available data, the results of these studies were used in the screening level assessment of risks to aquatic species and are consistent with the results of the Geiger et al. acute fish study.  As reported in the Geiger et al. study, these studies indicate that 1-BP presents a moderate hazard to aquatic species following acute exposure.  
                  No data were identified to characterize hazards to aquatic species from chronic exposure to 1-BP.  EPA estimated environmental hazard via chronic exposure by utilizing an acute to chronic ratio where an uncertainty factor of ten was applied to the acute toxicity data.  The physical-chemical properties of 1-BP indicate that it is expected to volatilize quickly from water.  Therefore, the potential for chronic exposures is limited.  
                  This slide outlines the approach EPA used to calculate an environmental hazard threshold and a screening level assessment of potential risks to aquatic species.  EPA calculated hazard thresholds known as concentrations of concern, or COCs, to provide a conservative estimate for use in the screening level assessment.  These COCs were calculated by applying an assessment factor to the most sensitive environmental hazard data according to EPA methods.  The assessment factor provides a lower bound affect level and accounts for differences in intra- and inter-species variability, as well as laboratory-to-field variability.  For fish and aquatic invertebrates, for example daphnia, the acute COC values were calculated by dividing the selected environmental hazard endpoint by an assessment factor of five.  For chronic COCs, an assessment factor of ten is used.  The EPA calculated these COCs using all aquatic toxicity endpoints.  According to the available data, the freshwater fish or rainbow trout study reported in the ECHA database represented the most sensitive species.  
                  This section discusses the environmental risk characterization for 1-BP.  This is discussed in Section 4.1 of the document and also relates to Charge Question 4.  This slide presents the results of the screening level analysis for the environmental risks to aquatic receptors.  Environmental risks are estimated by calculating a risk quotient, or RQ, which is defined as the quotient of the estimated environmental concentration and the affect level threshold for the taxa of interest.  In this case, the COCs for aquatic species were calculated by comparing the most sensitive aquatic COCs with the estimated environmental concentrations predicted using EPA's EFAST screening tool and releases reported in the TRI database.  
                  An RQ equal to one indicates that the exposures are the same as the concentration that causes effects.  If the RQ exceeds one, the exposure is greater than the concentration, and there is potential for risk presumed.  If the RQ does not exceed one, the exposure is less than the affect concentration and there is no risk presumed.  Based on the available data for aquatic species, all aquatic RQs were less than one, indicating risks were not identified for aquatic species under the conditions of use in this assessment.  
                  As we discussed in an earlier slide, EPA did not quantitatively evaluate potential risk to terrestrial environments or risks to aquatic species in sediment.  High volatility, high water solubility, and low log KOC indicate that 1-BP will only be present in terrestrial environmental compartments as a vapor and will not absorb to sediment in aquatic environments.  No specific conditions of use, that is systematic application to land, were identified that resulted in significant airborne exposures that overlap with terrestrial habitats.  As a result, airborne exposures are not a relevant route of exposure for 1-BP under the conditions of use in this risk evaluation.  Additionally, 1-BP is not expected to bioaccumulate, and high volatility from water means that exposure to terrestrial species through oral routes is negligible.  As a result, no quantitative analysis of risk to terrestrial receptors or sediment dwelling aquatic invertebrates was carried out as part of this evaluation, as risks from these pathways are not expected.  
                  This next slide outlines several of the uncertainties discussed in the document that relate to the environmental risk conclusion.  The use of proxy organisms to represent all organisms within a specific taxa and environment can be an oversimplification.  One study, an acute fish toxicity study, was evaluated for data quality.  No U.S. sponsor was identified for the study summarized in the ECHA database, so EPA did not obtain these studies.  And they have not been evaluated for data quality.  EPA has made copies of these study summaries available in the docket for consideration by the committee.  
                  Despite this, EPA incorporated all available data in the screening level analysis to thoroughly characterize potential risks to aquatic receptors from 1-BP.  The acute fish toxicity study evaluated for data quality and the results reported in ECHA are consistent in that they indicate that 1-BP presents a moderate hazard to aquatic species.  And use of these endpoints in the screening level environmental risk analysis did not identify risks to aquatic species.  
                  Environmental exposure data are based on estimates form the 2016 TRI release data and cannot be attributed to a specific condition of use.  However, as discussed above, a consideration of subsequent reporting to TRI indicates that these releases are similarly low and are not expected to result in risk concerns for aquatic species.  To summarize, qualitative and quantitative assessments of reasonably available environmental hazard data indicates that EPA does not find risks to the environment from the conditions of use for 1-BP.  
                  I will now transition into a discussion of human health, including the exposures, hazards, and risk characterization.  I will begin the discussion of human health with occupational exposure.  This information can be found in Section 2.3.1 of the draft risk evaluation and relates to Charge Question 2.  
                  There are several objectives to the occupational inhalation exposure assessment.  First objective is to estimate the number of potentially exposed workers and occupational non-users.  The second objective is to assess the inhalation exposures for these human receptors.  We assess central tendency and high-end exposure concentrations for workers and occupational nonusers.  Where possible, we consider the effect of engineering control on occupational exposure.  For each exposure scenario, EPA calculated the acute and chronic exposure concentrations.  This includes the acute concentration, AC; the average daily concentration, ADC; and the lifetime average daily concentration, LADC.  
                  To assess inhalation exposure for workers and occupational nonusers, EPA used both monitoring data and modeling approaches.  Where monitoring data are available, EPA used 1-BP personal breathing zone monitoring data from directly applicable scenarios.  Key sources of monitoring data came from NIOSH, OSHA, and the EPA SNAP program and the open literature.  These data were determined to be of either high or medium quality through systematic review.  Where EPA has information to construct a model, exposure modeling is performed to supplement monitoring data.  The model may be either deterministic or probabilistic.  
                  This slide is a summary of EPA's inhalation exposure assessment approach for each occupational condition of use.  As explained in the previous slide, EPA used a combination of monitoring data and modeling approaches.  For each condition of use, monitoring data are always used to assess exposure if they are available.  Personal breathing zone monitoring data are available for eight conditions of use.  Exposure modeling was performed for 12 conditions of use using six different inhalation exposure models.  Five of the six models were probabilistic in nature.  These five models were initially developed for the 2016 draft risk assessment of 1-BP.  Certain aspects of each model has been further refined to address CSAC peer review comments from 2016.  EPA developed a new deterministic model to assess exposure during container loading activities.  This model is used to assess exposure for import, certain industrial processing, repackaging and disposal conditions of use.  
                  For several conditions of use, EPA used a two-zone probabilistic modeling approach to assess inhalation exposure.  In this modeling approach, 1-BP emissions occur in the near field zone.  Workers are assumed to spend their time in the near field when handling 1-BP.  Occupational non-users are assumed to spend their time in the far field.  The near field is typically assumed to be a space with dimension of ten feet by ten feet by six feet.  Far field size is specific to each exposure scenario and varies on the models.  
                  For vapor degreasing, the far field volume is defined as a triangular distribution with a mode of approximately 17,000 cubic feet or 500 cubic meters.  Air exchange occurs between the near field and far field.  Model input parameters, such as the far field size, work activity pattern, 1-BP use rate, and environmental parameters, were defined using reasonably available data from the literature.  The EPA performed a Monte Carlo simulation to capture variability within the modeled scenario.  
                  The next few slides provide examples of acute eight hour time weighted average, or TWA, inhalation exposure concentrations determined using either monitoring data or modeling approaches.  For the conditions of use shown on this slide, such as manufacture and import, the acute inhalation exposure concentrations are below 0.3 part per million for both central tendency and high-end exposures.  This slide shows the exposure concentrations for open top batch vapor degreasing.  For this condition of use, exposures were assessed using both monitoring data and modeling.  Exposure results between the two approaches are within one order of magnitude.  
                  This slide shows the exposure concentrations for cold cleaning.  For this condition of use, exposures were assessed using both monitoring data and modeling.  The central tendency and high-end exposure estimates were found to be in good agreement among the two approaches.  
                  There are some uncertainties and limitations with the occupational inhalation exposure assessment.  For monitoring data, 1-BP personal breathing zone monitoring data are available for eight conditions of use.  However, certain conditions of use or job categories may have limited number of datapoints.  The representativeness of the monitoring data is generally not known.  For modeling, exposure modeling results are in general agreement with monitoring data where these data are available.  The modeling approach captures variability in the exposure scenario.  However, there is some uncertainty associated with each model input parameter, which is determined based on reasonably available information from the literature.  The model assumes each zone is well-fixed and that all emissions enter the near field.  In reality, it is possible that some emissions do not enter the airspace as relevant to worker exposure modeling.  
                  For dermal exposure, EPA assessed exposure to workers using the dermal exposure to volatile liquids model.  The equation of this model is given below.  Dermal exposure was not previously evaluated in the 2016 draft risk assessment, but a quantitative assessment was recommended during the CSAC 2016 peer review.  EPA did not assess dermal exposure to occupational nonusers.  EPA assumes that occupational nonusers do not directly handle 1-BP and are not exposed through the dermal route.  
                  The model is a modification of the peer-reviewed EPA/OPPT two-hand dermal exposure to liquids model.  This model version accounts for evaporation by incorporating a fraction of absorbed parameter.  This fraction absorbed value is based on experimental measurement from a 2011 dermal study conducted by Frasch.  The model also accounts for potential glove use by incorporating a glove protection factor, or PF values, from the European Center for Ecotoxicology and Toxicology of Chemicals' targeted risk assessment model.  EPA recently received comments on this model during the 1,4-dioxane peer review.  These comments have not yet been incorporated into the 1-BP assessment.  We look forward to receiving final written comments and specific recommendations on this model.  
                  This slide provides examples of dermal exposure estimates.  In the dermal exposure assessment, conditions of use were grouped or binned based on characteristics that are known to affect dermal exposure.  These bins are groups of multiple conditions of use and are different from exposure scenarios.  The slide presents example dermal exposure estimates for Bin 1, which covers large scale industrial uses that typically occur in mostly closed systems.  For these uses, dermal exposure is likely limited to loading and unloading activities and quality control sampling.  
                  There are some -- excuse me.  There are some uncertainties and limitations with the occupational dermal exposure assessment.  The assessment uses a fixed fractional absorption approach to quantify dermal dose.  In reality, dermal absorption may depend on skin loading conditions.  The model also assumes a single exposure event per day and does not address variability in exposure duration and frequency.  Further, there is limited information on glove type and glove use for most conditions of use.  Therefore, the actual exposure reduction from glove protection is uncertain.  
                  Moving on from occupational exposure, I will now focus on consumer exposures.  This information can be found in Section 2.3.2 of the draft risk evaluation and relates to Charge Question 3.  This slide provides a broad overview of the consumer uses, routes of exposure, and models used in the 1-BP risk evaluation.  Nine consumer uses were evaluated for 1-BP.  Two routes of exposure were evaluated, and three models were used as described in this table.  
                  Model selection was based on the availability of best fit predefined scenarios within each model for the consumer use evaluated.  For example, the consumer exposure model, or CEM, does not have a predefined scenario which aligns well with modeling inhalation exposure from a solvent pool.  While the multi-chamber concentration and exposure model, or MCCEM, has two predefined scenarios that align well with inhalation exposure from a solvent pool.  Based on the consumer use patterns identified for coin and scissors cleaner and an automobile AC flush scenario, the MCCEM's two inhalation exposure from a solvent pool aligned well with these two scenarios.  And therefore, we utilized the MCCEM for the inhalation exposure modeling.  
                  This flow chart depicts the general approach taken in the 1-BP risk evaluation for acute consumer exposure.  Acute inhalation exposure was evaluated for the consumer user and bystander.  Acute dermal exposure was evaluated for the consumer user, but only for three select scenarios where a continuous supply of product containing 1-BP against the skin could occur during use.  These select scenarios were depicted on the previous slide and include general cleaners and degreasers, coin and scissors cleaner, and automobile AC flush.  
                  Dermal exposure to the bystander is not expected due to the high volatility of 1-BP and therefore was not evaluated.  Neither inhalation or dermal exposure to the consumer user or bystander is expected to be chronic in nature, and therefore chronic exposure was not evaluated.  To avoid confusion as we continue through the slides, it is important to note that with regards to the consumer exposure results, inhalation exposure is presented as a concentration, milligrams per cubic meter converted to parts per million, and not a dose.  The concentration in the air is independent of age and other physical characteristics of the receptor.  Dermal exposure is presented as a dose and therefore is dependent on age and other physical characteristics of the receptor.  
                  Exposure was evaluated for both consumer use and bystander within a residential setting.  Consumer user was assumed to include youth and adults.  To further inform the reviewers, for this evaluation the youth group is further broken down within the CEM and MCCEM into two age groups: 11 to 15 and 16 to 20 years of age.  Adult is 21 plus years of age.  As noted on the previous slide, this is only a factor when looking at the three of the nine scenarios where dermal exposure was evaluated since the concentration related to inhalation exposure is independent of age.  Consumer bystander is not limited to any one age bracket for this evaluation, as any additional individual that is not a user present within the residence at the time of product use could be any age, infant to elderly.  Since bystanders are only evaluated for inhalation exposure because dermal exposure to the bystander is not expected under the conditions of use evaluated, these age references are only information to consider possible ages exposed.  
                  The image on this slide depicts the two-zone configuration utilized for the CEM for this evaluation, with Zone 1 representing the room of use and Zone 2 representing the rest of the house.  Zone 1 is further broken down by selecting the near field/far field options within the model.  The Zone 1 near field volume for this evaluation was set at one cubic meter, regardless of the room of use.  Zone 1 far field is the remaining volume of the room of use and can vary depending on the identified room of use.  Zone 2 volume varies depending on the room of use volume.  However, the whole house volume was held constant at 492 cubic meters for all consumer scenarios evaluated.  The arrows within the image on this slide depict the interzonal airflows and the direction of the airflows.  
                  This slide provides a broad overview of key concepts within the methodology used for this evaluation when assessing exposure using the CEM.  The CEM was used to model inhalation exposure for six of the nine consumer scenarios evaluated and all three dermal exposure scenarios evaluated for 1-BP.  The model profile utilized for inhalation exposure within the CEM for this evaluation considers an overspray fraction of the product, which is immediately available for uptake via inhalation, as well as that portion of the product evaporating from the surface over time.  For this evaluation, an overspray fraction of six percent was assumed for all scenarios.  The model profile utilized for dermal exposure within the CEM for this evaluation utilized a permeability method for consumer exposure.  The permeability coefficient used for this modeling was 0.00905 centimeters per hour.  
                  Also shown on this slide, in order to capture a broad spectrum of consumer use patterns, this evaluation varied three key parameters for inhalation and two key parameters for dermal exposure.  Values for duration of use and mass of product used were obtained from an EPA directed survey titled "Household Solvents Products: A National Usage Survey."  This survey received a high quality rating under the systematic review process and provided, among other information, duration of use, mass of product used, and rooms of use for 32 different product groups.  Many of the product groups from the survey crosswalk well with consumer scenarios evaluated for 1-BP.  
                  The amount of chemical in the product or weight fraction for each scenario was obtained from product safety datasheets when such information was available.  Modeling was conducted for all possible combinations of the varied parameters for each condition of use, resulting in a maximum number of 27 inhalation modeling runs and nine dermal modeling runs.  Results are presented for three of the modeling runs as defined on this slide, high, moderate, and low intensity use, to represent a low, mid, and high-end exposure concentration estimate.  
                  This slide provides a broad overview of key concepts within the methodology used for this evaluation when assessing exposure using the MCCEM.  The MCCEM was used to model inhalation exposure for two of the nine consumer scenarios evaluated for 1-BP, coin and scissors cleaner and automobile AC flush.  The model profile utilized for inhalation exposure within the MCCEM for this evaluation considers evaporation from a solvent pool with a fixed surface area.  The model profile utilized assumes a constant emission rate over the entire duration of product use.  
                  This slide also depicts the vessels within which the solvent pool was maintained during product use.  A bowl was assumed for coin cleaning, while a bucket was assumed for capturing spent automobile flush product.  For the two scenarios modeled, the emission rate is governed by the surface area and not chemical mass.  Therefore, the methodology used for 1-BP only varied one parameter: duration of use.  The remaining methodology is similar the CEM methodology, including a two-zone configuration, utilizing the near field/far field for the room of use, considering user and bystander inhalation exposures, and considering a whole house volume of 492 cubic meters.  
                  The insulation scenario model for 1-BP evaluated inhalation exposure to the bystander.  This exposure occurs due to off-gassing of 1-BP from the insulation product.  The general mass balance equation within the indoor environment concentrations in building with conditioned and unconditioned zones model, or IECCU, has six components affecting the change in concentration with any given zone.  However, due to the high volatility of 1-BP, only three of the six components were considered for this evaluation, including emissions from the sources in the zone, rate of chemical removed from zone by ventilation or interzonal airflows, and rate of chemical carried into the zone by infiltration and intrazonal airflows.  The three remaining components, rate of chemical absorption by interior surface, rate of chemical absorption by airborne particles, and rate of chemical absorption by settled dust were not considered, as it is assumed that once 1-BP is in the vapor phase it will remain in the vapor phase.  
                  The picture on this slide depicts the three zone configuration considered for this evaluation, and the arrows show the various interzonal flows between the three zones.  We used a three zone configuration for insulation as the product is assumed to be installed in two locations, attic and crawlspace, rather than used in one room of the house as with the CEM and MCCEM scenarios.  The modeled scenario assumes insulation is installed in the attic and crawlspace of a residence.  However, the only results presented in this draft risk evaluation were for the living space, as it is not expected that an individual bystander will live or spend any considerable amount of time in the attic or crawl space.  
                  This slide provides some information related to the sensitivity analysis conducted on the CEM during development.  A full description of the sensitivity analysis and the methodology used is provided in the CEM user guide appendices.  The link to the appendices is provided on this slide.  As discussed previously, to capture a broad spectrum of consumer use patterns, this evaluation varied key parameters for modeling: three for inhalation exposure and two for dermal exposure.  These parameters were systematically selected because they represent expected consumer behavior patterns.  The models used are highly sensitive or sensitive to these parameters, and EPA had published survey and safety datasheet information on these parameters, including the duration of use, mass of product used, and mass of chemical in product or weight fraction. While the models used are also sensitive to other data like room and building volume and air exchange rates, EPA assumes in any consumer residential setting these parameters should be relatively constant across building types where ventilation systems are properly design, maintained, and balanced.  
                  As with any risk evaluation process, there are uncertainties and limitations associated with the assumptions, modeled data and approaches used.  While all uncertainties and limitations could not be eliminated in this evaluation, efforts were made to reduce some of the overarching uncertainties associated with this evaluation.  Some of these efforts are summarized on this slide.  All three models used for this evaluation are peer reviewed and applied in the manner they are intended to be used.  The sensitivity of the CEM to certain parameters was considered in developing the overall modeling approach for this evaluation.  Three key parameters were varied.  Other parameters to which the models are sensitive, such as room house volume, Zone 1 and 2 air exchange rates were assumed to be consistent across typical residences and therefore were not varied for this valuation.  Instead, the default values within the models were used which were central tendency values.  While current tools utilized for the consumer exposure evaluation are deterministic in nature rather than probabilistic, multiple modeling runs were done across a spectrum of possible consumer use patterns and iterations to explore variability across the expected consumer scenarios evaluated.  
                  Some uncertainties and limitations remain, even if the impact may have been reduced through the efforts discussed on the previous slide.  The representativeness of consumer use pattern information, duration of use, mass used, and weight fraction utilized for this evaluation may result in an underestimate or overestimate of the actual exposure.  The math used for the present day may be less than the amount needed in the past, due to reformulations and the availability of highly concentrated product.  Similarly, the duration of use may be less for the present day than in the past because a highly concentrated product may achieve the same or better results in a shorter period of time.  This could result in an overestimation of the exposure.  Alternatively, the consumer movement toward more do-it-yourself projects could result in more consumers using more product over a longer duration for the present day than in the past.  This could result in an underestimation of the exposure.  
                  The representativeness of the crosswalk between the consumer uses evaluated and the product categories within the household product survey utilized for this evaluation may have some uncertainty.  However, while some aspects of the way a product is used may have changed, for example, an electronics cleaner may have been used to clean the VCR heads in the past but now is used to clean a DVD reading eye, computer motherboard, or the contacts in PS4 cartridges, the general product use category is still representative of consumer use categories evaluated.  The lack of monitoring data against which the modeled results can be compared is an unavoidable limitation for this evaluation, as such monitoring information was not identified during literature searches and the systematic review process.  This limitation does not mean the model results presented are unreliable or unrealistic or deficient.  It simply means, in the absence of monitoring data, modeled results are the best information from which we can characterize risk.  
                  I will now discuss the human health hazard associated with 1-BP.  This information can be found in Section 3.2 of the draft risk evaluation and relates to Charge Question 5.  This slide depicts the process EPA used to evaluate, extract and integrate 1-BP's human health hazard and dose response information.  This is similar to the slide shown previously for the environmental hazard.  The process is the same for human health, with the exception that the hazard ID and weight of evidence is followed by dose response analysis rather than a COC derivation.  
                  For this risk evaluation, all of the known human health hazard domains of 1-BP were described, reviewed, and evaluated, including toxicity following acuity exposure, liver and kidney toxicity, immunotoxicity, reproductive and developmental toxicity, neurotoxicity, genetic toxicity, and carcinogenicity.  Based on the review and integration of the best available data on human health hazards and using a strategy that included data quality evaluation, mode of action information and overall weight of the scientific evidence approach across studies, liver and kidney toxicity, reproductive and developmental toxicity, neurotoxicity and cancer were selected forward for dose response analysis.  EPA determined endpoints relevant for both acute and chronic exposures.  EPA determined that developmental endpoints were relevant to acute exposures.  EPA selected endpoints from each hazard domain considered relevant for chronic exposures.  And again, these included liver and kidney toxicity, reproductive and developmental toxicity, neurotoxicity and cancer.  
                  The developmental toxicity endpoints, points of departure, or POD's a reduced number of live pups per litter and increased post-implantation lose, were selected as the endpoints for assessing risks following acute exposure.  This endpoint was considered to be representative of a robust dataset supported by the general consistency of the effects reported across studies, which exists along a continuum of adverse and developmental and reproductive effects, including mortality, all occurring within a short window of exposure between ovulation and implantation.  While there is some uncertainty surrounding this consideration of applying this endpoint to acute exposures because the precise critical window of exposure is unknown, multiple publications suggest that some developmental effects may result from a single exposure during a critical window of development.  Therefore, without specific information on the mechanism of action or developmental windows of sensitivity for these specific developmental affects, EPA considered these developmental toxicity endpoints to be applicable to acute exposures.  
                  As previously acknowledged, all endpoints associated with the selected hazard domains of liver, kidney toxicity, reproductive and developmental toxicity, neurotoxicity, and cancer were considered for chronic exposure.  Overall, there is a high confidence in all endpoints selected as PODs for both acute and chronic exposure scenarios.  These endpoints were derived from three studies.  The studies were selected because they all scored high in data evaluation, were of longer duration, and the effects were observed more consistently than in other high quality studies.  In addition, these endpoints were identified as the most important endpoints relevant to acute and chronic exposures and were also the lowest available PODs.  And therefore, they would be protective of effects observed in other studies.  
                  EPA used benchmark dose modeling for the point of departure of decreases in live litter size as the most sensitive and biologically relevant developmental toxicity endpoint.  And it is considered to be representative of a robust dataset representing a continuum of adverse developmental outcomes such as pre- and post-implantation loss.  In an attempt to account for the litter effect, EPA used the BMDS nested dichotomous model, or NCTR.  However, this model can only be applied to increases in effects.  And therefore, increased post-implantation loss was the endpoint selected as the most relevant for calculating risks associated with developmental toxicity following acute and chronic exposures.  A BMR of 5 percent was used for decreases in live litter size because it is a combination of effects where a BMR of 10 percent would be used and post-implantation loss, which is considered a more severe effect like mortality, where a BMR of 1 percent would be used.  A BMR of 1 percent was used for the increases in post-implantation loss, which is considered a severe effect like mortality. 
                  While the majority of exposure pathways for 1-BP are occurring via inhalation and inhalation exposure is the most important, during the CSAC peer review of 2016 draft risk assessment, it was recommended that dermal exposures might be an important contributor to overall exposure and an estimate for dermal exposure should be included in the evaluation with all of the gaps, limitations clearly stated to address another potential workplace exposure pathway.  There are no toxicity studies for 1-BP by the dermal route.  The few toxicity studies available by the oral route are very limited, and there is no developmental toxicity study by the oral route.  There is a single chronic oral study, but it is not written in English.  Non-English language studies are excluded from systematic review in the population exposure comparator and outcome, or PICO, statement, and describes the study eligibility criteria that guides inclusion and exclusion decisions during systematic review.  
                  In addition, physiologically based pharmacokinetic, pharmacodynamic, or PBPK/PD, models that would facilitate route to route extrapolation have not been identified.  And there are no relevant kinetic or metabolic information for 1-BP that would facilitate development of dosimetric comparisons.  One second.  I'm having trouble with my mouse here.  Therefore, for route to route extrapolation for determining PODs by the dermal route in a noncancer assessment, EPA used the two generation reproductive developmental toxicity study by the inhalation route instead of the limited oral data.  
                  EPA benchmarked dose modeled data for all endpoints when possible.  Adult animal PODs were adjusted from study conditions for both acute and chronic exposures to calculate human equivalent concentrations, or HECs, and human equivalent doses, or HEDs, consistent with EPA guidance on allometric scaling.  The human equivalent dosimetric adjustment factor is the ratio between the animal and human blood to air partition coefficients.  For 1-BP, the blood to air partition coefficient for rats is greater than that for humans, so a default ratio of one was applies.  
                  Also consistent with EPA guidance, uncertainty factors were applied where appropriate.  For example, typically an uncertainty factor of ten is used to extrapolate a POD from a less than chronic study to a chronic exposure.  The same is true for a developmental toxicity study because the developmental period is recognized as a susceptible life stage where exposure during certain time windows is more relevant to the induction of developmental affects than lifetime exposure.  In this assessment, endpoints from chronic studies were available, and the uncertainty factor was reduced from ten to one for all of the HECs and HEDs.  
                  The next uncertainty factor accounts for uncertainties in extrapolation from rodents to humans.  In this assessment, a portion of the toxicokinetic uncertainty may be accounted for by the calculation of an HEC, an application of a dosimetric adjustment factor outlined in the RC methodology.  However, an uncertainty factor of ten is retained to account for additional toxicokinetic difference that remain unaccounted.  1-BP is irritating to the respiratory tract, and rodents exhibit physiological responses that differ from humans and may alter uptake due to hyper- or hypoventilation.  Therefore, an uncertainty factor of ten is retained to account for pharmacokinetic differences.  
                  The next uncertainty factor accounts for the variation in sensitivity within the human population.  In the absence of data, the default uncertainty factor of ten is adopted, which breaks down to a factor of three for toxicokinetic variability and a factor of three for pharmacodynamic variability.  Since there is no PVK model for 1-BP to reduce the human toxicokinetic/toxicodynamic variability, the total uncertainty factor of ten was retained.  
                  The last uncertainty factor accounts for the uncertainty in extrapolating from a LOAEL to a NOAEL.  A value of ten is a standard default value when a LOAEL is used, although lower values, for example, three, can be used if the effect is considered minimally adverse at the LOAEL or is an early marker for an adverse effect.  And a value of one is used when the POD is a NOAEL or a BMDL.  In this assessment, the default value was reduced from ten to one because a NOAEL or BMCL was used for all but one of the HECs or HEDs.  
                  For one of the reproductive PODs, the Amata et al. 2003, an uncertainty value of ten was used based on a minimally adverse effect, which resulted in a total uncertainty factor of 1,000.  However, this was not one of the selected PODs.  On this slide are the most sensitivity PODs representing each noncancer endpoint for acute and chronic exposures that EPA used for risk estimation, along with the benchmark MOEs that were derived based on the combination of uncertainty factors as described on the last slide.  
                  Now, moving on to the cancer assessment.  Significant increases in the incidence of skin tumors in male rats, rare large intestinal adenomas in female rats, and alveolar or bronchiolar adenomas or carcinomas in female mice were observed following exposure to 1-BP via inhalation in an NTP two-year bioassay.  NTP concluded that these data show some evidence for carcinogenicity in male rates, clear evidence for carcinogenicity in female rats, no evidence for carcinogenicity in male mice, and clear evidence for carcinogenicity in female mice.  No other laboratory animal data or human data were located on the carcinogenicity of 1-BP.  
                  In 2019, IARC concluded that 1-BP is possibly carcinogenic to humans, group 2B, based on the following characteristics: strong evidence that 1-BP is electrophilic or can be metabolically activated through reactive intermediate; strong evidence that 1-BP induces oxidative stress, induces chronic inflammation, and is immunosuppressive; moderate evidence that 1-BP modulates receptor mediated effects and is genotoxic; and sufficient evidence in experimental animals for the carcinogenicity of 1-BP.  Taken together, the available evidence supports that 1-BP is a multisite carcinogen in rats and mice.  By the criteria presented in EPA's guidelines for carcinogen risk assessment, 1-BP may be considered likely to be carcinogenic in humans based on the positive findings for carcinogenicity in more than one test species together with positive findings for the direct reactivity of 1-BP with DNA and evidence that both 1-BP and its metabolites are positive in immunogenicity studies and other types of studies that assess genotoxicity.  
                  Following EPA's guidelines for carcinogen risk assessment, the overall weight of the scientific evidence supports immunogenic mode of action for 1-BP induced carcinogenicity, including demonstrated alkylating activity, evidence of DNA damage in leukocytes taken from 1-BP exposed workers, demonstrated mutagenicity with metabolic activation in cultured mammalian cells, and in the AIMS assay, when test conditions included use of a closed system design for testing volatile chemicals, evidence indicated 1-BP may form DNA adducts in vitro, evidence suggesting 1-BP may be a direct acting mutagen including demonstrated mutagenic activity in bacteria and mammalian cells in the absence of metabolic activation, and evidence indicating 1-BP metabolites, for example, glycidol and propylene oxide are mutagenic in multiple test systems, for example, bacteria, yeast, drosophila, and mammalian cells.  
                  Other lines of evidence do not provide clear support for a mutagenic mode of action for 1-BP carcinogenicity, including recent data for 1-BP using transgenic big blue female mice following whole body inhalation exposure to 1-BP.  1-BP vapor did not induce mutations.  However, there are several methodological limitations in test design that limit the utility of these studies.  This is discussed in detail in the draft risk evaluation.  Other possible mode of actions, in 2013, NTP suggests that in addition to mutagenicity at least three other mechanisms, including oxidative stress, immunosuppression, and cell proliferation can contribute to the multistage process of carcinogenesis, although more research, for example, organ specific in vivo DNA adduct studies, is needed to ascertain mutagenicity as a key molecular band.  There is no definitive evidence which suggests that these other mechanisms play a more important role than mutagenicity.  
                  Although the exact mechanism and mode of action of 1-BP carcinogenesis is not established, reasonably available information and the weight of evidence analysis for the cancer endpoint was sufficient to support a mutagenic mode of action for 1-BP carcinogenesis.  Given the lack of specific mechanistic or dose response information, linear extrapolation from the point of departure to derive the inhalation unit risk, or IUR, is recommended per EPA guidelines of carcinogen risk assessment.  Benchmark dose modeling of the NTP cancer data was performed for all three statistically significantly increased tumor types, that is skin tumors in male rates, intestinal tumors in female rats, and lung tumors in female mice.  
                  In addition to the multistage modeling, two models averaging methods were applied, Frequentist and Bayesian, to assess the impact of model uncertainty.  The resulting model average benchmark concentrations associated with 0.1 percent added risk and their 95 percent lower confident limits are shown in Table 3-2 of the draft risk evaluation, the Frequentist model averaging column for each of the three cancer datasets.  Since the 2016 draft risk assessment, EPA has conducted additional modeling using the BMDS version 3.0.  More details are available in the supplemental file information on human health benchmark dosing.  
                  All dichotomous Frequentist and Bayesian models in the BMD software were fit to the incident data for each of the three tumor types.  The benchmark response levels used were 0.1 and 10 percent added an extra risk.  The BMR used in the 2016 draft risk assessment was 1 percent, 0.1 percent added risk.  
                  As mentioned on the previous slide, EPA derived the inhalation unit risk based on lung tumors in female mice.  Specifically, these are alveolar, bronchiolar, adenoma or carcinoma shown in the last row of the table.  On this slide, the benchmark concentrations, BMCs, and their 95 percent lower confidence limits, BMCLs, are shown for exposure durations of 40 hours per week, representing a typical worker, and 24 hours per day, representing a continuous exposure.  The results of the three modeling approaches are shown: the multistage modeling, model averaging with the Frequentist model, and model averaging with the Bayesian model.  
                  The BMCs and BMCLs are shown, the multistage columns for each of the three cancer datasets.  Also, the results of a BMR of 0.1 percent added risk are presented for comparison.  In the multistage column for each of the three tumor types from the NTP study, the BMR shown is 10 percent extra risk.  In the Frequentist model averaging column, the BMR shown is 0.1 percent added risk.  In the Bayesian model averaging column, the BMR shown is 10 percent extra risk.  The multistage and Bayesian model averaging approaches have the same BMR, 10 percent extra risk, and similar BMDs and BMDLs showing the modeling approach does not have a large impact on the BMD and BMDL values. 
                  Overall, the lung tumors are more sensitive than the other tumor types.  This endpoint was used for calculating cancer risk.  The BMCL of 25 parts per million from the multistage model was used to derive the cancer risks for a 40 hour per week exposure.  On this slide, the BMDs and BMDLs are shown for dermal exposures extrapolated from the inhalation BMCs and BMCLs for each of the three cancer datasets for the three modeling approaches for comparison.  The BMCL of 17 milligram per kilogram per day from the multistage model was used to derive the cancer risks.  
                  Strength of the human health hazard assessment, EPA considered adverse effects for 1-BP across organ systems.  The full list of effects was screened to those that are relevant, sensitive, and found in multiple studies, which include the following types of effects: liver and kidney toxicity, immunotoxicity, developmental and reproductive toxicity, neurotoxicity, and cancer as previously described.  In general, adverse effects were observed in all these systems in rats exposed to 1-BP by inhalation in the range of 100 to 1,000 PPMs for the LOAELs.  
                  Dose response modeling was conducted for decrease litter size and post-implantation loss.  Decreased litter size represents a continuum of effects, including post-implantation loss.  The similarly in BMCLs between the less specific litter size and more specific post-implantation loss provide confidence in the selected POD.  
                  Cancer effects were seen across two species, two sexes, and three target sites.  EPA conducted a weight of evidence consideration for 1-BP's carcinogenesis, which evaluated in vitro tests such as AIM studies, metabolic pathway information for animals and humans, structure activity relationships, and considered multiple mechanisms of action.  Dose response modeling was conducted for tumor incidence by three approaches, including two model averaging approaches and comparing BMRs of 0.1 and 10 percent.  The similar results between the different approaches and different BMRs supports that the selected BMCLs are not significantly impacted by the choice of modeling approach and provides confidence in the selected BMCL.  
                  While there are many strengths to the human health hazard assessment, there are some assumptions, uncertainties, and limitations.  For assumptions, EPA considered some developmental toxicity endpoints observed in a repeat dose two generation reproductive toxicity study applicable to acute exposures.  Exposures during other life stages, such as in childhood, may cause similar or related effects.  Without specific information on the mechanism of action or developmental windows of sensitivity for these specific developmental effects, there are uncertainties in extrapolating these effects for other life stages in order to refine dose estimates for these additional life stages.  
                  Uncertainties and limitations, oral repeated dose toxicity studies are insufficient for a quantitative dose response assessment.  However, data from these studies were used as qualitative support in the weight of scientific evidence for nervous system effects.  The data suggests that, at least for the nervous systems endpoints, the delivery of 1-BP via the inhalation and oral routes of exposure result in comparable toxic endpoints.  
                  PBPK and PD models that would facilitate route-to-route extrapolation have not been identified.  EPA chose to derive dermal HEDs for dermal exposures by extrapolation from the inhalation PODs.  EPA was unable to conclude with certainty that comparable toxic endpoints would be associated with the dermal route of exposure.  
                  Now, we will move into the human health risk characterization portion of the presentation.  This information can be found in Section 4.2 of the draft risk evaluation and relates to Charge Questions 6 and 7.  EPA estimated risk to workers, including occupational non-users, and consumers, including bystanders.  EPA considered PESS, including women of reproductive age and individual life stages, such as infants and toddlers.  EPA separately evaluated risks following acute and chronic exposures by both the inhalation and dermal routes.  
                  EPA used a margin of exposure, or MOE, approach.  The MOE is the ratio of the point of departure to the exposure and is compared to the benchmark MOE, which is defined by the aggregate of all assigned uncertainty factors.  When the MOE is below the benchmark MOE, risk is indicated.  When the MOE is above the benchmark MOE, risk is not indicated.  
                  EPA determined cancer risk for workers and occupational nonusers by multiplying the lifetime average daily concentration by the inhalation unit risk, or IUR.  For example, the high end, or 95th percentile, cancer risk is the 95th percentile exposure concentration multiplied by the IUR.  For most conditions of use, EPA assumes workers and occupational nonusers are exposed eight hours per day, 260 days per year, and either 31 working years for the central tendency exposure scenario or 40 working years for the high end exposure scenario.  Where exposure is modeled, the number of working years is varied as part of the probabilistic simulation.  
                  EPA typically uses a benchmark cancer risk level of one times ten to the minus four for determining the acceptability of the cancer risk in a population.  Since the benchmark cancer risk level will be determined during risk management, the occupational estimates for excess cancer risk were compared to the benchmark levels of ten to the minus four, minus five, and minus six incremental or extra individual lifetime risk.  The benchmark levels were one times ten to the minus six, which is the probability of one chance in one million of an individual developing cancer; one times ten to the minus five, which is the probability of one chance in one hundred thousand of an individual developing cancer, which is equivalent to ten cancer cases in a million; and one times ten to the minus four, which is the probability of one chance in 10,000 of an individual developing cancer, which is equivalent to 100 cases of cancer in one million.  
                  In summary, risk for workers and occupational nonusers were estimated following acute and chronic exposures via inhalation and dermal routes of exposure for workers and inhalation routes of exposure for occupational nonusers using monitoring data and modeling.  Risk estimates were presented for central tendency and high end scenarios.  Risk to consumers were estimated following acute exposures via inhalation and dermal routes using modeling.  Risk estimates were presented for low, moderate, and high intensity use scenarios.  
                  This slide summarizes the noncancer risk for workers and occupational nonusers following acute and chronic exposures.  The left hand side lists conditions of use where the risk estimates for noncancer PODs were above the benchmark MOE, indicating no risk.  For most of these conditions of use, risk estimates were above the benchmark MOE for both central tendency and high end exposure scenarios, even when respirators were not considered.  
                  The righthand side lists conditions of use where the risk estimates for noncancer PODs were below the benchmark MOE.  Risk estimates were below the benchmark MOE even when respirators were considered.  For workers, EPA considered respirators with assigned protection factors of 10, 25, and 50.  EPA assumes workers in small commercial facilities, such as dry cleaners, would not wear respirators.  EPA also assumes occupational nonusers would not wear respirators.  
                  This slide summarizes the noncancer risk to consumers and bystanders following acute exposures.  The left hand side lists conditions of use where the risk estimates for noncancer PODs were above the benchmark MOE, indicating no risk.  For the insulation scenario, risk estimates were above the benchmark by one order of magnitude.  For the low intensity use scenarios presented on the left side, risk estimates were one to five times the benchmark.  
                  The righthand side lists the conditions of use where the risk estimates for noncancer PODs were below the benchmark MOE.  For consumer use scenarios, risk estimates were below the benchmark MOE by one to two orders of magnitude for all conditions of use evaluated, except following inhalation exposure for certain low intensity use conditions as described on this slide.  This was the situation for both inhalation and dermal consumer exposure scenarios evaluated.  The number of users or bystanders at risk is not known but would be any user of a product or bystander located in a residence where a consumer has used the products containing 1-BP.  
                  Now, moving on to cancer risk summary.  The benchmark inhalation cancer risk estimate of one times ten to the minus four for workers and ONUs was exceeded for all conditions of use based on both monitoring data and modeling approaches, except for central tendency estimates for several conditions of use, without considering the use of a respirator.  When respirators were assumed to be used by workers, the risk estimates did not exceed the benchmark for several conditions of use.  EPA does not anticipate that occupational nonusers would be wearing respirators.  The benchmark dermal cancer risk estimate of one times ten to the minus four for workers and occupational nonusers was exceeded for all conditions of use, even when gloves were not used of PF equaling one.  As the PF increased between 5 and 20, the benchmark cancer risk estimate of one times ten to the minus four was not exceeded.  
                  Strengths of the draft risk evaluation, since the 2016 draft risk assessment, the current draft risk evaluation has been refined to include the application of the TSCA systematic review process: incorporation of new information since 2016 that provided additional weight of evidence information to assess the mode of action for cancer, inclusion of the dermal exposure route for both occupational and consumer exposure scenarios, assessment of additional conditions of use, the application of nested modeling for developmental toxicity, dose response modeling of two developmental effects, litter size and post-implantation loss, resulted in similar BMCLs, the application of three modeling approaches for cancer, multistage modeling and Frequentist and Bayesian model averaging.  Dose response modeling of the tumor incidence by these three approaches resulted in similar BMCLs.  For occupational exposure, specific input parameters for the two zone models were refined to address peer review comments and to reflect additional data that became available.  For occupational exposure where monitoring data were available, risk estimates based on monitoring and modeling data were similar in most cases.  
                  This slide depicts some of the assumptions and key sources of uncertainties.  The human populations considered include individuals of both sexes 16 years of age and older, including pregnant females, for occupational and consumer settings.  Although exposures to younger nonusers may be possible, the MOEs calculated for women of reproductive age are expected to be protective of this sensitive subpopulation.  Currently, there are insufficient data regarding specific genetic and/or life stage differences that could impact 1-BP metabolism and toxicity for further refinement of the risk assessment.  
                  Cross species and route to route extrapolations are precluded by the lack of data to inform a model of a species other than rat and a route other than inhalation.  EPA chose to derive dermal HEDs for dermal exposures be extrapolating from the inhalation PODs.  Inhalation studies were performed by whole body exposure rather than nose only exposure, which may have led to additional dosing by the oral and dermal routes of exposure due to deposition on fur and the grooming behavior of rodents, resulting in uncertainty of actual dose received.  EPA is unable to conclude with certainty that comparable toxic endpoints would be associated with the dermal route of exposure considering the expected quantitative ADME differences and the absence of an adequate PBPK model.  
                  As for exposure consideration in the risk characterization uncertainties, for the occupational exposure assessment, EPA has a high or medium confidence in the exposure estimates based on inhalation monitoring data.  The expected mitigation of any risk by PPE assumes that workers are properly trained and that PPE is properly fit and worn consistently for the duration of a work activity.  EPA believes that the range of consumer use patterns considered in this evaluation, the high, moderate, and low intensity use, address some uncertainties associated with the data used to represent consumer behaviors and residential environments, as well as potential variability across such behaviors.  
                  Regarding PESS, amended TSCA requires EPA to evaluate risk to potentially exposed or susceptible subpopulations identified as relevant to the risk evaluation by the Administrator under the conditions of use.  In considering subpopulations with greater exposure and accounting for PESS in the risk characterization, EPA identified workers and occupational nonusers, consumers, and bystanders.  Groups with great susceptibility than the general population to 1-BP toxicity include developmental life stages and pregnant women and their developing fetus.  EPA accounted for all of these factors in its human health risk characterization through stratification of risk ranges, use of age and subpopulation specific exposure factors, consideration of life stage specific exposure routes, and the use of a ten X intraspecies uncertainty factor.  
                  This slide presents the risk characterization considerations per the procedures for chemical risk evaluation under the amended Toxic Substance Control Act.  On this last slide, I would like to close the presentation by bringing forward the relevant factors the Administrator will consider in determining unreasonable risk as detailed in the procedures for chemical risk evaluation under the amended Toxic Substance Control Act.  These include the effect of the chemical substance on health and human exposure to such substance under the conditions of use, including cancer and noncancer; the effects of the chemical substance on the environment and environmental exposure under the conditions of use; the population exposed, including any susceptible populations; the severity of the hazard, the nature of the hazard, the irreversibility of the hazard; and finally, any uncertainties.  With that, I would like to conclude this presentation.  I thank you for your attention.  
QUESTIONS FROM THE PANEL
                  DR. KENNETH PORTIER:  Thank you very much.  Everybody take a breath.  That was an excellent summary.  Thank you.  We have about 15 minutes to challenge EPA with any clarifying questions, so I've got Dr. Vorhees as ready to go.
                  DR. CHARLES VORHEES:  Just a few points of clarification just so I understand some of the things that you said.  When you said that you -- you said to determine the fraction of this material in a consumer product, you said you determined it based on where the information was available.  So what do you do when the information is not available?  Do you just go with those where it is available?  Or do you make some assumption to fill in?  I just didn't understand "where available."  What did that leave in the not available cases for a product where it's not available? 
                  DR. KATHERINE ANITOLE:  I'll have to reach back to the consumer exposure. 
                  DR. KENNETH PORTIER:  Please identify yourself when you come to the board, and we can only have, I think, four of the speakers on at any one point in time when I give them the signal over here.
                  MR. KEVIN VUILLEUMIER:  My name's Kevin Vuilleumier.  I'm with risk assessment division at EPA.  And in response to the question, it's more of a general categorization because certain SDS sheets only had, for example, one weight fraction.  So we didn't have a range in that situation.  So on those situations, we would use just the one across all three iterations that represented the high intensity use, low intensity use, and moderate intensity use.  Others had, say, a range of 50 percent to 100 percent of 1-BP in the product.  And in that situation, we recommend to take the low end for the low intensity use.  We'd use the high end, if it was 100 percent 1-BP.  And then we'd just take pretty much the mean for the median -- for the moderate intensity use.  So that's where we got those weight fractions from. 
                  DR. CHARLES VORHEES:  Okay.  In the part on the housing, you show a house that you use as your model, which has the living space, an attic, and a crawl space.  Housing varies tremendously.  There are houses with full basements, and there are houses that are built on slabs.  So you say that a presumption of exposure is that it will occur not in the crawl space where the insulation is containing 1-BP.  That's not where people are.  Well, I agree.  In crawl spaces, people don't generally dwell there or in the attic.  But what about with houses with basements?  People spend a lot of time in basements in a lot of housing.  So what do you do -- how do you decide that that's the model house? 
                  MR. KEVIN VUILLEUMIER:  Good question, and there is a lot of variability.  We had to put together a general house layout.  I agree.  Some basements are finished.  We have a finished basement where we have a TV and VCR and all down there.  But we initially set it up to represent the insulation installed in the attic.  It's installed in the crawlspace.  Although, it could be installed in a basement, and that would be an additional situation.  We can look into that deeper in how much of an effect it would have for those living in that area.  If it's installed in the basement and it's a finished basement, then clearly the exposure, which we do have information on -- we just did not present it in the risk evaluation.  We can present that type of information in the final risk evaluation. 
                  DR. CHARLES VORHEES:  So you talked about the fact that in these insulation products, it's the off-gassing that determines exposure, right?  Am I correct about that? 
                  MR. KEVIN VUILLEUMIER:  Correct. 
                  DR. CHARLES VORHEES:  So how do you take climate into account?  I don't think the off-gassing in a house in Minnesota versus Arizona versus Florida would be similar.  How do you account for those kinds of differences, or are they not significant?  Or do you know that they're not significant? 
                  MR. KEVIN VUILLEUMIER:  The model that we used for the insulation scenario, IECCU, absolutely considered seasonal variations.  I'll have to review in more detail how exactly we approached the average temperatures, but we did model it over a course of a year.  We can get more information on the details of what averaging, like temperatures, we used and things like that. 
                  DR. KENNETH PORTIER:  Thank you.  So we have a lot of flags up, and I didn't see the total order in which they come up.  So I'm going to kind of go left, right, left, right.  But I know Dr. Johnson was the next one up. 
                  DR. MARK JOHNSON:  Hi.  I apologize in advance if I couldn't find this in the text, but it looks like for noncancer you use the post-implantation loss or reduction in litter size as your endpoint, for acute and for chronic.  Is that correct? 
                  DR. KATHERINE ANITOLE:  Yes. 
                  DR. MARK JOHNSON:  Okay.  I'm trying to match up -- where I'm having the trouble is trying to match up the BMR in the supplementary information and the value you actually used for your POD or your HEC.  So which data from which model did you use, post-implantation loss or reduction in litter size? 
                  DR. KATHERINE ANITOLE:  We used both of those.  Both of those were used for acute and chronic. 
                  DR. MARK JOHNSON:  So did you average the two up or use the lowest one? 
                  DR. KATHERINE ANITOLE:  We calculated the risk separately for each one. 
                  DR. MARK JOHNSON:  Okay.  So you calculated risk for each, for post-implantation loss and reduction of litter size? 
                  DR. KATHERINE ANITOLE:  Mm-hmm. 
                  DR. MARK JOHNSON:  Okay.  Thank you. 
                  DR. KENNETH PORTIER:  Dr. Cobb? 
                  DR. GEORGE COBB:  I've got several questions, but I'm going to need to pass and come back to give other people a chance.  So the first question I think I know the answer to.  For the general population exposures, that was simply to consumer products and not to others? 
                  DR. KATHERINE ANITOLE:  We did not do general population exposures. 
                  DR. GEORGE COBB:  Okay.  Did you do consumer exposures? 
                  DR. KATHERINE ANITOLE:  Mm-hmm. 
                  DR. GEORGE COBB:  Okay.  That answers that question.  And then one additional one about the environmental assessment, it appeared that there were a couple of studies in the report, not in your presentation necessarily, in the report that you said that there were endpoints, but you couldn't get the full study.  So it was unclear to me whether those were used or not in the assessment? 
                  DR. KATHERINE ANITOLE:  They were used in the assessment, and they were consistent with the Geiger study.  And we provided summaries for those.  Those were the ones that we could not obtain the full study reports on. 
                  DR. GEORGE COBB:  Okay.  I was unclear as to whether or not those -- and they were consistent with the Geiger study?  Okay.  Those are my two main questions.  I'll pass. 
                  DR. KENNETH PORTIER:  I added you back to the end of the list.  Okay.  Dr. Bennett? 
                  DR. STEVEN BENNETT:  I jumped in line.  I appreciate it.  A question with respect to the dermal models and the dermal exposure assessment, within the draft risk assessment I know there's scenarios where there's quite a bit of potential for occlusion of the 1-BP under the surfaces or under the materials.  And I was wondering how that was incorporated into the model itself, either directly or whether uncertainty factors in the different scenarios of the different types of gloves, or et cetera, that could protect against that? 
                  DR. NHAN NGUYEN:  Are you asking about the occupational exposure? 
                  DR. STEVEN BENNETT:  I think it was primarily with the occupational because that was where the dermal was identified. 
                  DR. NHAN NGUYEN:  Yes, we did consider the potential occlusion, but then there were a lot of uncertainties about duration exposure and frequency and so on.  So we didn't include it. 
                  DR. KENNETH PORTIER:  Dr. Bennett, is that your total question? 
                  DR. STEVEN BENNETT:  That was my only question.  Sorry. 
                  DR. KENNETH PORTIER:  Okay.  Dr. Davies? 
                  DR. HOLLY DAVIES:  I have a couple of questions that I think will be quicker.  I was wondering where you got your reproductive age of 16 to 54? 
                  DR. STAN BARONE, JR.:  So with regard to -- you're talking about occupational exposures and the range, so again, age 16 is a work permit age, again, according to OSHA.  Those are youth.  Those are not adults, and so we included 16 and up in our worker exposures. 
                  DR. HOLLY DAVIES:  And on page 145 when you talk about using some acute exposures -- well, using acute exposures for developmental because of the critical window of fetal development, it only applies to some of the activities, which I was just wondering about that.  You don't list all of them. 
                  DR. KATHERINE ANITOLE:  For consumer or --? 
                  DR. HOLLY DAVIES:  This is occupational.  Well, occupational or consumer use of spray adhesive, dry cleaning, or degreasing products.  You can answer these later.  You don't have to -- you can come back. 
                  DR. STAN BARONE, JR.:  We'll look at that more closely, but, in general, as Dr. Anitole pointed out that we looked at acute risk estimates and chronic risk estimates for workers and the occupational nonusers.  We only looked at acute risk estimates for the consumers.  So back to the question that was asked earlier, we did use the most sensitive and robust endpoint, which were the developmental points of departure for our MOEs for both acute and chronic for workers and nonusers. 
                  DR. HOLLY DAVIES:  Yes, and I like the point about the critical window of development, which is why it caught my eye. 
                  DR. STAN BARONE, JR.:  Again, the Agency guidance -- developmental risk assessment guidance and literature supports using an acute effect or a short-term window for a chronic exposure scenario because any time during the lifetime you could have that kind of exposure and lead to a long-term, lasting consequence. 
                  DR. HOLLY DAVIES:  And then my last question, I was wondering about the significance of estimating the number of pupal for the exposures because they don't factor into the determination of risk.  And you spend a lot of time, it seems, estimating the numbers of pupal. 
                  DR. STAN BARONE, JR.:  So again, we do obtain that information to where it's possible and do consider that in the risk determination, but we're not calculating -- we're not making any additional calculations at this time for the population size and the impact.  That's done later.  If we go to unreasonable risk and risk management, we do an impact assessment in the regulatory arena.  That's another exercise.  But we gather that information where possible. 
                  DR. HOLLY DAVIES:  So I guess that -- which was leading to because it wasn't affecting the risk determination now, if we don't comment on it now, does that imply that we think it's a good estimate for the next possible step? 
                  DR. STAN BARONE, JR.:  So again, this is about the data, and you are allowed.  And of course we would like additional information if you know of additional information about the size of the worker population or the number of people exposed.  That's important information for us to consider.  That's data.  That's important information for us to consider both in the revision of the risk determination, as well as future activities if future activities are called upon. 
                  DR. KENNETH PORTIER:  Thank you.  And that was Dr. Barone speaking for EPA.  Just a reminder to say your name for the audio record.  Dr. Schlenk? 
                  DR. DANIEL SCHLENK:  Yeah.  A couple questions, on the TRI data that was used to estimate the waterborne exposures, it said 100 percent of that was stormwater derived.  So was the facility that was used a wastewater treatment plant, or was it an industrial treatment plant?  And how did stormwater affect that, I guess, is the question? 
                  MS. EVA WONG:  In modeling out the releases from the facility, there was one reporting facility.  It is, I believe, a manufacturing facility.  And in the TRI information, they report the release to water as 100 percent to stormwater, so in our modeling, we did not apply any wastewater treatment removal to this.  It was zero percent removal as a protective assumption. 
                  DR. DANIEL SCHLENK:  Okay.  I'm just curious how stormwater would actually cause that.  The other question I had I think I know what the answer is, again.  Because this is in insulation, I'm assuming that landfill was not considered in part of the disposal in the conceptual models.  Is that correct? 
                  MS. EVA WONG:  We did not model that scenario. 
                  DR. DANIEL SCHLENK:  Thanks. 
                  DR. KENNETH PORTIER:  You want a follow-up?  Dr. Twiss? 
                  DR. MICHAEL TWISS:  With reference to water solubility, following up from the last question, throughout the document I encountered ten references to solubility of 1-BP.  And most of them consider it to be high; although, there's a reference to being slight soluble, low solubility, and moderate water solubility in three other reference.  And this morning, Dr. Anitole considered it slightly soluble in water.  So is it slightly soluble, moderately, or highly soluble? 
                  DR. STAN BARONE, JR.:  Again, thank you for that observation.  We were working very diligently, given the comments from our last peer review, about making those qualitative calls consistent with the PChem properties section.  So the PChem property section the call is moderately soluble.  Again, thank you for making those notations.  We have a very large team, and, again, adjectives lead us into different calls.  And that's something we'll take into account as we revise the assessment. 
                  DR. KENNETH PORTIER:  I can only think of a 100 page document with 20 authors, they probably don't all get their adjectives right.  Dr. Gilbert? 
                  DR. KATHLEEN GILBERT:  Going back to the insulation question, if I think I understood your slide right, you said that as far as consumer use the risk to bystanders was not unreasonable.  The risk to users was.  So how are you dividing users and bystanders when it comes to the insulation? 
                  MR. KEVIN VUILLEUMIER:  We actually only evaluated the insulation in the context of a bystander because, as I discussed earlier, it was an off-gassing.  So user, we didn't really -- a user would just be during installation.  You wouldn't really have an actual exposure for inhalation for the users.  You're just putting it in.  It's the off-gassing over time, and in that situation, it's really a bystander, unlike a consumer product like a spray where the user's right there using it.
                  DR. KATHLEEN GILBERT:  Okay.  Thanks.  I have a second question. 
                  DR. STAN BARONE, JR.:  So if I may clarify that, that's captured under the commercial use.  So the installation of the insulation, the 1-BP exposure, is captured under the commercial use. 
                  DR. KATHLEEN GILBERT:  Thanks.  So we had heard previously that some of the risk assessment for the general population, in terms of air quality, was going to be handled by the Clean Air Act.  Do we have any more information on that or updating? 
                  MR. MARK HARTMAN:  I'm not going to read it, but I think that the description that you see in the introduction on page 27, paragraphs 3 and 4, is our position on why we chose in this particular case to not look at general population exposures from air releases. 
                  DR. KATHLEEN GILBERT:  But there was a petition to have it listed as a hazardous air pollutant, and it's not clear from what we've heard here where that hazardous air pollution would come from. 
                  MR. MARK HARTMAN:  Oh, the source? 
                  DR. KATHLEEN GILBERT:  Yes, the source. 
                  MR. MARK HARTMAN:  So two possible sources would be through the stack releases from manufacturing and also potential releases from, for instance, a dry cleaner to offsite. 
                  DR. STAN BARONE, JR.:  So manufacturing and processing, those are two sources, and then the dry cleaning commercial use is another source category, just to be more explicit. 
                  DR. KENNETH PORTIER:  I think Dr. Blando. 
                  DR. JAMES BLANDO:  Just two quick questions.  On one of the slides, Dr. Anitole, you presented occupational scenarios above the benchmark MOE and below.  And I just might not have caught it on the slide.  Where did dry cleaning fall on that slide?  Where does dry cleaning -- the occupational use of dry cleaning -- is it above or below the benchmark MOE?  I just didn't see it on the slide. 
                  DR. STAN BARONE, JR.:  It's below for both users and nonusers. 
                  DR. JAMES BLANDO:  Okay.  And then I just had one other clarifying question.  In your presentation, you mentioned that personal protective equipment.  It sounded like, from what I've read and from your presentation, that for some scenarios you assumed workers were using PPE properly, and then in some scenarios, you were assuming they didn't use PPE.  I just want to clarity that I understood that correctly and would just like to hear a quick snippet about the logic of how you decided that workers would be using the PPE correctly. 
                  DR. NHAN NGUYEN:  Yeah.  I think in the original evaluation, the way we assess exposure considering PPE, if PPE is worn properly with a PF of 5, 10, 50, then these are the exposures. 
                  DR. JAMES BLANDO:  So I guess my question is how did you decide which industries use PPE and which ones don't? 
                  DR. NHAN NGUYEN:  Well, based on information from published literature and/or judgement on some uses.  And generally, you know that they won't use PPEs.
                  DR. JAMES BLANDO:  Right.  I was just going to say my experience has been that PPE is not often used properly or well.
                  DR. STAN BARONE, JR.:  So as we have spoken about this previously and we have received numerous public comments, again, our general approach is to look at the technical safety datasheets, look at the publications, look at industrial hygiene practices in different sectors and what we know about those sectors.  Industrial sectors tend to have larger industrial uses, tend to have industrial hygiene plans; whereas, smaller commercial sectors tend not to have any industrial hygiene plans.  So those are the two extremes.  
                  We're looking for more information about PPE and PPE uses, as well as the other stages in the hierarchy of controls.  And any information that could be provided would be helpful to our evaluation.  But we are looking for more information in that regard.  We do make it clear in the assessment that there are certain conditions of use where we do not expect any PPE usage.  Dry cleaning is one of those, and there are other commercial activities that we specifically call out that PPE is not used in any routine or correct fashion. 
                  DR. KENNETH PORTIER:  Thank you.  Dr. Kissel?  And we're a little bit into our break here, but we're going to go ahead and get these comments out.  And then we'll take our 15 minute break. 
                  DR. JOHN KISSEL:  So my question has to do with the consumer dermal exposure.  I couldn't follow it.  There's no worked example.  The spreadsheets that are in the supplemental information aren't really spreadsheets.  They're just tables with de facto numbers in them without any equations which lead to those numbers.  So did I miss something, or can you give me an example of how you actually did the consumer dermal exposure assessment? 
                  MR. KEVIN VUILLEUMIER:  Thank you for the question.  For dermal exposure, we used a CEM model, and it has inputs for select -- we did an inhalation route.  And we also have a dermal route.  We selected the particular models that we're going to use, and then we input the various parameters that we varied.  And then we have default parameters.  In the case of the dermal, we had a permeability coefficient, so that was an input that we added to it.  
                  And the tables that we provide in the supplemental documents, as well as the appendix, are designed to give you the list of inputs that were utilized.  And then you'd go into the CEM model and input those parameters there.  And that's how it'd run it.  So we tried to lay out all the inputs because they vary by scenario and by the routes of absorption.  That's exactly what you're going to put in there.  So it's kind of hard to describe a full example of the whole process as it's incorporated in the CEM model, which is publicly available. 
                  DR. JOHN KISSEL:  Well, I tried to do all of that.  So the text says that you're using PDER 1B, which they're four dermal possible models in CEM.  Actually, it turns out there's many more.  I looked in the CEM user's manual.  In the user manual itself, PDER 1B doesn't even exist.  It's in the appendices.  And the example given there is for a showering scenario where somebody is exposed to chloroform in a shower via dermal route in addition to inhalation.  And the application here I don't think is from aqueous solution, but maybe it is.  
                  And that's where my confusion lies because there's no worked example, and the permeability coefficient that you're using seems to be from -- it's the Wilcutt (phonetic) modification of the Robinson equation that NIOSH uses.  And that is from aqueous solution.  So I'd like to know did you multiple that by a concentration in aqueous solution?  It seems like all of the exposures here are to something that's 80 percent up of 1-bromopropane?  So if you're using an aqueous permeability coefficient, then that's probably wrong.  So I'd like to know what you did.  And if you can't give me a numerical example, I'm not going to be able to figure it out. 
                  MR. KEVIN VUILLEUMIER:  We can certainly look back and try to provide some additional information for you, if that's okay, because we'll have to go back and look in the models. 
                  DR. JOHN KISSEL:  Yeah.  And I only need one example.  I don't need all 27 or however many there are.  Just show me one that allows me to trace what it is you did. 
                  MR. KEVIN VUILLEUMIER:  Okay.  We can certainly try to pull that together for you. 
                  DR. KENNETH PORTIER:  Dr. Eastmond? 
                  DR. DAVID EASTMOND:  Thanks.  I'm trying to understand kind of the literature view on this.  So you have the systematic review, and it describes how there's the data search, et cetera.  That seems to have happened for new information, new studies or fairly recent studies.  There also seems to have been sort of a grandfathering in of other studies that took place -- and these would be like ATSDR or ECHA.  And maybe there's a previous document.  Is that correct?  Now, were those reviewed as well -- went through sort of the data quality and that type of aspect of the review, or was that information just taken as is and plugged in?  The reason is I'm finding some errors in the thing, and I'm trying to figure out where they came from -- the errors in the document. 
                  DR. KATHERINE ANITOLE:  So are you referring to on the human health hazard side? 
                  DR. DAVID EASTMOND:  Yeah.  It's the human health hazard. 
                  DR. KATHERINE ANITOLE:  So for that, since we had just gone through the peer review in 2016, the studies that were in the 2016 risk assessment that had gone through the systematic review at that time and then peer reviewed by the CSAC were also put through the current systematic review process.  Those were the ones that were carried forward for dose response analysis on the human health side.  So they, in a sense, went through a second systematic review process, and, as a result of that, they all ranked high quality that were eventually then carrying forward as we did in 2016.  
                  Any new studies that came since 2016, we did a comprehensive literature search on anything since the 2016 peer review.  And then those were likewise run through the current systematic review process.  And those happened to be the mode of action studies for the big blue.  Those were to two new ones that came through since 2016. 
                  DR. DAVID EASTMOND:  Yeah.  I mean, there's a whole series that's sort of on the mode of action genetic toxicology test.  I know there are many studies, and it's actually hard to keep track of them.  And some of them are published and some aren't, and that's what I'm trying to figure out.  So those would have been reviewed in this 2016 peer review and then largely carried forth based on that?  So for example, the statement that 1-Bromopropane binds to DNA calf thymus DNA in vitro, that actually goes back to an article that doesn't show that.  
                  But it's mentioned in its discussion, and you follow the reference in the discussion that goes to an abstract from a SOT meeting, which is not really available.  So that's kind of a key piece of information in this argument that I couldn't chase down.  Now, there are two new studies that have been published more recently that actually show that information.  So the information itself is not incorrect, but the references and the basis for it is not correct. 
                  DR. KATHERINE ANITOLE:  So we can certainly provide further clarification on that and go back to check to see if there's an inconsistency.  I should mention that, if there's anything that came out -- I think the stopping point was around 2000 -- well, since the problem formulation document, which was a year ago.  So anything that's come out since then obviously didn't get picked up in our literature or go through systematic review.  However, we would be more than happy -- if there are any additional studies, we would be more than happy to evaluate those and include them in the draft risk evaluation. 
                  DR. KENNETH PORTIER:  I think following up on that the grandfathered documents bypassed the screening, but they go through evaluation, right?  They just don't get screened.  They just get put into the pool where you get into the evaluation and the data extraction component.  So it gets a critical review but at a different level. 
                  DR. STAN BARONE, JR.:  There's one clarification here I want to make sure.  As Dr. Anitole pointed out, it's the studies that were critical and key supporting studies that went through that screening from the prior assessment.  Not every study that we had identified, hundreds of studies that we had identified in the previous assessment, were not screened through the tool.  So let's be clear.  We didn't have the resources or the time to go through the hundreds of studies that were in the previous assessment.  We did take all the previous key and supporting studies for the critical effects through the screening. 
                  DR. KATHERINE ANITOLE:  And if I could just add one more thing to Dr. Barone's statement.  We were in sort of a unique position with 1-BP because we had just gone through the rigorous CSAC peer review, so we felt fairly confident in using that approach. 
                  DR. KENNETH PORTIER:  Dr. Johnson, then Dr. Doucette, and then Dr. Cobb and then a break.  Dr. Johnson? 
                  DR. MARK JOHNSON:  All right.  Two totally disjointed but hopefully quick questions.  When you assessed the factors for the aquatic receptors, you used five for acute and ten for chronic.  That was just to apply for duration, study duration.  Is that correct?  It's not designed to apply to interspecific variation?  You didn't have enough data to do specie sensitivity distributions.  Is that correct? 
                  DR. KATHERINE ANITOLE:  I'm sorry.  I missed the first part of your question.  Was that for environmental tox? 
                  DR. MARK JOHNSON:  Yeah.  For aquatic receptors, the assessment factors you used are five and ten?  Okay. 
                  MR. GARRETT JEWETT:  You might need to repeat your question.  I didn't hear it correctly. 
                  DR. MARK JOHNSON:  Sure.  The assessment factors you used for aquatic receptors were five and ten, five for acute, ten for chronic.  Was that just to make an adjustment for study length duration to apply to a chronic situation, or was that to consider interspecific variation? 
                  MR. GARRETT JEWETT:  There were two factors used in this case.  The first one was ten, which was used to account for the duration adjustment between acute and chronic.  And then the other one varied based on whether or not we were calculating the acute or chronic COC.  In this case, the acute COCs were calculated using an assessment factor of five, and the chronic COCs were calculated using assessment factor of ten.  That was to account for interspecies variation.  So in the case of estimating the chronic COCs, there were actually two assessment factors of ten applied, one to account for duration adjustment to make the estimation from acute to chronic.  And the other was to provide interspecies -- to account for interspecies variation. 
                  DR. MARK JOHNSON:  Great.  And all the data that you applied them to were LC50s.  Did you have an EC50s or non-lethal criteria?  Like NOAECs, LOAECs?  
                  MR. GARRETT JEWETT:  In this case, so the estimated chronic value, it's a CHU which is based on a geometric mean between a NOAEC and LOAEC.  But in this case, because it's estimated, you can't say exactly whether or not it was a NOAEC or LOAEC.  But typically, it's LC50s.  Or if it's based on algae, it would be an EC50. 
                  DR. MARK JOHNSON:  Okay.  The other question we had -- and a couple of us were talking about this -- on Table 426, we noticed that typically for higher exposures, high intensity use, particularly from coin cleaners, automobile AC flush, the MOE tends to be lower than for the low intensity use.  In these two circumstances, it's the opposite.  In fact, for the overall maximum, the MOE is much higher than it is for the minimum.  So I'm wondering is that just a calculation or is it something that we're missing? 
                  DR. STAN BARONE, JR.:  I'd have to look at that.  We'll get together at break and get back with you. 
                  DR. MARK JOHNSON:  Okay.  Thank you. 
                  DR. KENNETH PORTIER:  Dr. Doucette? 
                  DR. WILLIAM DOUCETTE:  Yeah.  Just a quick question on the use of dry cleaning.  So if it's used as a dry cleaner and yet it was not assumed that there would be leaks from dry cleaning facilities in terms of a potential route into the environment.  Am I understanding that correctly? 
                  DR. STAN BARONE, JR.:  So you're talking about the general population exposure, not the actual worker exposure? 
                  DR. WILLIAM DOUCETTE:  Yes.  So in the environmental, another pathway for an environmental release, other than the stormwater that was brought out from the one facility. 
                  DR. STAN BARONE, JR.:  So again, as part of that --
                  DR. WILLIAM DOUCETTE:  You don't have to consider that, is that right? 
                  DR. STAN BARONE, JR.:  Right.  Because part of the HAP listing, that will be covered in those particular pursuits.  In our original scoping, it was in.  And since then, it has changed because the Agency has decided to move forward with the HAP listing. 
                  DR. WILLIAM DOUCETTE:  So that is related to the HAP listing? 
                  DR. STAN BARONE, JR.:  It is very much related to the HAP listing where the emissions will be controlled under the air --
                  DR. WILLIAM DOUCETTE:  And they would take care of vapor intrusion issues at that point then, there, assuming it leaks? 
                  DR. STAN BARONE, JR.:  You mean vapor intrusion into co-located facilities or adjacent facilities? 
                  DR. WILLIAM DOUCETTE:  Right. 
                  DR. STAN BARONE, JR.:  Yes. 
                  DR. WILLIAM DOUCETTE:  Okay.  Thanks for the clarification. 
                  DR. KENNETH PORTIER:  Dr. Pessah on the phone raised his hand.  Dr. Pessah? 
                  DR. ISAAC PESSAH:  Yes.  I was wondering if -- and I didn't hear this during the presentation.  Some of the data for occupational exposure indicates neurological outcomes that tend to be sex specific, or at least females seem to be more affected than males.  Was that in any of your modeling? 
                  DR. KENNETH PORTIER:  They're looking around. 
                  DR. STAN BARONE, JR.:  The sex specific -- which sex specific effects are you talking about, Dr. Pessah? 
                  DR. ISAAC PESSAH:  In particular in Ishihara 2011 they showed that females had more neurological symptoms. 
                  DR. STAN BARONE, JR.:  You're talking Ishihara.  I believe that was the rodent hanging -- no, not the hanging -- grip strength? 
                  DR. ISAAC PESSAH:  No, this is an acute and occupational exposure. 
                  DR. STAN BARONE, JR.:  This is the occupational exposure.  Okay.  I'm trying to remember exactly which study was which, and I'm flipping around at the same time.  The outcomes I think you're talking about, again, are primarily in the domain of peripheral neuropathies.  And again, there was a larger population of females versus males in that worker population, if I remember correctly.  Whether that's a reflection of the number of cases or the actual sensitivity -- gender specific sensitivity, I don't think we have enough information to really go after that.  We can look further at that based upon your comments. 
                  DR. ISAAC PESSAH:  Okay.  I was just wondering if you had tried to accommodate a sex difference in any of your modeling. 
                  DR. STAN BARONE, JR.:  So again, the Chinese factory worker studies are not principle studies that we are basing our point of departure on in this evaluation for a number of data quality issues.  I'll just remind you of that because of data reporting.  It was a little hard to reconstruct some of the exposures.  And then there were some methodological concerns also raised about the studies in the way the measurements were taken, an apparent lack of blinding in the testing by the clinicians who were doing the testing. 
                  DR. ISAAC PESSAH:  Thank you. 
                  DR. KENNETH PORTIER:  I have a follow-up question real quickly.  You mentioned at this meeting and in the previous that certain references that are not translated into English get excluded.  For example, on the eco tox you said the studies were in French and Japanese with no U.S.A. sponsor.  Therefore, EPA decided not to make a further attempt to obtain the full study reports.  So what's the policy on that?  Is that just kind of a blanket, or is that a case by case, situation by situation? 
                  DR. STAN BARONE, JR.:  So again, I wouldn't say necessarily policy.  But with regard to our systematic review framework, that is called out specifically in the systematic review framework that we do focus on English reported studies first.  We have received and did evaluate the English robust summaries of those studies.  We did not have the full study reports in English and were not able to identify who the owner of the data was, the source of the data, to get further information for a translation.  And that's, again, quite a challenge with the ECHA database.  We are working with our partners across the pond on some of those issues.  But it is a challenge at times. 
                  DR. KENNETH PORTIER:  Dr. Cobb, last questions. 
                  DR. GEORGE COBB:  So I actually have several.  I may just pose -- can I give them in writing to the EPA and have them answer them later?  Or do we want to go through them all right here? 
                  DR. KENNETH PORTIER:  If you really need to take the break, after the public comments, we'll open for a few more clarifying questions at that point.  If that works, we can do that. 
                  DR. GEORGE COBB:  So my first question is about the algae toxicity, and it was about an order of two and a half lower than the fish toxicity.  And there was a rationale given for not including that, but that would only hold if you're only trying to protect fish, I think, or maybe fish and invertebrates.  So can you clarify why that was excluded a little more? 
                  DR. STAN BARONE, JR.:  We'll follow up after the break. 
                  DR. GEORGE COBB:  And then I have kind of a general question about the way that the data were evaluated for inclusion or exclusion.  I've got to find the right one here.  Hold on.  So when you were doing the search terms, as I was reading through, there's places where it identifies chemical uses and then exposure types of terms.  And did the strategy require both the chemical and the use terms in the strategy to be true before then allowing the exposure and engineering and fate terms to then kick in?  Or conversely, was it chemical identifier or use terms before the rest kicked in? 
                  DR. STAN BARONE, JR.:  So I think you're talking about systematic review and the screening procedure for systematic review.  There are two bins that are important for our systematic review for exposure engineering and also fate, but exposure engineering.  The great sources of literature are much more important and more helpful to us.  We also look at the published literature.  So again, the chemical identifier, cast number, are the chemical specific information, but what we need for exposure assessment and engineering assessment is the use information, the activity information, the other.  So it's and, not or.  
                  And we do -- and we've talked about this previously with systematic review with the published literature.  When you don't have that PICO, you don't have those and statements, you end up capturing -- pulling down a lot more literature that's cast number or chemical specific or synonym related.  And that ends up being a whole lot of extra work to go through in the screening process.  So we're refining our approaches as we speak on these very issues, again, learning from AI techniques and others.  So hopefully we'll have more to talk about on that subject later.  Thank you. 
                  DR. KENNETH PORTIER:  Okay.  I have about -- what do I have here?  11:30.  Let's reconvene at 11:45.  We have about 45 minutes of public comments, so we're going to push the lunch break back to 12:30 or so.  So we'll take the public comments before we take the break.  Let's reconvene in roughly 15 minutes, 11:45.  Thank you. 
                  
                                             (BREAK)
                                                
PUBLIC COMMENTS
                  
                  DR. KENNETH PORTIER:  Let's reconvene.  Thank you.  This next section we're going to entertain comments from the public, and Ms. Gibson's going to moderate that section for us.  Each of the speakers have a designated about of time -- limited amount of time to talk to the panel.  
                  MS. TAMUE GIBSON:  Thank you, Dr. Portier.  So we will begin our public comments session at this moment.  You are limited to five minutes.  There are a number of public commenters, so I'll ask you -- there's an order in your agenda.  So if you could move after the speaker has concluded their talk.  Please move into the seat directly in front of me where you have a placard that indicates public commenter.  So if I could have Ms. Lindsey -- I'm sorry.  Ms. Stephanie Schwarz, please, from -- a legal fellow from the Environmental Defense Fund. 
                  MS. STEPHANIE SCHWARZ:  Hello.  My name is Stephanie Schwarz, and I am a legal fellow at Environmental Defense Fund.  I will make brief comments today with more detail in the written comments we've already provided to the SACC.  Recently, EPA has publicly stated that a number of the topics discussed are in the realm of policy and are therefore not relevant to the SACC's charge.  These include EPA's decision to exclude all exposures from releases to land, air, and water based on the assumption that other statutes adequately address the exposures; EPA's decision to assume that appropriate personal protective equipment is always used based on the authority of the Occupational Safety and Health Administration; and EPA's decision to use a benchmark cancer risk level of one times ten to the negative four to define unreasonable risk to workers.  
                  EDF strongly disagrees that these issues are beyond the scope of the SACC.  In fact, they fall squarely within the SACC's charge.  All three decisions have major direct scientific consequences as they clear to underestimations of chemical's risk to the environment, the general population, workers, and vulnerable subpopulations.   Charge Question 7, among others, expressly directs the SACC to address the uncertainties and assumptions that EPA uses in the draft risk evaluation.  All three of EPA's decisions I just described represent assumptions that EPA has not verified or adequately explained and then introduce major uncertainty into its risk evaluation that EPA has not analyzed.  It is vital that the SACC consider and address in its report the scientific consequences of these assumptions and decisions on EPA's characterization of exposure, hazard and risk.  
                  Let me address each of these issues a bit further.  First, the statutory based exclusions, EPA has asserted that exposures to the general population are adequately managed without any analysis whatsoever of the standards under the other statutes, including standards that are not strictly health-based, unlike TSCA's standard.  EPA devoted less than three pages to justify its decision to eliminate entire pathways and provided no data or analysis of the exposures and risks that remain and our contribution to total exposure and risk.  EPA has failed to provide any scientific rationale for this assumption, and the SACC has been charged with commenting precisely on the adequacy of the support EPA has provided for just such assumptions.  
                  Second, EPA has also assumed that OSHA ensures that suitable PPE is always used in order to find no unreasonable risk to workers, even though EPA has state elsewhere in the draft risk evaluation that few literature sources indicate the use of respirators in 1-BP conditions of use.  And EPA does not know the actual frequency, type, and effectiveness of glove use in specific workplaces with 1-BP conditions of use.  Despite the lack of data regarding the use of PPE, EPA has made broad assertions about OSHA's authority.  In order for the SACC to fully evaluate these assumptions, the SACC should request that EPA provide any feedback EPA has received from OSHA and NIOSH on its assumption regarding PPE use and, more generally, any input they have provided EPA regarding the extent and sufficiency of OSHA's authorities.  
                  Third, EPA's unprecedented use of one in 10,000 as the cancer risk benchmark for workers also clearly underestimates risk and flies in the face of EPA's longstanding policy that it should reduce risk to less than one times ten to the negative six for as many exposed people as reasonably possible.  In sum, TSCA specifically states that the purpose of the SACC is to provide advice on scientific and technical aspects related to the implementation of TSCA.  And EPA's sweeping assumptions regarding environmental exposures, PPE use, and adequacy of a less protective cancer risk benchmark have direct impacts on the scientific integrity of EPA's implementation of TSCA.  The SACC needs to address the scientific consequences of each of these decisions.  At the very least, it is the SACC's responsibility to state clearly in its report that the result in serious underestimations of risk.  Thank you for your time. 
                  MS. TAMUE GIBSON:  Thank you.  I actually would like to make note that for you public commenters, if you could -- we will definitely email those comments to all of the committee members as well as placing them into the docket.  So for those who have not submitted their comments to me, please email them to me so that we can make those arrangements to have them submitted to the committee as well as into the public docket.  Okay.  Ms. McCormick.  She's program manager, Environmental Defense Fund.  Please turn your mic on, please.  Yes. 
                  MS. LINDSAY MCCORMICK:  That would help.  Okay.  Good morning.  My name is Lindsay McCormick.  I'm also with Environmental Defense Fund.  In 2016, Congress strengthened EPA's authority and mandate under TSCA to protect workers by expressly identifying them as potentially exposed or susceptible subpopulation.  Yet, EPA's draft risk evaluation grossly understates the risk to workers and overstates OSHA's protections.  I will discuss four concerns, which are most germane to Charge Questions 2 and 6.  
                  First, EPA ignores real world limitations of PPE and distorts OSHA requirements.  EPA heavily relies on the assumption that all workers will always use gloves and respirators, and they'll be universally effective in most scenarios.  For example, EPA assumes that workers are properly trained and fitted on the respirator use and that they wear respirators for the entire duration of the work activity where there's potential exposure to 1-BP.  These assumptions are wholly unwarranted.  
                  First, any OSHA requirement for employers to provide respiratory protection from 1-BP exposure will apply only extremely rarely for many reasons, including the fact that OSHA PEL does not exist for 1-BP.  EPA distorts development OSHA requirements when it invokes OSHA respiratory protection standard at 29 CFR 1910.134, which only applies for chemicals with an OSHA PEL.  Dr. Finkel's (phonetic)comments to this fact discuss this issue further.  
                  Second and more broadly, even where OSHA respiratory protection requirements do apply to a chemical, OSHA's database of inspections demonstrate significant noncompliance with those requirements.  In physical year 2018 alone, OSHA cited 2,892 violations of the respiratory protection standard identified in 1,281 separate inspections.  Violations of the respiratory standard were the fourth most common type of violation in OSHA inspections that year.  EPA overstates or distorts other OSHA requirements, as well.  For example, EPA implies that safety datasheet recommendations for PPE are mandatory, when, in fact, OSHA standard mandating SDS specifically states that there is, quote, no requirement for employers to implement their recommended controls.  
                  Second, EPA conflates the risk evaluation and risk management processes by assuming use of PPE.  TSCA intentionally divides risk evaluation and risk management into two distinct processes whereby regulatory measures are considered after EPA finds an unreasonable risk.  However, by choosing to make risk determinations based on the assumptions of universal effective use of PPE for most scenarios, EPA conflates risk evaluation and risk management and leaves EPA either not to find unreasonable risk or to underestimate the magnitude of that risk in a number of scenarios thereby denying itself the authority to impose mandatory requirements sufficient to control workplace exposures.  
                  So I want to just look at one example on the slide here.  On this table from EPA's draft on page 237, you can see that for cancer risk from dermal exposure EPA has actually found excessive cancer risk in every scenario, even using its one in 10,000 benchmark.  You can see that in the pink.  Yet when it comes to the risk determinations, EPA finds no unreasonable risk in several scenarios, including manufacturing import, which is displayed on the slide by stating that there is, quote, no unreasonable risk when PPE gloves of PF of five are used.  EPA's failure to make an unreasonable risk determination will mean it will lack any authority to require that such gloves are actually used.  
                  Third, EPA understates risk by using the one in 10,000 benchmark.  So let's look at the same EPA table through another lens.  Again, the shading shows that the EPA found excessive risk using the one in 10,000 benchmark.  Recall my colleague Stephanie's earlier comment that EPA should be using a much more protective cancer risk benchmark.  So if EPA were to use even one in 1,000 -- in 100,000, the table shows it would have found excessive risk in every scenario assuming gloves of PF up to ten.  And then using EPA's longstanding policy of aiming to reduce risk for as many exposed people as possible to one in a million, EPA would have found excessive risk in every single scenario, even if gloves of PF of 20 were used.  
                  And as a side note, I do want to note that EPA's supplemental Excel file in this table shows that included scenarios have risks as high as eight times ten to the negative two, which is almost one in ten.  Yet in the draft risk assessment, EPA does not make a single unreasonable risk determination based on cancer risks from dermal exposure.  And then my fourth and final point, which I have no time to make, is that EPA underestimates exposure by failing to consider combined exposures from workers through different routes.  So it does not combine inhalation and dermal exposures.  Thank you. 
                  MS. TAMUE GIBSON:  Thank you.  Mr. Denison. 
                  DR. RICHARD DENISON:  Thank you.  I'm Dr. Richard Denison, also with EDF.  Let me start by making one statement that reflects on what was heard this morning.  Every time you hear EPA say that there is uncertainty or assumptions or modeling or estimations that had to be made, had to be made because of a lack of data, that is a reflection of the EPA's failure to use the authority that Congress gave it under TSCA to require the information to have been developed.  So we are now three years out with major data gaps that affect EPA's assessment, and that is a result of how they have approached this and every other one of the risk evaluations that you are all reviewing.  I do want to reflect on two serious concerns about the risk evaluation.  
                  First, EPA fails to assess cancer risks from acute, short-term exposures.  EPA acknowledges that 1-BP is a mutagenic carcinogen and that linear extrapolation is warranted.  Yet the Agency has chosen not to estimate cancer risks from short term exposures.  And hence, it completely excludes that risk for consumers.  For a chemical with a mutagenic MOA, even a single direct reaction, specifically a single hit on a single target, can be sufficient to cause cancer, and, hence, acute exposures pose a cancer risk.  
                  For EPA's sole rationale for ignoring these risks is that published methodologies for extrapolating from chronic to short term exposures have uncertainty.  EPA cites a 2001 NRC report for support, but that same NRC report goes on to say that cancer is a potential effect from short-term exposures.  And it cites NRC guidance specifically saying that acute exposure guideline levels should be developed using extrapolation, linear low dose extrapolation from chronic exposures.  EDF is concerned the EPA has ignored these basic principles in deciding to exclude this risk.  
                  EPA erroneously assumes that acute exposures to 1-BP, including to consumers, pose zero cancer risk.  It departs from sound science by effectively adopting a threshold for 1-BPs dose response, in this case based upon the duration of the exposure as opposed to dose per se.  The only scientifically supportable approach here is for EPA to be assessing cancer risk from acute exposures using linear low does extrapolation.  
                  The second issue I want to raise that came up several times is EPA's lack of access to full studies for aquatic toxicity.  All but one of EPA's aquatic tox studies it cites as sourced from dossiers available through the European Chemicals Agency, or ECHA.  EPA claims that ECHA dossiers are existing chemical assessments equivalent to EPA or ATSDR.  And it lists ECHA as the author of these studies.  These statements are false and highly misleading.  In fact, ECHA dossiers are not assessments.  They are not government documents.  They are compilations of industry studies and information submitted to ECHA that have not been evaluated for quality by ECHA or any other governmental entity.  What ECHA provides are not full studies, only study summaries prepared by the industry registrants.  EPA acknowledges it lacks access to the full studies, and it did not subject them to systematic review.  And yet it proceeds to use those studies in its analysis.  
                  EPA has access to only a single full acute study on fish toxicity, and it has no chronic data whatsoever.  Instead, it extrapolates from that one study.  And then it uses the acute studies that the industry summarized to estimate chronic toxicity by applying an acute to chronic ratio of ten.  It provides no justification or citation for that value.  And even a cursory look at the literature indicates that that ratio should often be set at at least 100 in order to be protective.  
                  EPA has no tox data on sediment or terrestrial organisms and instead resorts to vague arguments to assume there is no exposure, and therefore they need not assess it.  And then EPA leaps to a final conclusion that there is no risk posed by 1-BP to the environment as a whole based on what I've just described.  The SACC needs to make clear that EPA's assessment is insufficient to make that conclusion.  Thank you. 
                  MS. TAMUE GIBSON:  Thank you, Mr. Denison.  Operator, do we have Robert Sussman on the line. 
                  MR. ROBERT SUSSMAN:  I am on the line.
                  MS. TAMUE GIBSON:  Could you please provide --? 
                  MR. ROBERT SUSSMAN:  Can you hear me? 
                  MS. TAMUE GIBSON:  Yes.  Is this Mr. Robert Sussman? 
                  MS. ROBERT SUSSMAN:  This is he.
                  MS. TAMUE GIBSON:  Could you go ahead with your comment, please?  You have five minutes. 
                  MR. ROBERT SUSSMAN:  Thank you.  Great.  I'm Bob Sussman, and I'm here today representing Safer Chemicals Healthy Families.  We are concerned about the exclusion of air emissions from the scope of the draft risk evaluation.  EPA's 2018 problem formulation document for the evaluation highlights 1-BP's high volatility and emphasizes, quote, air as a primary medium of environmental release, close quote.  
                  1-BP was added to the toxic release inventory in 2015, and during the initial round of reporting, 43 facilities reported fugitive emissions totaling 394,000 pounds.  And 26 reported stack emissions totaling 232,000 pounds.  The problem formulation notes that the TRI reports likely understate the air emissions because many admitting facilities were below the threshold for reporting.  This would include small facilities using 1-BP as a vapor degreaser, as a component of adhesives and sealants, and in dry cleaning and other cleaning operations.  For example, the draft risk evaluation estimates between 500 and 2,500 establishments using 1-BP for vapor degreasing and between 1,000 and 5,000 using the chemical as an aerosol spray degreaser.  
                  The magnitude of the health risk for 1-BP air emissions is demonstrated in an exposure and risk assessment submitted to EPA by the Halogenated Solvents Industry Alliance.  Focusing on five representative facilities, the assessment concluded that, quote, 9,000 people are estimated to have a cancer risk greater than one in one million, close quote.  If more facilities were included in the analysis, the at risk population would be far higher.  
                  Let me turn to EPA's rationale for excluding air emissions, which is the premise that the Clean Air Act adequately assesses and effectively manages risk to the general population.  As the HSIA assessment shows, general population exposure could be a significant source of risk and would be additive to be unacceptable cancer and other risks that users of consumer products already face.  Failure to account for this adequate risk to consumers in highly exposed subpopulations living near admitting facilities would result in an incomplete picture of 1-BP's public health impacts and lead to inadequate and unprotective mismanagement measures.  
                  In addition, it is unlikely that the Clean Air Act will adequately protect against unsafe air admissions.  No chemical has been added to the HAP list since the amendments of the Clean Air Act were enacted in 1990.  The petition to list 1-BP has been pending since 2010.  While to Obama EPA made a preliminary decisions to grant the petition in early 2017, the Trump EPA has taken no action since then.  There is considerable uncertainty whether there will in fact be a HAP listing.  We shouldn't be excluding an important risk pathway from the risk evaluation based on something that will not happen.  
                  Moreover, even if 1-BP is listed as a HAP, EPA would mandate technology based, not risk based, emission limits.  And they would apply only to major sources and not to the large number of smaller establishments that account for substantial 1-BP emissions.  For all of these reasons, EPA's risk evaluation should include air emissions as a significant risk pathway, and SACC should recommend that these emissions be included in the final evaluation.  Thank you. 
                  MS. TAMUE GIBSON:  Thank you.  Ms. Suzanne Hartigan from EarthJustice.  Oh, I'm sorry.  I apologies.  Excuse me -- American Chemistry Council.  I apologize. 
                  DR. SUZANNE HARTIGAN:  Thanks for the correction.  Good morning and thank you for the opportunity to speak to you today.  My name is Suzanne Hartigan, and I am a senior director in the Regulatory and Technical Affairs Department for the American Chemistry Council.  ACC appreciates the opportunity to provide feedback on the next draft TSCA risk evaluation for consideration by the SACC.  As the SACC peer review has been scheduled before the end of the public comment period on October 11, we've had an abbreviated time to review the documents and provide feedback for SACC's consideration.  But we will provide further detailed comments by the October 11 public comment deadline.  
                  I'd like to highlight several points for the committees consideration as it reviews the draft risk evaluation for 1-Bromopropane.  First, EPA has incrementally improved some of the weight of evidence narratives in draft risk evaluation for 1-Bromopropane.  However, much work still needs to be done regarding the process of integration of evidence for cancer mode of action.  On page 158 and -9 of the draft, EPA describes their rationale for the evaluation of the evidence supporting a mutagenic mode of action.  However, the draft risk evaluation for 1-Bromopropane does not fully evaluate, discuss, or weigh the scientific evidence of the potential alternative interpretation of the data and analyses using the cancer mode of action approach.  As a result, EPA has not transparently characterized the potential risks associated not only with the default option but also the potential risks from biologically plausible alternatives as required by the risk evaluation rule.  In its TSCA risk evaluations, EPA should more clearly and transparently present biologically robust weight of evidence assessments where data integration is required, such as the OACD adverse outcome pathway methodology or the mode of action approach initially championed by the WHO international program on chemical safety, which has been adopted by EPA's Office of Pesticide programs.  
                  The evaluation of the cancer mode of action should begin with evaluation of the quality of relevant studies and evidence.  However, EPA does not appear to perform the full systematic review process, including data quality evaluation and data integration according the TSCA systematic review principles and guidance for the in vitro studies.  None of the in vitro genotoxicity assays are included in the supplemental review file -- data quality evaluation of human health hazard studies.  We note that EPA does have data evaluation criteria for in vitro studies but either did not apply them or did not publish the evaluation in the docket.  There is a discussion of evidence in Appendix I 5-6.  However, it is not supported by a data quality evaluation.  The SACC should provide feedback to EPA on evaluation of the data quality of these studies and why EPA preferred to rely upon limited in vitro assays in its analysis over in vivo animal studies.  
                  Further discussion in Appendix I 5-8 in the supplemental file for the human health hazard assessment, a systematic review of this point does not appear consistent with the EPA systematic review guidelines.  The SACC should consider the weight of evidence supporting a non-mutagenic threshold mode of action for tumor formation resulting from exposure to 1-Bromopropane in laboratory animals.  And in sum, the SACC should provide feedback to the EPA regarding the sufficiency of information for acceptance of an alternative mode of action.  
                  Secondly, regarding the exposure assessment, we believe the SACC can provide additional input to improve the overall quality of the assessment.  ACC encourages the SACC to provide EPA feedback on its assumptions within the exposure models themselves.  We note that in the draft risk evaluation of 1-Bromopropane the assumptions are not always well described and often do not represent real world conditions.  
                  Ventilation controls, for example, are not considered within any of the inhalation models, very likely leading to exposure overestimates.  Similarly, EPA's assumptions for dermal exposure are not well-explained.  The draft risk evaluation uses a permeability model rather than fraction absorbed model for consumer exposures.  The opposite, however, is true for the dermal occupational exposure calculations where EPA uses fraction absorbed.  The context for each model is important because permeability models ignore evaporation, and 1-Bromopropane is a high volatility chemical where evaporation is expected to play a factor in exposure.  
                  Finally, the risk determination section for 1-Bromopropane could use further streamlining and clarification.  We recognize the document is draft and EPA has the opportunity to further edit the document based on feedback from peer reviewers and stakeholder comments.  However, as we commented on for the previous risk evaluations, the risk determination sections are critically important for communication to the public.  Therefore, EPA should give particular attention to the final risk determination sections and should seek to clarify the basis for the determinations and better summarize and highlight the overall determinations in order to improve the public's understanding of them.  Thank you for your time. 
                  MS. TAMUE GIBSON:  Thank you.  Operator, do we have Mr. Jonathan Kalmuss-Katz on the line from EarthJustice? 
                  MR. JONATHAN KALMUSS-KATZ:  I am on the line.  Can you hear me in the room? 
                  MS. TAMUE GIBSON:  Yes.  Hi.  Could you go ahead and proceed with your public comment, please? 
                  MR. JONATHAN KALMUSS-KATZ:  Thanks very much.  Good morning and thank you for the chance to speak.  I'm Jon Kalmuss-Katz from EarthJustice.  Today I will focus on three ways in which EPA fails to account for known exposures and risks in the 1-Bromopropane risk evaluation.  
                  First, as you've heard from other commenters, EPA ignores exposures to 1-BP from ambient air, water, and soil.  As a result, EPA does not evaluate any risks to the general population.  EPA acknowledges that these risk exist.  The risk evaluation states that people who are exposed to 1-BP in ambient air also have higher levels of 1-BP metabolites in their urine.  EPA has also warned that 1-BP is, quote, sufficiently volatile and toxic to pose inhalation risks via vapor intrusion from contaminated soil and ground water.  Yet the draft risk evaluation does not consider any of those exposure pathways.  
                  EPA cannot disregard 1-BP exposures merely because they could be regulated under other environmental laws.  Instead, TSCA requires EPA to evaluate all risk associated with a chemical's known, intended, and reasonably foreseen conditions of use.  The law permits EPA to consider potential regulation under other laws only at the risk management stage after EPA has completed a TSCA risk evaluation and found unreasonable risk.  Moreover, EPA has not regulated 1-BP under the Clean Air Act or the Safe Drinking Water Act.  And there's no federal law limiting the migration of soil vapor into indoor air.  
                  For the last nine years, EPA has failed to act on a petition to list 1-BP as a hazardous air pollutant.  Even if that petition were granted this afternoon, it would likely be years before any new regulations are finalized, and EPA may never set risk based emission standards for many 1-BP sources under the Clean Air Act.  The unrealized possibility of regulation under other environmental laws does not satisfy TSCA's mandate to protect vulnerable subpopulations from the risks currently posed by 1-BP.  
                  Second, EPA overlooked widespread evidence -- sorry.  EPA overlooked evidence of widespread exposure to 1-BP.  Biomonitoring studies have detected a urinary metabolite of 1-BP in most of the people tested, including 99 percent of pregnant women.  Despite describing this metabolite as a valid biomarker for 1-BP exposure, EPA did not consider these background exposure levels in its dose response analysis claiming that in 2016 the peer review panel for a prior 1-BP risk assessment had advised against using such data.  
                  As the SACC members who served on that panel may recall, EPA's position fundamentally misrepresents the committee's recommendation.  The minutes from that Chemical Safety Advisory Committee meeting supported the use of biomarker data in order to, quote, avoid overlooking a significant exposure scenario and stated that such consideration would benefit a revised risk assessment.  Dr. Kissel said, "I think you should maybe do a little more with the existing biomarker data that you already have identified."  
                  Since 2016, the ATSDR and IARC have both used that same urinary metabolite as a biomarker for 1-BP.  To the extent EPA had questions concerning the specificity of the biomarker, it has had more than three years to resolve them and ample authority to do so.   EPA's failure to do so does not excuse the Agency from considering relevant and reasonably available exposure data.  
                  Finally, EPA's selected points of departure characterized 1-BP's risks based solely on animal studies even though available epidemiological data reveals neurological effects and cellular damage at much lower doses.  EPA scored all of those human studies as acceptable under its systematic review, but it questioned their reliability in an appendix to the risk evaluation.  Many of EPA's stated concerns, including the reporting of time weighted averages as opposed to peak concentration and the use of median exposure levels at the lower end of the reported range, would have biased the studies away from a finding of risk.  Yet they still reported adverse effects at levels below EPA's point of departure, confirming that EPA has understated 1-BP's noncancer risks.  
                  There is no question that 1-BP presents unreasonable risk.  Only by considering all exposure pathways and all relevant data can EPA evaluate and protect vulnerable subpopulations from those risks.  The draft risk evaluation fails to do so.  We urge the SACC to advise EPA to correct those flaws.  Thank you. 
                  MS. TAMUE GIBSON:  Thank you.  There's a slight change to the public commenter.  Ms. Swati Rayasam will not be providing her comment.  However, Dr. Tracey Woodruff will take her five minutes.  So in total, Dr. Woodruff will be speaking for ten minutes.  Dr. Woodruff, are you on the line?  Operator? 
                  OPERATOR:  Yes, Ms. Woodruff is connected. 
                  MS. TAMUE GIBSON:  Dr. Woodruff, are you on the line now? 
                  DR. TRACEY WOODRUFF:  Can you hear me now? 
                  MS. TAMUE GIBSON:  Yes. 
                  DR. TRACEY WOODRUFF:  Okay.  Great.  Thank you.  Hi.  Good morning.  My name is Tracey Woodruff, and I'm a professor at the University of California San Francisco School of Medicine and the director of the program on reproductive health and the environment.  Today, my comments will focus on the application of systematic reviews and our concerns that will follow on to the others related to exposure assessment in the 1-Bromopropane draft risk evaluation.  Can you go to the next slide, please?  I have no conflicts to disclose.  Next slide, please.  
                  EPA is required by the TSCA statute to use the best available science and the weight of scientific evidence to make decisions about chemical risk assessment.  And in the risk evaluation rule, which they codified in 2017, EPA defined this weight of scientific evidence as a systematic review method that uses a pre-established protocol to comprehensively, objectively, transparently, and consistently identify and evaluate each stream of evidence, including strengths, limitations, and relevance of each study to integrate the evidence as necessary and appropriate.  Next slide, please.  
                  However, there are several important parts of this definition that are key to EPA having the best evidentiary basis for its decisions in terms of using systematic review for the evaluation, not only of 1-Bromopropane, but of the other chemicals that are being evaluated under the first ten chemicals.  So the problem with the current approach that EPA is using is that it misses critical pieces of the required methods that are required as part of systematic reviews.  So one of those critical missing pieces is creating a protocol for all the review components.  
                  That is supposed to be done before EPA engages in the risk assessment.  This helps minimize bias and ensures transparency in the decision making.  This is also part of best practices that is used by all established methods for systematic reviews and has been recommended by the National Academy of Sciences in their report related to the IRIS risk assessment program.  Other elements of the method that are not consistent with best practices include a nonempirically based scoring system, which is also not recommended, use of metrics not relevant to study quality that are actually about reporting and not relevant to the quality of the studies, and they have methods that are questionable and not transparent that can exclude potentially relevant studies.  We have written about this already in a peer reviewed paper on the problems with the risk assessment method -- the systematic review method that EPA is using.  Next slide, please.  
                  This is a flow diagram of the -- literature flow diagram for the environmental releases and occupational exposure data sources.  You can see that EPA has identified 52 key sources that were taken forward to data extraction and evaluation.  And 37 of these are excluded by this phrase or this process of hierarchy of preferences.  First of all, there's a lack of clarity on how EPA chose to -- and evaluated the key sources.  That's the 52.  We have previously given comments on both 1,4-dioxane and HBCD risk evaluation about how EPA has failed to give -- to have a consistent protocol despite the risk evaluation rolling out clear guidance on how to select studies.  
                  Additionally, this 1-Bromopropane risk evaluation only briefly mentions this hierarchy of preferences, which is mentioned in a caption, and it is not explained.  The risk evaluation does not provide transparency in terms of what sources of information EPA relied on and which ones they excluded, which goes against the critical tenant of transparency within the systematic review process.  This hierarchy of preferences I will also note has not been through a public comment or peer review process.  So it is difficult to understand how EPA is using this method to select these studies.  So given that there are concerns about how the method is using and collecting studies, it is very difficult to have confidence in the final risk evaluation of the evidence base and thus lead to potential problems with the risk assessment itself.  Can we please go on to the next slide?  
                  As I said, we've previously commented on the systematic review methods that EPA is using.  I'd like to talk about the issue around vulnerable populations and issues in EPA's exposure assessment for 1-Bromopropane.  There are many factors that can make the population at risk or more vulnerable to exposure to toxic substances, including proximity to polluting industries, social vulnerabilities, and biological susceptibility such as age, gender, preexisting conditions.  We agree with EPA that there are important susceptible time periods, but we are concerned that EPA is not adequately considering them in this draft risk assessment.  As previously has been noted, 1-Bromopropane -- the manufacturing was between 18 million to 26 million pounds, and exposures come from primarily spray adhesives, dry cleaning, and degreasing.  Can you go to the next slide, please?  
                  As has been noted by the previous commenters, there are several areas that EPA is not considering in the exposure assessment.  This will have a high likelihood of underestimating exposures to 1-Bromopropane.  One is that they're not considering inhalation exposure pathways.  The reason that's given is because it's supposed to be listed as a hazardous air pollutant.  That was previously commented on by a previous commenter.  The other is is that there has been reported measurements of the metabolite of 1-bromopropane widely detected among pregnant women, as well as the population, and that there is data to indicate that there is potential for very widespread nonoccupational exposures to 1-Bromopropane.  Can you go to the next slide, please?  Next slide, please.  
                  The other thing I wanted to comment on which has not been raised by the reviewers is that EPA is assuming that children may only be present for four hour periods from 3:00 to 7:00 after school, which they could have exposures that are similar to occupational non-exposure levels.  However, this has two problems.  One, it does not consider children who are too young for school or not in school and therefore may be exposed for a longer period of time and that children in family owned dry cleaners are likely to spend their time outside of school in the dry cleaner for a much longer time than is indicated by EPA.  
                  So according to EPA, dry cleaners are open for 12 hours a day, six days a week.  That's 3,672 hours a year.  According to the National Center for Educational Statistics, children of school age are in school around 1,195 hours per year.  That means there are 2,476 hours that children spend not physically in school.  And there is a high likelihood that they are in the business.  This would represent a majority of children's time spent, 67 percent, in the dry cleaner with the same exposure as the other nonoccupational -- occupational nonuser's exposure and could mean a higher risk because they are also both developmentally more vulnerable to the exposures to 1-bromopropane.  So while EPA talks about presenting the full range of estimates, the lack of consideration of possibility of widespread exposures in the population, both as indicated from admittance into air as well as measurements through biomonitoring and exposure to the most vulnerable subpopulations, this can lead to scientifically inappropriate underestimated risk to the public.  Next slide.  
                  Along the lines of these comments, we have recently published a paper in PLOS Biology on population susceptibility, highlighting three ways that EPA's approach under TSCA is not consistent with current science, both that there are biological and social factors that can make people more susceptible to chemicals.  EPA is not fully integrating this information into their risk assessment -- that EPA needs to consider the multitude of exposures that can lead to a totality of exposures that can lead to increased risk of exposure to chemicals, in particular 1-Brompropane, and that these practices together can lead to EPA underestimating risk and thus leading to increased risk among the population.  Next slide, please.  That's it for now.  Thank you. 
                  MS. TAMUE GIBSON:  Thank you.  There are some members of the general public who wish to provide a public comment.  There are three in total.  Mr. Bob Miller, Jr., could you please come forward and provide your comment, please?  Thank you. 
                  MR. ROBERT MILLER, JR.:  Thank you.  Good afternoon.  Thank you for the opportunity to speak today.  My name is Bob Miller.  I'm the Senior Director for Global Product Stewardship for the Albemarle Corporation.  We appreciate the opportunity to provide comments on the U.S. EPA draft risk evaluation for 1-Bromopropane at this peer review panel meeting of the Science Advisory Committee on Chemicals.  Albemarle Corporation is a global manufacturer of specialty chemicals among whose product line includes 1-Bromopropane as an important part of our portfolio.  
                  As a manufacturer and in conjunction with leading experts and key users of our products, Albemarle has developed considerable expertise with respect to 1-Bromopropane.  For today's comment, I would like to take the opportunity to provide some additional details on behalf of my technical team in support of our written comments submitted on August 30.  Specifically, I'd like to address the determination of pulmonary tumorgenicity in humans and provide some salient points for the committee's consideration as it reviews the draft risk evaluation.  
                  Albemarle has collaborated with outside experts in human pulmonary pathology to evaluate the applicability of lung tumors in mice towards predicting a potential pulmonary tumorgenicity in humans.  Lung tumors in mice induced via chronic inhalation of certain chemicals are not relevant to the risk of developing human lung cancer because mice are super metabolizers of certain chemicals as compared with rats or humans.  Mice metabolize these chemicals through cytotoxic metabolites that injure cells, thereby increasing cell proliferation rates.  Increased cell proliferation rates amplify background mutation rates resulting in a significant false positive rate for tumor development seen in the two year NTP rodent cancer bioassays.  This mechanism has been extensively described by Cohen and Ellwein (phonetic), Aims and Gold, and Mulguvocar (phonetic), among others.  Recent work by Tomasetti et al. have supported the relevance of increased cell proliferation as a risk factor for development of human cancers.  
                  A number of chemicals are known to be metabolized in mice by the pulmonary P450 enzyme, CYP2E1, for example, styrene, nepheline, ethylbenzene, and coumarin.  Chemicals metabolized in the mouse lung by this have been shown to induce lung tumors through the formation of cytotoxic metabolites that damage Clara cells, induce cell proliferation, and result in the formation of bronchoalveolar carcinomas.  1-BP has been shown to be metabolized in the liver by this same mechanism.  The existence of a high level of metabolites of 1-Bromopropane in mouse liver strongly implies that pulmonary metabolism is also occurring.  
                  1-Bromopropane forms the same tumor at the same location following administration via inhalation, as does the styrene, nepheline, ethylbenzene, and coumarin.  Since these substances form lung tumors in mice via enzymatic formation of cytotoxic metabolites, it can be safely assumed that 1-Bromopropane is forming lung tumors via a similar mechanism, since mice have been shown to form cytotoxic metabolites via liver metabolism.  The lowest risk numbers reported in the 1-Bromopropane draft risk assessment were derived from lung tumors in female mice only as tested in the two year inhalation study conducted on rats and mice by the NTP.  
                  In our opinion, expressed in a recently published peer review manuscript, mouse lung tumors represent a level of sensitivity to chemical carcinogenesis that is much higher than would be expected in humans.  That is the NTP two year inhalation study on 1-BP suffers from interpretability issues due to problems in extrapolating bronchioalveolar lung tumors in mice to pulmonary adenocarcinomas in humans.  Reliance on this data has the effect of developing a calculated risk that is higher than warranted, and we would encourage the SACC to consider this information as you review the draft risk assessment and respond to the charge questions, many of which seek your review in support of these very points.  Thank you again for the opportunity to provide comments today. 
                  MS. TAMUE GIBSON:  Thank you.  Mr. Ben Gann from American Chemistry Council, please. 
                  MR. BEN GANN:  Good afternoon.  I'm Ben Gann, Director of the American Chemistry Council in the Chemical and Products in Technology Division, or CPTD for short.  We are pleased to provide comments on EPA's draft risk evaluation for 1-BP.  CPTD represents more than 60 chemical specific groups focused on the business of chemistry and issues relevant to chemical manufacturers and downstream users.  As the committee conducts the peer review process for the 1-BP risk draft risk evaluation, CPTD recommends that the Agency consider the following items.  
                  First, halogenated solvents are used in industrial and commercial settings because they are essentially nonflammable and reduce the overall fire risk.  A significant need exists in the marketplace for cleaning solvents with the wide solubility parameters and excellent cleaning capabilities of 1-BP.  Limiting solvent choices could result in an abrupt and significant change for industrial and commercial facilities that are designed to handle materials rated as nonflammable.  
                  Second, a condition of use that may not have been adequately evaluated by EPA is processing of 1-BP for use as a recyclable reaction solvent.  Under this condition of use, it is not intended to be incorporated into a solvent but rather is used as an intermediate in which the reaction takes place.  Third, greater context is needed for less recent air sampling data.  In January 2014, the American Conference of Governmental Industrial Hygienists decreased the threshold limit value time weighted average for 1-BP from ten parts per million to 0.1 parts per million.  Thus, engineering controls and personal protective equipment, or PPE, used to meet the recommended occupational exposure levels at the time a study was done may not meet current recommended exposure levels.  It is within this context that data for inhalation exposure pathways should be evaluated.  
                  Fourth, we should not lose sight of the fact that EPA found that there was no unreasonable risk under some conditions of use, including domestic manufacturing, importation, processing as a reactant, distribution disposal, and recycling.  In addition, EPA found that there was no unreasonable risk to the environment under all conditions of use.  Fifth and finally, exposure levels for each condition of use vary depending on volume, engineering controls, and use of PPE.  EPA acknowledges that it does not have information on the effectiveness and prevalence of engineering controls despite recognizing their effectiveness in reducing occupational exposure. Although PPE is the last means of worker protection in the hierarchy of controls, EPA also recognizes the effectiveness of PPE and reached a finding of no unreasonable risk for most conditions of use where respirators were used and risk estimates did not exceed the risk benchmark for inhalation exposure.  CPTD looks forward to continuing to work with the Agency as it moves forward with the risk evaluation process.  Thank you for the opportunity to speak today. 
                  MS. TAMUE GIBSON:  Thank you.  And Richard Morford from Enviro Tech International. 
                  MR. RICHARD MORFORD:  I can waive my comments. 
                  MS. TAMUE GIBSON:  Oh, I see.  Thank you.  Okay.  This is the end of the submission of the public for public comments.  At this time, I have the time now as 12:36.  We are ready for lunch, so we'll be back here at 1:36 to begin the charge to the panel. 
                  
                  (BREAK)	
                  
                  MS. TAMUE GIBSON:  Hello, everyone.  Hello, everyone.  The peer review meeting for 1-bromopropane -- we will begin, and I will turn it over to our chair, Dr. Portier.
                  DR. KENNETH PORTIER:  Thank you.  I guess before we start with the issue questions, I open it up for one last -- see if there's anyone from the public that wanted to make a comment before the panel.  We kind of, I don't think we rushed through it, but we had a lot of commenters this morning.  We had a lot of topics to discuss.  We want to make sure that everyone who wanted to have a say before the panel has had that opportunity.  I don't see anybody jumping up.  We'll give them 15 seconds.  Courage is waning.  Okay, we'll close that, and we'll go on with the issues. 
                  EPA has provided the panel with a fairly large list of charge questions, and our DFO has set an ambitious agenda for today for us to get through four questions.  I have my doubts, but I'll do my best to do it.  And so we're going to go ahead and start, and, I guess, Dr. Anitole, you're going to be reading the questions.  And I've asked her to read the whole question for 1, 2, and 4, but she's going to shrink Number 3 because it's way too long.  Dr. Anitole.
                  
                  
                               CHARGE QUESTION 1
                  
                  DR. KATHERINE ANITOLE:  Okay.  Beginning with Charge Question 1: Systematic Review.  This relates to Section 1.5 and the supplemental documentation.  To meet the scientific standards required by TSCA, EPA applied systematic review approaches and methods to support the Draft Risk Evaluation of 1-BP.  Information on the approaches and/or methods is described in the Draft Risk Evaluation as well as several supplemental documents.  
                  Question 1.1: Please comment on the approaches and/or methods used to support and inform the gathering, screening, evaluation, and integration of data and information used in the Draft Risk Evaluation for 1-BP.
                  Question 1.2: Please also comment on the clarity of the information as presented related to systematic review and suggest improvements as warranted.
                  DR. KENNETH PORTIER:  Thank you.  And Dr. Blystone has the lead on this.
                  DR. SHERI BLYSTONE:  Yes, so I'm going to make a couple of comments, and then we would like it to go over to Dr. Davies.  She has some stuff to show before we get to the other members, so I want her to talk through her stuff.  
                  In general, I think you've heard this comment before.  You're going to keep hearing it until it happens is that peer review is required for the screening criteria.  
                  I think we have continuing challenges with clarity.  If I go to 1.2, at least I had that because I'm not necessarily an expert on systematic review, but I tried to follow some stuff through.  And, whenever I tried to follow a reference, either from the systematic review up to where does it end up in the assessment or vice versa, I seemed to get either lost or run into a roadblock or something.  And then, at some point, I just sort of threw up my hands and gave up.  But there's some challenge to following it all the way through that I think I'll spend some more time to try to come up with some ideas of how we can prove the clarity, but it is a challenge, I would just say.  
                  I will mention that Dr. Portier pointed out one particular instance to me that I also found in the Systematic Supplemental File for Fate and Transport Studies.  In Table 3, there was a reference to a study in hydrolysis that was rated as low, and, under comments, it said that an article was not available.  We found that article within two minutes on Google, so it seems to suggest that I don't know how thorough all of this stuff is.  If you have one of those instances, is there more instances of that?  But I verified it with -- I have the study.  I'll give it to you, so you can have it.  It's publicly available.
                  I've seen other in -- public comments also sort of suggested there are times where things don't quite seem to match up with the criteria in the systematic review.  There was one public comment that was talking about two studies done at the same laboratory.  The same people were rated differently, and the clarity of why that was may need to be improved.  
                  So I kept hearing those kinds of questions about it.  It's difficult to sort of follow it through, and I think Dr. Davies has more to say about that.
                  DR. HOLLY DAVIES:  Yes, thank you.  I tried to follow it through.  So I do want to start by saying I do have a lot of the same comment, of course, that I had for 1,4-Dioxane.
                  DR. SHERI BLYSTONE:  Excuse me, I think we wanted to show some stuff, right, Don?
                  DR. HOLLY DAVIES:  So I can get to that, yeah.  I'll tell you.
                  DR. SHERI BLYSTONE:  So he can pull it up while you're talking.
                  DR. HOLLY DAVIES:  Okay.  I do have a lot of the same comments, of course, as for 1,4-Dioxane as EPA, of course, has not redone the systematic review, and we don't want to lose those.  Some of the parts are too stringent, some not stringent enough, not enough explanation of terms and the procedure, simplifying used more like the existing systematic reviews, confusion on the gray literature; and Tracey Woodruff also reiterated a lot of that this morning in her public comment.
                  Also, I wanted to mention the systematic review lays the foundation for the risk evaluation, the discussion of the rest of the risk evaluation will provide insight into that.  I had looked through the Draft Risk Evaluation kind of from the perspective of the systematic review, and I'm going to comment on that and not on the quality of the risk evaluation.  I'm going to try to kind of limit to not is this a good paper or did this lead to a good risk evaluation but more on how the systematic review went and look forward to insights on how the risk evaluation went.
                  So I really wanted to follow through like how this worked because I was expecting it to flow.  I was expecting a comprehensive search with the baseline of their other evaluation that were done fairly recently.  And then a comprehensive search like what was on topic and then do the data quality evaluation and then the synthesis leading.  And I'm just kind of expecting that, and I never really saw that.  As Sheri just said, that flow wasn't really there.  
                  Okay.  So maybe Don could show this.  Okay.  So start with this.  This was the first figure, literature flow for fate and transport.  I started with this one.  It was smaller.  I thought it would be easier.  There's nine on-topic peer-reviewed references from the bibliography.  And then there's seven in the data quality evaluation; but, to my surprise, the seven in the data quality evaluation are completely different from the nine on-topic peer-reviewed references from the literature search.  And so it's like, where did those seven come from?  There's just this complete disconnect.  
                  And so I eventually ended up coming up with figures, which is why I made this, to figure this out because there's no mention where the seven papers came from in the data evaluation, nor can I figure out this -- I never could figure out any of these.  I could not follow any of them, so I took everything, and I put it into Excel spreadsheets, which Don also has, but I don't think we need to look at all the Excel spreadsheets.  So I have Excel spreadsheets of all the data quality evaluations and all the references and then other things too so I could compare them.  So I have like 12 spreadsheets in my Excel file.
                  So, for this Figure 1.6, many things -- so, in the bibliography, there was about 981 search results.  I assume that's because they did the search again.  That's not in the bibliography.  I don't know what data screening means so much, but they excluded 98.6 percent of the references.  So I have kind of both big and small comments.  So that implies that the initial search criteria were too broad.  You're excluding 98.6 percent of them there.  The data evaluation says 18.  There's only seven in the data quality evaluation, so I don't know where that is since there's only seven in the document.  The text says they entered full screening when relevant and went into data evaluation.  
                  The 12 that were unacceptable based on the evaluation criteria, none of them were unacceptable in the Data Quality Evaluation document, so I don't know what that means.  And the data extraction's not mentioned in the data quality.  And zero were added from the key in supporting, so I don't know where those extra ones came from, so I never figured that out.  I could never get the numbers to match up even by putting everything into Excel spreadsheets.  
                  So I finally -- if you want to go to the next slide -- I did try to kind of figure out where things came from.  Here's my figure.  Another minor comment: it's better to use the same words even something like going from environmental hazard to ecological hazard.  It's hard to know in this when we're trying to follow it.  I guess that means something different or the same thing.  The strategy has different words than the beginning of the section versus what's in the bibliography versus what's in the data quality evaluation are called.  So this is one I was hoping it was going to kind of flow nicely with different bubbles.  
                  So the nine fate didn't go anywhere, although, one of them does show up in the references and is used later in the risk evaluation.  But none of them go to the data quality evaluation.  Four of them come from existing OPP sources; and three of them, I don't know where they came from.  So that's where they came from.  So I was trying to figure all of those out.  
                  I also looked at where they were used.  Table 1.1 Physical Properties -- almost none of that was in through the data quality -- the DQE.  None of those references had been evaluated.  These were used in Appendix C, and some of them in Table, I think, 2.1 in one of the tables.
                  So for occupational -- the next one's occupational -- on environmental release and occupational exposures, again, this had -- I just could never get these numbers to work for all the same reasons.  The numbers in the data quality evaluation didn't match this.  The ones that were not integrated didn't match.  Where did the key environmental ones come from?  The other thing that caught my eye here -- well, if you look down -- next slide -- you can see where these come from: Some of them come from engineering.  Some of them come from human health hazard search.  
                  I ended up getting distracted here because of the criteria.  Because, what I noticed when I was looking through these, looking through the Data Quality Evaluation, was that a lot of them had scored really high for methodology, and they scored high for methodology because they were an EPA study or they came from a federal agency or they were from -- they had been published in a peer-reviewed journal or, my favorite, because it came from King County.  And so they must have used good methods because, as a former King County employee, it's always nice to know that King County is so highly thought of.  
                  So I went back to the methodology document and that is not the criteria for a methodology for high, and so they weren't following the methodology.  And I did look at -- and this is where methodology is not being followed.  I don't think following it would have improved the risk assessment.  And they're different, so some evaluators scored it high for any federal agency: TURI, CARB, consultant funded by CARB, academic labs.  One of my favorites was just four letters: H-S-I-A.  That's it.  It must be high because the Halogenated Solvents Industry Association [sic] did it.  But then other ones, not ECHA, not NEWMOA, so somehow TURI was good but the NEWMOA not.  ATSDR and OSHA were sometimes were not rated high, so it was just weird, inconsistently done things.  
                  So I looked and, to actually be high, the sampling or analytical methodology has to be approved OSHA or NIOSH method or well described and found to be equivalent to approved OSHA or NIOSH methods.  And one of the key things is that for the occupational -- and for this type of study, EPA explained -- and this, again, is all back in the 2018 application of systematic review -- the reason for the approach is to avoid admitting potentially valid data or information since occupational exposure and release data is often sparse.  So really, they just set this incredibly low bar, and they said that in the methodology, and so then it was very few things.  Many metrics had no unacceptable.  There was no unacceptable for sample size.  There was no unacceptable for metadata.  There was no unacceptable for several of the metrics.  And it was really hard to get the low.  And so it was setting a low bar and then not having a lot of different kind of high, medium, low and then not following the criteria they did have.  
                  And then sample size and minor is another comment -- sample size is not actually sample size but whether there're statistics to describe the sample size, so it would be helpful to have criteria that are named in a more descriptive way.  Metadata is one of the criterion, and it actually is for sample types, exposure types, sample durations, exposure durations, worker activities, and exposure frequency.  So it's one criterion to cover like everything in the study, and there is no way to get an unacceptable rating in the occupational exposure rating.  So it's a very -- and this just really conflicts with the human health criteria where, if you don't bind to receptors, it's unacceptable, the whole study is out.  So it's a very different criteria for the different kinds of studies.  
                  And what had caught my eye about this was, as I was going through it, it just seemed like it was not a very useful criteria when everything was high, and it wasn't differentiating.  So I went through in my spreadsheet and I looked at -- I put down for all of the occupational ones what it was.
                  Also, the models: while I was looking at the criteria, I was really surprised that one of them for the other one was models for the fate EPI Suite with one of the seven was EPI Suite.  Most of the criteria was NA, and so it got a pretty high score.  And going back to the methods document, there are no criteria for models under fate, which is why it was hard to -- for the person to rank it.  
                  But there's actually, there are criteria for models for engineering and occupational exposure, and, yet, in the document, they used tons of models for occupational releases and occupational exposure, but they didn't rate any of those models.  So none of those went through this data quality evaluation even though they do in fact have criteria for them.  I don't think the criteria would have been particularly useful because, in order to be unacceptable, the model has to have like mathematical mistakes in it, so they all would have passed this incredibly low bar.  Okay. 
                  So consumer and environmental, again, -- you could go to the next one -- it's going to be the same.  I can't figure out this flow, where the ten -- where they come from.  Three were from the exposure on-topic literature search.  One's from human health hazard, two or three from the existing OPPT sources.  Two of them aren't in the bibliography.  One of them was unacceptable.  A couple of them aren't acceptable actually.  Three of them had data extracted, but it's not clear why.  Several with perfect scores didn't have data extracted, so then it's not really sure how this criteria works; if they're perfect and then they don't have data extracted.  Also one of these is the draft ATSDR profile which always says, "Do not cite"; so we should probably not cite that one.  And the final ATSDR profile has the wrong year because they're both 2016, and one of them, the final, is actually 2017.  
                  Also something that would improve the systematic review is consistent citations.  So some things where -- especially since the EPA has the citation source, and so things change citations.  As you're going through, it makes it hard to find them.
                  In the consumer exposures in the document, it doesn't explain -- so like when they look for consumer exposures, they don't explain why they didn't use the ones that they have that went through the data quality.  They found uses.  They didn't find them through the systematic review process, and the sources didn't go through the data quality evaluation that they mention in the text.  
                  And there's other statements like, on page 115, "all three parameters had a range of documented values within literature identified as part of the systematic review."  I don't think that's true.  
                  For ecological hazard, so this is -- I wonder what my -- go down to my picture, Don.
                  DON:  This one?
                  DR. HOLLY DAVIES:  Yeah, the next one, this is -- oh, that's where the consumer exposure came from -- from all different places.  The next one's ecological hazard which is a little more straight forward because there's only one of them.  You go to the next -- but again, here it says there were two in the data quality evaluation when there's only one in the document, like that's not hard to count.  I don't know where the other one is.  
                  And this had more explanation in Appendix 5 -- I mean F, Appendix F.  So we know the key in supporting studies, those five are the ECHA studies.  So that at least is a little more explained but not totally.  And all of these again, have so many excluded.  But, if you go to the next slide, you can see in the environmental hazard, the ones that came from the literature search: one is the Geiger study; five of them go to human health hazard showing that all the search terms for environmental hazard included all the model organisms.  So some of the key -- like the Hanma et al., 2003, which was used for one of the PODs, was an environmental hazard.  All the model organisms, the mice and rat studies, were in environment instead of as human health, so the search terms should be fixed there.  And, of course, Richard Denison mentioned just the issue with the studies not being fully looked at, the ECHA studies.
                  So it was in human health hazard, the next one, that I finally -- a little bit more became clear as I was looking through it because there's more explanation later in the evaluation in the human health section and in Appendix I.  So it named the foundational documents, the draft risk assessment, the NIOSH criteria, and the ATSDR tox profile.  And it was more clear about the systematic review process only being used for dose response and talking about several places.  The papers for dose response were then put through the data quality evaluation, and ones that weren't used for dose response weren't put through that.  
                  There's still a lot of terms that I'm not sure are being defined, the key and supporting studies, the influential studies; it also mentions an informal evaluation for overall data quality.  At some point, I had put in all of the -- everything that had a data quality evaluation and all of the references so I could see that less than a quarter of the references went through the data quality evaluation.  
                  So it's really just -- I mean, I really had been expecting from this systematic review, I had been expecting this systematic -- that this review, data quality evaluation, and integration and that was what was used to do this.  So going through this also -- and it says, papers for dose response had the data quality evaluation, other papers didn't.  Other papers are there for supporting evidence.  And some of them -- it was not mentioned in the risk evaluation, but there's one of the papers that failed the data quality evaluation was still used as supporting evidence.  I don't think it should have failed because it failed for one of those overly strict criteria where the -- it was a male reproductive study that was six weeks instead of eight weeks.  But it failed under the existing criteria and was unacceptable and then was mentioned five times in the rest of the document.  And so all of the explanation of the hazards which were just a mix of both: ones that had gone to the data quality evaluation and ones that had not.
                  Also there was an unacceptable criterion that was not rated unacceptable overall, and then the Hanma et al. has an issue.  That one was high for acute, but it failed for chronic; and then it ended up being used as the chronic POD.  So I'm looking forward to that and what people think about that in the human health because it's described in the document as high for this chronic endpoint that it's looking at, but it was rated -- it was evaluated twice in the document again; once for acute and once -- it was failed just on the one endpoint.  
                  So, looking at this, there were 29 in the data quality evaluation.  It's hard to tell where they came from.  I got totally distracted again.  I couldn't find where they all came from, but the next slide shows kind of where they came from and, again, just to show where they -- it wasn't this kind of expected kind of flow of looked for human health, found human health, evaluated human health.  They kind of came from all the different bibliography topics, and I didn't end up finding where all 39 came from.  
                  The human health epidemiology came from, I think, one exposure, one environmental, and one human health on the bibliography.  
                  So I did look.  Not all of the studies in the data quality evaluation got into Table I-2, and I don't know why because I-2 is supposed to be the big summary.  I didn't check to see if like all the good studies.  It just seemed like there's throughout the whole thing, there's a lot of studies that were rated good in the data quality evaluation that were not integrated into the document, and there was never any explanation of why similar to what Dr. Woodruff was saying earlier in -- like they had all the studies and they were dropped off.  Even after the data quality evaluation, there were these really good studies by their criteria, and they just kind of left, so you don't see them later.  In other tables in, like, the fate and transport and such, where the tables in the document are populated by studies that didn't go through the data quality evaluation.
                  Oh, another question, and I'm trying to -- this will all be a lot more -- when I tried to put this together, one of the things I'm trying to do is gather like questions on like what does this word mean?  Like confidence rating, there's a bunch of confidence -- like is confidence rating the same thing as the data quality score?  It seems like it is, but I'm not sure.  But some tables will suddenly start saying confidence rating.  
                  So then I stepped back with the big picture.  Well, if you go to the next slide, this is what I find -- that was the human health hazard assessment that they said.  The next slide is like what I think is going on.  So, looking at this, I kind of thought like, okay, so you have these foundational studies.  They talk about these key and supporting studies.  The influential ones go to the data quality evaluation but some of them just go right to the Draft Risk Evaluation.  They don't get evaluated and the Draft Risk Evaluation talks about narrative, synthesizing and integrating evidence, supporting qualitative, and more than 75 percent of them don't have data quality scores.  
                  They also talk about the backward referencing, searching, and targeted supplemental searches.  We don't know what those are.  Some of those, I think, go into the data quality evaluation.  The systematic review available literature with the PICO scores that are on topic and at least in the human health ones, it's suitable for dose-response analysis; they went to data quality evaluation.  But some of them that were on topic went into the Draft Risk Evaluation without going through the rest of it.  And some of them, even if they weren't on topic, went right there.  And then there just seems to be this ongoing literature searches, communication with industry.  I put the five ECHA studies there, this informal evaluation, like this ongoing thing that's also kind of going in.  
                  So this is what I kind of came up with because I kept trying to kind of think what was happening.  I think the next one was what -- this is what I hoped was happening.  This is what I thought was happening.  I thought everything was fitting into kind of different sources because it totally makes sense that you'd start with the assessments that are done and that you'd go into the data quality, but I thought the systematic review with the new systematic review was going through the data quality.  And, of course, you'd also have ongoing new information coming, but I thought it was kind of funneling through this data quality evaluation in going through this integration.  And that's kind of what I was being confused on when I kept trying to find things.  
                  Anyway, so I kind of took a step back with going back to TSCA.  So the big picture is I went back, and I thought, what does TSCA really say about systematic review?  I didn't remember systematic review in TSCA, but I was mostly focused on state preemption at the time.  So, of course, TSCA says nothing about systematic review, which you guys know.  There's a new part in TSCA that says scientific standards, a new part of scientific standards, that EPA wrote into their rule into the definition of weight of evidence.  And, in the rule in the beginning of Section 1.5, it talks about, kind of implies, that TSCA requires systematic review.  
                  So this systematic review doesn't show the scientific standards as Section 26 requires in TSCA.  Since most of what's going into the Draft Risk Evaluation isn't going through the systematic review.  So I don't think it's meeting EPA's description of systematic review and goes from the preamble.  I don't think it's meeting TSCA's -- showing the scientific standards.  I don't think it's bad science.  I think they're good papers; they're just not being shown through the systematic review process.  
                  So that's my -- I think that would be everything so me trying to actually dig in and figure everything out.
                  DR. KENNETH PORTIER:  I'm glad you're going to write that thing up.  I mean, I followed most of it and the diagrams help, but it would be good.  Yes, Dr. Blystone.
                  DR. SHERI BLYSTONE:  This is Dr. Blystone.  Just to follow up in that you can sort of see from the way that she graphically did this why it can be a challenge to try to figure out how any particular reference was used and where it came from and where it went through.
                  DR. HOLLY DAVIES:  So, if you go to the previous one, Don, the previous one is, I think, what they actually did because you can find references and you're like, why didn't it go?  
                  DR. SHERI BLYSTONE:  And it's not necessarily that any of that is incorrect; it's just is it clear?  Is it understandable why anything was done in a particular way?  And I think that that's where we -- there's room for improvement there to understand in each of those cases where the arrows went in the way that they did that that's very clearly described in the document.
                  DR. KENNETH PORTIER:  Dr. Cobb, do you have anything to add?
                  DR. GEORGE COBB:  I have a few things to add, but I will say that what we just heard covers a lot of my questions or comments, so I'm going to try to be as brief as I can.  A couple of things: on page 20, there's a part beginning industrial and commercial, and it seems to contradict the exclusion of cleaning and degreasing products, exclusion that's written on page 20.  That's more of an inconsistency comment, and I have a couple of those.  And I just entered that one to say that I have a couple of inconsistency kind of things I'll provide comment on.  
                  So, on page 44 in the literature review for fate and transport, it appears to me that only atmospheric transport routes were considered.  And given that 1-BP can be applied to ground or can be introduced to the soils, groundwater is something I think we should be considering in that literature review and that's -- we can expound on that as much as we'd like but that's -- to get that in the record.
                  And I'll echo just so that I'm echoing what we just heard from a previous presenter, the exclusion criteria that were used for excluding specific papers, whether it's individually or collectively, somehow, needs to be in there that these were excluded because they didn't have measurable data or because the species wasn't properly identified or whatever that exclusion criteria would be very helpful in understanding why things were excluded.
                  Also, on page 42 of the assessment, it states that literature search terms included environmental fate and transport, engineering releases and occupational exposure, and exposure to general populations.  We heard this morning that general populations were not considered in the evaluation, so, to say that the data screening included that is misleading at best.  And so I suggest that that wording be somehow repaired, reexamined, or that a paragraph stating why specifically those things were no longer included because I believe that was part of the problem formulation that this was part of what should be considered.  I think that is it.  Thank you.  
                  DR. KENNETH PORTIER:  Ms. Rudel.
                  MS. RUTHANN RUDEL:  Thanks.  I did not have anything to add.
                  DR. KENNETH PORTIER:  Do any of the panelist on the phone want to comment?  Dr. Anderson, Cory-Slechta, and Dr. Pessah?
                  DR. ISAAC PESSAH:  No comment.
                  DR. HENRY ANDERSON:  Oh, I think we've covered it.
                  DR. KENNETH PORTIER:  Dr. Pessah, did you have something to add?
                  DR. ISAAC PESSAH:  No, I just said I don't have any comments.  It's been covered pretty extensively, the concerns I mean.
                  DR. KENNETH PORTIER:  Thank you.  Anyone else on the panel wish to comment?  I don't see any flags up.  I think this would be probably our most comprehensive comment on systematic review that we've had so far.  So I look forward to having your review organized.  I look forward to having that.
                  DR. HOLLY DAVIES:  Well, actually, I forgot to say something, believe it or not.  Table 2.1 and just so that I can put it into my notes -- Table 2.1 -- so the first one says the study quality's high but, it's not in the Data Quality Evaluation.  And there's another one that says low that's also not in the Data Quality Evaluation, and I'm not sure if there's -- there's two references in one cell and the low could just be referring to one of the two references and the other ones are EPI Suite, which is just kind of weird to have done.  So there's just other examples like that.  That was when I started and said, how is it used?
                  DR. KENNETH PORTIER:  Okay.  And I'll look to EPA, do you have any comments or clarification questions of the panelists on their response.
                  DR. STAN BARONE, JR.:  Yeah, this is Stan Barone.  So thank you, Dr. Davies and Dr. Blystone, for your synthesis here.  I think you've done a very good job identifying some weaknesses in the narrative.  I think your graphics here, if we could get those sooner rather than later, that will help us in discussions that we're having on the next group of assessments.  I think you've outlined in your graphics some weaknesses in our explanation.  I think, if you looked at your last graphic, that was our goal and that was the approach that we were taking with the different disciplines.  
                  If you go to Number 13 where all the key and supporting studies, all the critical studies for the different disciplines, go through the data quality review.  Now, we apparently did not show the flow very clearly, and I think that's what you've outlined are some weaknesses that we need to address.  And some of this has been brought up in the other comments in a sort of one-off individual kinds of issues, but I think you've done a more systematic job.  So thank you very much, and I think this will be useful for us in developing our protocols which we work currently underway for the next one, so thank you.
                  DR. KENNETH PORTIER:  Okay.  I think we're done with Question 1.  Let's move onto Question 2 on occupational exposure assessment.
                  
                               CHARGE QUESTION 2
                  
                  DR. KATHERINE ANITOLE:  Question 2. Occupational Exposure Assessment, Section 2.3.1 of the Draft Risk Evaluation.  EPA evaluated acute and chronic exposures to workers for conditions of use in industrial and commercial settings.  For exposure via the inhalation pathway, EPA quantified occupational exposures for both workers and occupational non-users based on a combination of monitoring and modeled exposure concentrations.  For exposure via the dermal route, EPA modeled exposure for workers, accounting for the effect of volatilization and glove use.  EPA assumed dermal exposure would not occur for the occupational non-users. 
                  EPA assumed that workers and occupational non-users would be adults of both sexes -- greater than 16 years and older, including women of reproductive age -- based on occupational work permits.
                  Question 2.1: Please comment on the approaches and estimation methods, models, and data used in the occupational exposure assessment.  
                  Question 2.2: Please provide any specific suggestions or recommendations for alternative data, or estimation methods that could be considered by the Agency for conducting the occupational exposure assessment.
                  DR. KENNETH PORTIER:  Thank you.  Dr. Blando, you have the lead on this.
                  DR. JAMES BLANDO:  Yes, I do.  Thank you.  This is Jim Blando here, so I guess I'm the lead discussant for Question 2.1 and 2.2, and, I guess, we'll just take Question 2.1 first, obviously.  So I know we have to verbalize all of our comments at this meeting.  I have some very specific comments that I wouldn't necessarily characterize as substantial but I think, perhaps, the best approach for us would be to have all of the discussants list their specific points and maybe rather than me leading off -- although, I'm happy to do that -- maybe we'll start with the other discussants if they want to bring their points up to the full committee.  
                  The first person on our list is Dr. Henry Anderson.  He's on the phone I know.  Henry, I don't know if you had any specific points that you wanted to raise at this time for discussion.
                  DR. HENRY ANDERSON:  Well, I have a couple of points that I think we've heard a lot about the occupational exposures.  One thing I haven't seen -- and it may be in the document, but I just haven't been able to search properly to find it -- is some more information on the OSHA inspections that I think we heard a little bit in the public presentations about the number of respiratory protection violations specifically.  And, in all of the four compounds we've had so far, one of the key issues raised, both publicly and by a number on the committee, has been the appropriateness or the use of assumptions of full respiratory use continuously rather than having it be quite so either everybody's using it or we're not using it.  
                  It would be helpful to put some of the information from OSHA.  I've been onto their website to try to deal with their data set on their inspections, and I just haven't been very successful at it.  But, I think, I would also ask, has EPA queried OSHA for what has been their experience with 1-BP or specifically with the respiratory protection standards that they have?  How well has that adhered to worker understanding?  There's a number of -- there's quite a bit of literature on some of that.  
                  So one of my things I would say is I think the discussion in the document is not sufficient to justify the making a determination of acceptable risk around acceptable risk based on whether people are using respirators or not.  I think that's problematic, and, I think, getting more quantitative data on that would be helpful.  
                  The other thing that I was struck by there's a lot of information on this chemical, specifically -- and some of the others -- on occupational information.  You have the NIOSH criteria document for 1-BP, and you have considerable experience by OSHA.  And I'm assuming, as part of the interagency review, there must be some written comments from NIOSH or OSHA about the documents, and I would -- maybe there isn't any, but, if there are, I think that would be very helpful for the committee to have copies of those comments what their -- because they're really the boots on the ground as it relates to what's going on in industry and less so in the consumer side of things.  But it would be very helpful to get what they may have responded in their comments to these documents.  So, once those exist, I would ask if EPA couldn't make those available to us because, while we have -- or I have some -- general experience in industry, I certainly don't have as much experience as both NIOSH and OSHA.  And OSHA databases don't seem to be utilized here very much.  
                  So that's one of things that -- and then the other comment I had, or my question would be, it would seem would be useful in the document to have a combined inhalation and skin exposure estimate, and the skin exposure is calculated based on the air or the respiratory risk.  So that basically says that, if it's in the air, it will also get deposited onto the skin.  So it's pretty hard to have an inhalational hazard that doesn't contribute to the skin hazard.  So it would seem to me, as far as workers are concerned, it's pretty difficult to have a skin exposure without a respiratory exposure.  So it would be useful to have some combined consideration for worker health and safety.  That, I know, makes it a bit more complicated rather than just deal with skin or just deal with lungs or ingestion, but they all do combine to deal with the exposures.  
                  And then, I think, as I recall, the review of some of the EPI data, specifically on the neurological side, it appeared in those studies that the humans were quite a bit more sensitive than the values that were chosen out of the animal studies.  I wasn't overly convinced that there was fatal flaws in the EPI studies.  I think they could be looked at or offered alternative set of estimates even if you end up feeling more comfortable using animal data.  
                  So that's kind of some of my comments that I was -- I think the assessment or the evaluation of some of the studies as we already -- we just went through with problems, but I would think it would be helpful to have some discussion about kind of a holistic exposure to the worker which would combine the respiratory exposure and the skin exposure.
                  DR. JAMES BLANDO:  Henry, thank you very much.  If I might just add to Henry's comment because it's similar to my first comment as well.  I know we've had a lot of discussion about PPE previously, and I'm sure you guys are looking forward to hearing more discussion about PPE, but I personally have had difficulty utilizing the OSHA inspection database to do this kind of assessment for me personally.  But what I can tell you is that for myself, as personal experience, going to a lot of these facilities that handle bromopropane is that I understand the logic that an industrial facility with professional staff would be more likely to have a robust industrial hygiene program properly utilizing PPE and perhaps the small facilities or family-owned mom and pop shops would likely not have a good program.  
                  I would just point out that I've also been to industrial facilities that have fairly poor industrial hygiene programs, and, in fact, I would point out that the 2008 MMWR from Dr. Perrone on the two cases, the index cases, that we had: the one being the dry cleaner and the other one being an industrial worker in a wave solder room.  That while he worked for a fairly large circuit board manufacturing company, their personal protection, the PPE, was not very effective at all.  In fact, their vapor immersion degreaser, the cooling coil, didn't work, didn't wear any PPE at all, and their ventilation was essentially basically just general ventilation in the building; so big surprise, he had very, very high exposures.  
                  So, I guess, the question I put out to everybody on the panel is, while I think it's important to point out in the risk assessment document that PPE can greatly reduce risk and makes some estimates about how that might impact exposures, I do think it's also worthwhile for the risk assessment to acknowledge to any decision-maker or user of the risk assessment that, in fact, we shouldn't assume that PPE will be utilized effectively.  I do know the risk assessment and the written documents does point out estimates of pre-engineering control or PPE use or no PPE use; however, for me personally, this is maybe just a minor editorial comment is, as I read through the document, it seemed to imply to me that you really did expect that any industrial facility would, in fact, be using effective ventilation and effective PPE.  And I don't know that that should necessarily be implied.  The thought being that any decision-maker who's going to make a decision by looking at this document or anybody who might utilize this document, I think, it's important to just clearly acknowledge that we really don't know what the compliance would be with personal protective equipment, and that can make a very distinct and very large difference in the exposure and risk estimates that you might generate from this document.  So I don't know how you could reword the document or what you could add to the document to just make that even more clear, but that would be my comment piggybacking on Henry's comment.
                  Henry, I guess that's all that you had for now?
                  DR. HENRY ANDERSON:  Oh, just you've reminded me another would be, while their determination is there's no unreasonable risk in some of the manufacturing facilities, I would just point out that, even though they're a closed system, there are still repair and maintenance workers that spills or leaks or things like that are occurring at the same time.  So we have to pay attention to the types of workers in the various facilities that a manufacturing facility has all sorts of different kinds of workers that need to be looked at either as individuals with a specialized or unusual opportunity for exposure even though it may be a small proportion of the workforce.  
                  And the other thing that seemed to be discounted was the use of urinetic biomarkers sort of things.  I think there's quite a bit of literature on those and of course we have the NHANES data that showed that general population is exposed to this compound as well.  So that kind of would build on what is the risk of people living in the neighborhood of facilities either manufacturing or using these materials.  
                  The ATSDR review specifically mentions people who may be living in the vicinity of a facility.  And again, dry cleaning and those kind of things as smaller plants often will have a discharge.  There may be not too many people living right nearby, but that is a general population exposure that I don't think we ought to completely discount.
                  DR. JAMES BLANDO:  Thank you, Henry.  One thing about the biomarkers that were discussed, and the toxicologist would know better than I would; I was always under the impression that especially from talking to Kevin Hanley (phonetic) that the biomarkers that they looked at -- the urinary biomarkers -- weren't necessarily specific for just 1-bromopropane.  So I will tell you I have wondered, when we hear some of those estimates about general population measurements of that particular urinary biomarker, I mean, I would ask the toxicologist on the panel if, in fact, that is specific for one 1-bromopropane or less, and, Dan, I see you have a comment.
                  DR. DANIEL SCHLENK:  I brought this up three years ago, and, no, it's not specific.  
                  DR. JAMES BLANDO:  Okay.  So that's important to note when we talk about those urinary biomarkers.  So why don't we move on to our next discussant just to make sure we get everyone.  Dr. Barton.
                  DR. CHARLES BARTON:  Thank you.  One of the questions I had was regarding ventilation and engineering controls.  Ventilation wasn't discussed much in the various scenarios as an engineering control, and so I wonder, was it taken into account in the modeling, you know, ventilation-wise?  I see you're nodding your head.
                  DR. STAN BARONE, JR.:  Stan Barone.  A quick answer: yes.  In many of the scenarios where we did modeling, we're looking at room exchange rates, and those room exchange rates are surrogates for ventilation engineering controls.  That's in addition to the monitoring information assessment which included room exchange engineering controls, pre and post, so there are published studies and both NIOSH studies that also looked at that.  So we have probably, out of all the assessments you've seen thus far, this one has the most analysis for the conditions of use looking at engineering controls by condition of use.
                  DR. CHARLES BARTON:  Thank you.  No further questions.
                  DR. JAMES BLANDO:  Okay.  And how about Dr. Bennett?
                  DR. STEVEN BENNETT:  Yeah, sorry.  It's Steve Bennett.  A couple of questions or clarifications: I think one of the things that I looked at this and I'm going to focus a little on the aerosol spray degreasers -- to some on there who have some of the most background -- but I want to make sure that the estimates that they're describing are representative of today's market.  I think they reference a 2002 study for an estimate for about a thousand to five thousand business using 1-bromopropane, and it's unclear if that's still the current practice or the current number of the businesses.  And I think of the changes of the institution of the OSHA PELs as well as the AC -- I never get that acronym -- ACGIH TLV as well as the 2016 risk assessment, how much an impact that's had out in the marketplace?  And I don't know.  That an area of, I think, that if there's a way to clean some of that up, that would be helpful.  
                  Certainly, I think, in comparison to the 2016 products that were listed in the document versus those that are listed in the preliminary document for here, there's fewer.  And I certainly know that within our membership there's fewer companies that have an interest in this methodology, so I just want -- I don't want to belittle the concerns but make sure that that's accurately representative of the current marketplace.  
                  Another piece that I'll raise is of a little of our same issue we'll go into the consumer question which we have in Number 3 is oftentimes these products, especially in the aerosol space, our design for commercial industrial uses; but they may be available to consumers.  They're not designed for consumers, but they may be available and try to vary how to differentiate between those, clearly understand that you need to take the appropriate risk to ensure that they don't get into the wrong users.  That a trained users, if they're going to using it in there, but there's clearly you have to have that differentiation there with as much as possible, granted, I know, there's a variety of regulatory authorities that come into play and certainly you want to make sure.  But that's a piece that's somewhat in there.  I'll just kind of put those out there.
                  DR. JAMES BLANDO:  Dr. Kissel.
                  DR. JOHN KISSEL:  I have multiple comments.  First, I'd like to note one of the public comments that we got that wasn't presented orally today was from Adam Finkel who's the second former highly placed official at OSHA who has bothered to kind of take issue with the way EPA is dealing with the PPE here, and I think some of those comments should be taken to heart.  A couple of points that he made -- one has to do with -- which I haven't heard anybody make -- had to do with breakthrough.  Compounds with this kind of vapor pressure tend to break through.  This is the same problem with methylene chloride that you can't tell when your respirator isn't actually doing anything anymore and it may actually be exhausted in way less than an actual work shift.  And so there's lots of opportunity for people who think that they're wearing a respirator to be exposed to basically whatever the ambient concentration is; and I think that's cogent point because my experience is, even in places where people are encouraged to wear respirators, there's not checking as to what the capacity of that respirator is on a regular basis and their replacement may, in fact, lag the exhaustion of the device.  
                  Secondly, we have -- there's reference to a hazard alert from OSHA which I quote says, "1-BP easily travels through most glove materials."  Now, that's pretty vague, but apparently only relatively sophisticated and expensive gloves will actually stop BP from getting to skin, so we have another level.
                  Many places aren't going to provide gloves, that it would be up to the workers to provide their own gloves.  And they may, for economic reasons, not change those gloves out very often, or they may buy the cheapest things that they can get at the hardware store because they think they're protected when they're not.  
                  And so, unfortunately, the reference or the citation for this alert is a textbook which I couldn't get to online, and so I don't know how thoroughly that's been tested.  I don't know how materials have been challenged with 1-BP, but this seems like a compound for which the glove protection factor is even a worse assumption than usual.  So that's a problem.
                  Another thing pointed out by Adam Finkel was that actual OSHA data tend to have air concentrations in the tens of parts per million, and I didn't find air concentrations like that in this document.  And part of that is it's kind of unclear sometimes whether the air concentrations are post-control or pre-control so it's a little hard to follow.  It would be better if the document was a little more clear on what the starting concentration is and what the effective concentration is after some type of protection.  
                  Also with respect to the dermal exposure, most of the dermal exposure seem to be starting with a number in the vicinity of two grams per day of exposure to the material.  In our previous encounters with HBCD and 1,4-Dioxane, I think those kinds of estimates were coming from EASE, which is a European exposure model, but I'm not sure where that came from.  In this document, I couldn't find any basis for that starting point.  Once I had the starting point, I could reproduce the calculations and see how you got the doses, but I didn't see justification of what that starting point was and that's, again, that's that transparency issue.  
                  With respect to multipathway exposures, this compound -- generally, you would anticipate on the basis of its physical/chemical properties that inhalation would swamp dermal vapor uptake.  But one, there's some reason to believe that the theoretical estimates of how available this thing is dermally might be off the mark a bit.  
                  And, secondly, once you put respiratory protection on, then you elevate the importance of those other pathways.  So, if you decided that you didn't have to worry about dermal vapor because it was only ten percent of what inhalation would be, but then you put a ten X correction factor on the inhalation, you haven't cut the exposure to ten percent of what it was; you've cut it to 20 percent of what it was because you have to add back in the 10 percent that was dermal vapor.  
                  And so, if you're going to do calculations where you're making assumptions about these protection factors, then it's important to add back in those other pathways which may have been discounted because you would traditionally expect, in an unprotected version, that they wouldn't be important, which, again, plays into the aggregate dose.  It would be good to see aggregate doses.
                  The last comment I had was about the Frasch studies.  So you're using Frasch et al. for a fraction absorbed, which there's been comment about why flip back and forth between fraction absorbed and permeability approaches and you really don't need to flip back and forth.  You could use the Frasch data to get to a permeability approach.  Basically, you can see in the Frasch data -- that's really a good paper --  you can see in the data that you've got competition between stuff evaporating and stuff being absorbed.  And what's going on in those experiments is the evaporation rate is high enough that, basically, the experiments get cut off very rapidly.  And the dermal absorption is basically, how long is the reservoir still on the skin?  But the uptake that Frasch found was actually quite rapid.  
                  So the permeability coefficient that is implied by the NIOSH equation, that EPA is using and the consumer data, produces a permeability coefficient from water that's about a third of what the Frasch data suggests; and this is quite possible.  The overall maximum flux, the experimental maximum flux, reported by Frasch is about 30 times the theoretical maximum flux that you would get from using the NIOSH approach.  And that's actually within the air bounds of these kinds of calculations.  When you're doing prediction of behaviors based upon physical/chemical properties and the physical/chemical properties themselves can be an order of magnitude off when you're talking about log Kow and some of the other things, vapor pressure.  
                  There's pretty wide error bars on those estimates, and so there's suggestions here that this compound goes through skin much faster than from the Frasch empirical results.  There's reason to believe that this compound is underestimated by the traditional methods, and that's potentially a problem, both on the occupational side and on the consumer side because those numbers are getting used.  
                  One other thing to mention -- so it is a kind of timed experiment.  Basically, you can get dermal exposure.  If you unoccluded skin, it's basically how fast does this stuff evaporate because that's when the experiment is basically called off.  Frasch, to his credit, actually measured the wind speed in the hood where he was conducting this work and came up with .3 meters per second, 30 centimeters per second, which is about three times the number that Frasch and Bunge recommend as a typical indoor velocity for -- and I brought this up last time -- hood velocities tend to be faster than actually human experience indoors.  
                  So it's possible that that one of the reasons that that stuff evaporated so fast in the Frasch experiments is that the wind velocity was actually quite high, and the mass transfer coefficient on the gas side varies as the velocity to the .78 power.  So, basically, the evaporation rate would go down by a factor of about two and a half if you cut that wind speed down to a tenth of a meter per second instead of three-tenths of a meter per second.  And so that would allow more uptake and so that .0016 percent absorbed would be too low probably for a scenario in which the wind speed wasn't that high.  So those are my comments for now.
                  DR. JAMES BLANDO:  Thank you, Dr. Kissel.  So I just had a couple of comments to add as well.  I just wish I could see what I'm doing here. 
 Sorry.  
                  UNIDENTIFIED FEMALE:   It's okay.
                  DR. JAMES BLANDO:  All right.  I just had a couple specific comments, and I know I have to verbalize all these comments for them to be included in the record.
                  The first comment that I had in addition to some of the things that have already been said, on page 39 of the supplemental occupational information, there's discussion about a 90 percent reduction from a push-pull LEV-type ventilation system.  I was just curious, and I think it would be important for EPA to describe in the document why they decided on that ventilation system for that degreasing open-top degreasing system.  I know, in some of the facilities that I've been to, I haven't seen that used with degreasers -- not that they aren't; I'm sure they are -- but that 90 percent reduction, I would presume, assumes that the operation of that system is done properly.  I know that with push-pull ventilation systems, if your supply and exhaust does not balance properly which is, from an engineering perspective, is very common with those systems; you can actually make situations worse.  So you certainly don't want to blow the vapor out into your workspace, so, I think, EPA should maybe add some more discussion about why that system was picked and perhaps just make a note about the operation of that system and some of the challenges that sometimes people face in the real world with operating those systems.  
                  The second point I had was on page 84 in the supplemental occupational information, EPA stated that using the NAICS and SOC codes may overestimate the number of workers exposed which is certainly very reasonable, although, I would just -- in my opinion, it's just an editorial comment -- I thought that perhaps that rather than saying overestimates since you really don't know if you don't go to the facilities, it might be better to say you might not have a very precise estimate as opposed to assuming it's an overestimate.  
                  Page 86 in the supplemental information, EPA notes and assumes that the workers remain in the respective exposure zones, and that this may overestimate their exposure.  While this true for the workers, I would point out that for the occupational nonusers, it's very possible that they too might not remain in their nonworker far-field zone.  
                  I know when I've been at facilities, it's not uncommon for workers to be friends with people in the plant.  I know I was at one of the facilities that we sampled and one of the clerks spent a lot of time talking to the operator of the dry-cleaning machine, so there was definitely mixing both ways.  So it might be better to rather than characterizing that always an overestimate, although it might overestimate for the worker, it might underestimate for the occupational nonuser.
                  Page 87 on the supplemental information for occupational, EPA discusses using a constant emissions scenario and points out the exposures that likely vary over time -- which, of course, I'm sure they do -- and indicated that this might overestimate exposures by using a constant emission scenario.  I would just argue from a variability in air levels, as we showed in our 2010 paper with the dry cleaners, you can have -- if you have really significant magnitude in your variability, it's kind of hard to say that assuming if constant exposure's going to be an overestimate; it might actually be an underestimate.  If you look at our real-time TBA analysis that we did with those dry cleaners, in that paper, I think, one could argue that unless you assumed a really high constant emission rate, I'm not so sure that it would be an overestimate.  So perhaps you could be more tentative in saying that would be an overestimate, maybe rather say that it all depends on the specific scenario whether it would be an over- or underestimate.
                  Page 87 supplemental information post engineering controls assuming 90 percent reduction was based on TCE.  I would just say for the engineers that it would be important to keep in mind that 1-BP is -- if I understand and remember correctly -- 1-BP's much more volatile.  I don't know what impact that would have on your assessment of engineering controls.  I do know that in some cases when we had folks who were heating this product that the higher volatility resulted in a higher emission rate when they compared it to using TCE.
                  Page 88, one of the points that I had, which is towards the end the supplemental info, for the dry cleaning under the uncertainty point number 2, EPA stated that the small near-field volume results in a large spike in concentration when the machine doors open, find the residual 1-BP solvent is assumed to be instantaneously reused into the near field.  In reality, the residual solvent will likely be released continuously over a period of time.  I would not agree with that.  I think our video exposure monitoring data clearly shows that, in fact, the scenario you do have when somebody opens up the front door is you do get a really rapid spike in that near-field zone, and I would argue that the operators and machine are not really going to -- when they're trying to pull clothes out of the machine, they're not necessarily going to be able to just walk away from the machine.  So I would not agree with that uncertainty point number 2 that it would be a constant release over a period of time.  That's not what we observed in our video exposure monitoring results which, if anybody would like to see, I could show them.  I have it here on my computer.
                  Page 84 in the Draft Risk Assessment -- I'm going to skip that; that's not that important.  One of the other clarification points -- EPA, in the Draft Risk Assessment, discusses and utilizes the terminology of the 8-hour TWA and the 12-hour TWA.  As you guys know, OSHA has a specific policy on extended work shifts, and I think it would just be clear when you're using the terminology of, especially for the 12-hour TWA, to make sure to the reader that -- and I'm assuming what your 12-hour TWA is a denominator of 12 hours -- but the only confusion that, I think, could occur is that for an industrial hygienist looking at that, if they think you're following the OSHA extended shift policy, they might think it's actually over an eight-hour period and that you're sampling the eight hours of worst exposure.  What you're really doing is computing it over 12 hours, I just think you need to be clear in the document when you're saying 8-hour TWA or 12-hour TWA that you just clarify for the reader, especially if it's an industrial hygienist that's reading it, that you're not referring to the OSHA extended shift policy where, even though it might a 12-hour shift, you're only supposed to sample 8 hours to follow the OSHA policy.  So, if you're not doing that, I would just make sure that that's clear.
                  And then on page 103 -- so I'm getting to the end here -- on page 103 in Appendix B, the calculation for acute concentration computation, to me, was a little unclear.  You essentially had an eight-hour TWA multiplied by the exposure duration divided by the averaging time.  What felt unclear to me about that was that the eight-hour TWA implies that time is already figured into the calculation.  So it's sort of you have an eight-hour TWA that is trying to account for a time-weighted average -- time's already included -- and then you're multiplying it by another timed term which is exposure duration divided by averaging time.  
                  I presume what you're trying to do with that calculation is adjust your eight-hour TWA for somebody who works longer than an eight-hour shift if that's what you're trying to do there.  But what I thought was confusing about it, to me, is when I was looking at that computation was that the eight-hour TWA already presumes that you have time figured into the calculation.  So, primarily editorial, but I would just clarify so that a reader can really understand what you're really talking about.  
                  And then the last comment was from Dr. Portier who sent me an email about the Tank Truck and Railcar Loading and Unloading Release and Inhalation Exposure Model.  Dr. Portier noted that -- not to speak on your behalf, but I'll just mention I didn't forget -- that in the draft document, it wasn't cited anywhere although it was described in the supplemental information.  So, if there is no citation because that's something custom that you guys have done, at least in the draft document, you might want to refer or put a citation to the supplemental info so, if somebody wanted to get that information, they know exactly where to look.  Did I cover --
                  DR. KENNETH PORTIER:  I was going to say, subsequent to my note -- this is Ken Portier -- I kept tracking -- I don't know.  You get obsessed with some of these things.  I wanted to know where that model, that Tank Truck and Railcar Loading and Unloading Release and Inhalation Exposure Model, came from.  First, it's got the longest name of a model I've ever seen.  But it refers to the supplemental information, and you go to the supplemental information and you, essentially, have the same information as the document.  It doesn't take you anywhere, but I think I tracked it down to ChemSTEER.  And there's loading and unloading module in something called ChemSTEER, which is a model that's out there and it's been evaluated.  And so I'm assuming all these parameters that are talked about go into that document somewhere, but it's not clear.  I mean, you really have to search to come up with that.  You dig and dig and dig.  
                  One of the questions I wanted asked back to you and other panelists is there are a lot of models used in this especially the occupational exposure and what your opinion was about the adequacy of the use of those models.  I mean, I think this particular tank truck model sounds reasonable to me.  I didn't see anything about the model that raised a flag, but I don't know the models like some of you know those models.
                  DR. JAMES BLANDO:  I would just say -- not being a modeling guy, more of a field guy -- I have an inherent bias that I like to see monitored data actual, what I would call real data -- all the other models would hate me for saying that.  But I did note when I read your report that you seemed to indicate that you had fairly decent agreement for the models and the actual sampling data that you had, and that felt like a good thing to me.  I don't know if other folks have an opinion on that, but I felt that was very strong.
                  DR. KENNETH PORTIER:  This is Ken Portier.  I just remember back in 2016 when we looked at this.  The panel had a lot of discussion especially about the exposure modeling -- the open and the close, and the inner and the outer space -- and all of that is back in here and I think is cleaned up and is written clearer than the document back in 2016.  And I just kind of wanted to say that I appreciate the EPA putting the effort to kind of clean that up as one of the recommendations from the panel.  Dr. Davies.
                  DR. HENRY ANDERSON:  I would agree with -- it was helpful with how -- well, they didn't provide the data for that as to how close it was that there seemed to be reasonable consistency between the actual measured data and the modeled data.  I think that's an important statement to make as a tendency is to use models and then pretty soon you really are -- there's no data that it's based on but earlier when the model was first constructed with different chemicals and things.  So I think that's helpful.
                  But the other thing that I think is getting better that way is not to be quite too emphatic about the results that come out that you have to be willing to accept that, qualitatively, the high and the low are different but actually having a linear relationship or whatever between them may not be quite so accurate.  So I think the variability in it is something you need to be able to say, if they're usefully used, then they're very helpful.  But they tend not to be quite so quantitative as they're often used.  I would say qualitatively they're very helpful.
                  DR. KENNETH PORTIER:  Dr. Davies.
                  DR. HOLLY DAVIES:  I was just going to say that I also wanted to get other people's opinion because it was laid out well, the pictures were nice, the end, the agreement, was nice, so it was explained well, but I do like the monitored data also.  
                  DR. KENNETH PORTIER:  Dr. Kissel.
                  DR. JOHN KISSEL:  So I would like to add that I think everybody who was just talking about modeling was talking about air models, and, you know, air is a relatively friendly medium.  It's well mixed.  You can assume that you have a well-mixed tank and not be too far off with air.  Surfaces don't work like that all.  And I would say that the dermal encounter modeling as the current state of the art is extremely primitive.  And, if anybody can guess within one or two orders of magnitude how much stuff you're actually going to contact on your skin, you're doing pretty well.  
                  Oh, and if I could add, I amend my prior comments because I noticed on my -- I left one thing off.  In the evaporation study that Frasch et al. did, they put gauze on top of their device because they were trying to mimic a rat study that had been done.  One possibility is that the gauze actually wicked the materials away from the surface and caused it to evaporate more rapidly than it would have otherwise, so that's another little caveat.
                  DR. KENNETH PORTIER:  This is Ken Portier.  I'm a modeling guy rather than a monitoring guy because I've worked with people collecting data.  There's so many opportunities for bias, whereas, a model is a, I think, exercise where you're really putting it together, and it should come together and work.  But I'm also one that says you've got to calibrate it and validate it, and that's why I like the fact that they're using this kind of more peer-reviewed models which in my mind equates to something more like a validated model.  
                  But it's only as good as the inputs going into the model, right.  And then we've seen some previous examples where we didn't agree with the quality of the inputs going in.  And I just kind of question back to you guys was whether you felt that the data going into these models was of enough quality that you could believe what came out of the model, like the EPI Suite stuff.  I mean, these things seem to depend a lot on the EPI Suite outcomes, and I, personally, don't know how to judge the quality of that, but I'm hoping somebody on the panel does.
                  DR. JAMES BLANDO:  I could add that I know when we've done things like interworker comparisons or even with the same worker over different shifts, we have found you can have really wide variability.  I know a lot of the sites that I go to there's always one guy or girl -- I'm very bad -- is like the sloppiest of the sloppiest worker.  And, when you compare their exposures to somebody else who's a little more diligent, you can find -- I know we have found coefficients of variation greater than a hundred percent between workers on the same shift on the same floor.  So that's one issue you could have when you're trying to model that that you could have this extremely wide variability that, I imagine, would impact the models as well or the outputs that could come out of a model.  I don't know if other folks have the same experience, but that's definitely been my experience.
                  MS. TAMUE GIBSON:  I just want to say something real quick, please introduce yourself when you speak, yeah.
                  DR. JAMES BLANDO:  That was Jim Blando.
                  DR. KENNETH PORTIER:  Thank you for reminding us.  Ken Portier here.  The other aspect is the high, medium, low exposure levels.  I mean, do you agree or feel that those are a fairly good description of the range of exposures, especially in the occupational setting which is the -- did you guys look at that?  Dr. Anderson, do you want to join in?
                  DR. HENRY ANDERSON:  No, I mean, my concern with all of this is there tends to imply precision in modeling that isn't necessarily there, and I would agree that, depending on what goes in and what you use, it makes quite a bit of difference as to what comes out of the model.  And high, medium, and low -- I didn't specifically try to go through all of the literature to see what's out there.  That might alter that.  I'm not sure the high levels are really truly reflective as how high things can get or did get in the past.  
                  I think that's the other comment that I didn't say much about as it's hard to know how much outreach was made to industry to get information from them or how responsive they were, but we did have some written comments about how the discharge rates when you rely on TRI data from 2016 and we're now three years later what they currently are.  It would be helpful if there was some kind of a survey through the industry, or industry organizations, and kind of ask them, and they pull together some information, and then kind of take a look at that as part of the added information.  I mean, just relying on what's in the literature, the choice of where you go to do your studies is often made by who will let you in or agree to doing this study, so it isn't necessarily reflective of the industry, especially when you get into the smaller facilities.  I mean, the dry-cleaning studies, I think, really show that.
                  DR. KENNETH PORTIER:  Thank you.  Dr. Blando, I'm going to ask whether you've got enough for Question 2.2 as well or whether you had additional comments on that.
                  DR. JAMES BLANDO:  Just for me personally I just had one short brief comment for Question 2.2, but I don't know about the other discussants if they had significant -- to add so we'll do that, okay?
                  So for Question 2.2, please provide any specific suggestions or recommendations for alternative data estimation methods that can be considered by the agency for conducting occupational exposure assessment.  I just have one quick comment, and then we can ask the other discussants.  
                  For me, the only other sources of data that we've used in the past, that it wasn't clear to me if you guys used them or not used them, was I know we identified -- in the dry cleaning study that we did, we identified users of 1-BP through the minor source permit program through our Department of Environmental Quality.  The air permits.  People have to submit those air permits and those have identifiers and have information that can be useful for identifying.  
                  For the number of dry cleaners in the state of New Jersey, when we did that study, we used the Dun & Bradstreet.  I think it used to be called the "I" Database, but now I think it's called Selectory (phonetic) database.  But it was basically a business database that told us, based on searching for NAICS codes, told us how many facilities there were in the state that had -- you have to have DUNS number -- but how many facilities they had in the state and how many employees they had.  So that's how we were able to come up with our estimates at that time for how many dry cleaners we expected.  In the state of New Jersey, we used a Dun & Bradstreet database.  
                  So I don't know if you utilized those databases if they would be helpful.  But those are two other sources of information that I thought, if you hadn't used them, they might be helpful in doing some of the worker estimates and so on.  
                  So, if we go to the other members of the charge question folks, Dr. Anderson, did you have anything you wanted to add for Question 2.2?
                  DR. HENRY ANDERSON:  The only thing, there was -- I think, it's part of the OSHA database now, but there's no OSHA Consultation Program which specifically is targeted to small businesses.  And so that might also be a source of information.
                  DR. JAMES BLANDO:  Okay.  Thank you.  How about Dr. Barton?
                  DR. CHARLES BARTON:  No further comment.
                  DR. JAMES BLANDO:  Dr. Bennett?
                  DR. STEVEN BENNETT:  No further comment.
                  DR. JAMES BLANDO:  Dr. Kissel?
                  DR. JOHN KISSEL:  I think I already said this, but I'll just reemphasize that the empirical data in the Frasch et al. experiment with respect to fluxes is not being utilized by EPA in this analysis and giving difference to the fraction absorbed rather than looking at the fluxes is a failure to utilize all the information that's actually in that study.
                  DR. JAMES BLANDO:  Okay.  I think that's it unless anybody else on the panel wants to comment.
                  DR. KENNETH PORTIER:  Any additional comments?  Yes, Dr. Davies.
                  DR. HOLLY DAVIES:  I wanted to add a few things.  Dr. Anderson had mentioned people living near daycares.  I wanted to add daycare -- living near dry cleaners.  Sorry.  And I wanted to add daycares onto living near dry cleaners.  Also, with the number of workers for the aerosol cleaners, in particular, it seems the aerosol degreasers -- because there's not big pieces of equipment to switch out, it seems like those workers, in particular, could be more workers because it's harder to estimate those.  They could switch products pretty easily.
                  And then I'm not sure where to put this comment, and so this seems like a good place to put it.  We're talking about estimating uses.  I like mass balance, and how much 1-BP is there?  And so I don't know where else to put this in the comments.  One of the public commenters, I think in the written comments, had said that the estimates that EPA's using through the public disclosures they do is overestimating it.  It's hard to know that.  Hopefully, they'd be able to give more information.  EPA can only estimate the information that it gets.  
                  I was looking at it from a different perspective.  There was a point where EPA said there's only five pounds, I think, going into the water, and I thought, well, that's hard to believe.  You know, TRI -- TRI is what TRI is.  It's not all of the releases in the country, but it would be nice to really see kind of where it's really going.  If we're looking at the releases, is there really one five pounds going into the water?  If 24 million pounds are being imported or manufactured, it's hard to believe that little's going out, but a better kind of mass balance, better estimate, and maybe not that much is being imported or manufactured.  
                  Maybe the commenter's right and only five pounds are going in.  And the list of uses on page 100 that are kind of niched, they aren't really estimated.  Some of those sounded like they could also be fairly large uses, but it would be nice to have better estimates on there.  And I think some of the other public commenters also commented on it.  In fact, Richard Denison's media had something about the toxics release inventory and how much was coming in from different places.  That could be added in.
                  Okay.  Yeah, I think Dr. Anderson had said a little bit about that too in his last comment that I wrote down.
                  DR. KENNETH PORTIER:  So you're kind of thinking like an accountant, right.  You had X amount millions; how do we separate them out?  What's leftover that's lost to follow on, right?
                  DR. HOLLY DAVIES:  Well, especially if we're looking at aggregate exposures to the whole population that we should be looking at where everyone is exposed, it would be nice to see where all of it is as we're looking at this.
                  DR. KENNETH PORTIER:  Dr. Blystone.
                  DR. SHERI BLYSTONE:  Sheri Blystone.  Just to kind of comment on that, I think there's uncertainties in all of that sort of data.  I have my own personal opinions of whether TRI data is even relevant for doing the types of things that they're doing because there's a lot of uncertainty in that and certainly overestimation.  
                  The CDR data swelled.  There's potential for double-counting or triple counting as that public commenter mentioned.  So some of the uncertainties around that, that maybe there's an opportunity to expand on that.  But there's definitely challenges for all of that data.
                  DR. KENNETH PORTIER:  Dr. Cobb and then Dr. -- well, no, I'm sorry.  Ms. Rudel was next.
                  MS. RUTHANN RUDEL:  Thank you.  So I'm thinking about what the document says to sort of what it conveys to a risk manager also to the public who might be interested in this and how the uncertainty in the assessment overall is conveyed, and we talked about this with PV29 as well.  As I was sort of reading it, on the one hand there's one place where they use, I guess, an informal communication or a letter from Eviro Tech to say that there's maybe 27 dry cleaners that are using 1-bromopropane, but then other sources of information would suggest that maybe, are there 20 thousand?  Or like how many dry cleaners are using this now?  And on the one hand, it looked like sales data show production in sales are increasing over time, and then in other kinds of streams of information would suggest it's being replaced because of the hazard.  
                  So I don't read this and feel like I have any real sense of what is happening with this chemical.  Where it's going?  How many facilities are using it?  Where it's ending up?  And so I'm echoing Dr. Davies' comment about -- and EPA does, I guess, according to the commenters, have the authority to get manufacturers to give sales data.  I think in addition to doing a better job of estimating sort of worker exposures and environmental releases that, well, then environmental releases will be important for characterizing exposure to the general population, which I know EPA has sort of excluded from this, but we'll come to, I guess, later in the discussion.  So I'm just recommending to avoid a misleading sort of representation of the precision in these assessments, also, given some of the uncertainties and also the uncertainties in the models that we were talking about.  
                  And then I had, in terms of the modeling, there's a statement that there's consistency between monitoring and modeling, but I think it would be helpful to see that more clearly presented what that was so that people can evaluate how much uncertainty or confidence that they have in that.  
                  I propose that the estimates of worker exposure are done assuming no PPE and no ventilation and assuming PPE and ventilation, that all of those results would be very clearly presented as separately so that some idea of the range of exposures and risks would be presented, and that the inhalation and dermal would be combined.  And there are different ways to represent uncertainty in the exposure estimates, but you could add another -- you could do more -- try to represent more of a worst-case and try to just be -- I don't know.  Like the slides that were shown this morning about the worker exposures and they don't say on them, this is assuming with PPE or this is without PPE.  But that should be sort of a big feature of any representation of this data of what the worker exposure is.  Like I would expect that to be on the figure.
                  DR. KENNETH PORTIER:  Dr. Cobb.
                  DR. GEORGE COBB:  So to a comment that was made earlier about double or triple counting on the TRI, based on the fact -- my estimation -- the report says that 15.4 million pounds of 1-BP were produced in 2012, yet we only have facility that's reporting discharge.  So, unless we get double or triple counted up to 15.4 million pounds, I'm personally not concerned about that.  As a matter of fact, I would say in the absence of data suggesting how much is released, perhaps the Agency should consider that all of the 15.4 million pounds is either released into air or water on an annual basis because it is used to spray or --
                  DR. KENNETH PORTIER:  -- it's not converted.
                  DR. GEORGE COBB:  -- so it's got to be disposed of somewhere.  So if it's not inventoried as being properly disposed, maybe that's the route that should be taken to be conservative.
                  DR. HOLLY DAVIES:  We were just saying some is exported.
                  DR. DAVID EASTMOND:  I believe quite a bit is exported.
                  DR. HOLLY DAVIES:  Hence the mass balance.
                  DR. DAVID EASTMOND:  I think that's one of the ones that we had public comments that were made.
                  DR. KENNETH PORTIER:  That was Dr. Davies and Dr. Eastmond.
                  DR. DAVID EASTMOND:  And exports would certainly fall into that category.
                  DR. SHERI BLYSTONE:  This is Dr. Blystone.  The double and triple counting wasn't around TRI data, it was around CDR data.
                  DR. KENNETH PORTIER:  I think we've kind of handled this question.  At this point, I'm going to turn to EPA -- and I need a break.  So I'm going to turn to EPA and say, do you have any clarifying questions or comments?  The panel will clearly write all this up.  Dr. Blando's got it already ready to go.
                  DR. STAN BARONE, JR.:  So I think we've heard some really constructive and robust discussion.  I do have some clarifying comments, and they cut across a number of the speakers' comments.  So I wanted to be sure that we're on the same page.  
                  Number one, there was several discussants mentioned an OSHA PEL.  There is no OSHA PEL for 1-BP.  There is a hazard alert which refers to the ACIGH value.  That is not a regulatory limit.  That is a suggestion.  So that's not a PEL.  
                  So let's just be clear about that.  Yes, we consider -- and the committee has also discussed about temporal issues with a PEL.  We'll be talking about that in the future.  We've received public comments about risks above and below the PEL -- not a PEL.  There's no PEL.  
                  So again, we can look at the ACIGH value, and what is current practices and what's happening, and where does the monitoring data fall and those kinds of things.  That actually can be derived from what we have in the assessment.  
                  There was mention of our exchange or information exchange with OSHA, NIOSH, and others and, in fact, we have a very active engagement in the federal family.  On their monitoring data, we actually incorporated monitoring data from OSHA, NIOSH, DOD, HSIA, and HSIA being an industry association that provided us monitoring data for some conditions of use.  And, again, that's incorporated into our assessment as well as publication data so monitoring data.  Again, as the presentation outlined, representativeness of those samples is always a question, and that's why we also look at modeling and what we said was looking at the modeling and looking at the monitoring data for at least six conditions of use where they overlap, there's pretty good agreement in those cases.  
                  There was a discussion about workers and workers in the broadest sense and PPE.  And I still think we have some confusion about what we're presenting as far as risk estimates.  We are presenting risk estimates along a continuum, and it's in our supplemental files actually.  You have all of that in the supplemental file and the risk calculator with and without PPE and with different types of PPE.  And, again, the modeling efforts really, really help us facilitate this process by looking at the different combinations where monitoring, we can't always do that.  So we do have that there, and we actually show in the summary tables with and without PPE, and we show the range of PPE that may reduce the risk estimates.  
                  But we also show -- and I think this gets to Dr. Anderson's point -- that not all workers are expected to use PPE and that the occupational nonusers -- again, the person or persons who are working within that far-field zone -- will not necessarily be expected to be wearing PPE, and we're presenting those risk estimates -- no PPE for those workers.  And you will also note that oftentimes the risk driver is the ONUs which has also received some comments.
                  I think that's pretty much all the clarifying points.  I wanted to make sure that -- there was a lot of good discussion -- I wanted to make sure we didn't lose some threads or overgeneralize what's in and what's out.  I think we, again, are trying harder to hyperlink our document.  In hyperlink, we've figured out a couple of tricks -- the hyperlink to the supplemental files too so to help you and the public be able to find the information because it's a lot of information to present.  And we're trying to get the summary information in the assessment file, but, yet, provide transparency to where the calculations are done and where the more detailed information exists in the appendixes and in the larger supplements.  And the exposure supplements and the risk supplemental files are debts.  We acknowledge that, but again, we're trying to show all the data we got including the monitoring data that we have received which has been sanitized, and we also recognize that it is difficult oftentimes for the public to understand what's in those inspection databases.
                  DR. KENNETH PORTIER:  Okay.  We're due a break.  I have 3:25.
                  DR. HENRY ANDERSON:  Just a quick question.
                  DR. KENNETH PORTIER:  Dr. Anderson.
                  DR. HENRY ANDERSON:  In all the data, is there a table somewhere that summarizes kind of by count how many monitoring samples you get from air, the various sources?
                  DR. STAN BARONE, JR.:  Dr. Anderson, you are asking for a number of samples per site or what are you actually asking?
                  DR. HENRY ANDERSON:  Just a total number.  I mean, are the number of samples or the data that you have to work with in the tens of thousands?  I think they'd be helpful.
                  DR. STAN BARONE, JR.:  We can get --
                  DR. HENRY ANDERSON:  I just don't have a sense of how much data you have.  Mention it.
                  DR. STAN BARONE, JR.:  Yeah, we can pull that out for you.  But it's not tens of thousands at all.  It's probably up to close to a thousand.  Again, I'm just guessing.  This is one of the cases where we have the most monitoring data points of any of our assessments.
                  DR. HENRY ANDERSON:  That's what I thought.
                  DR. STAN BARONE, JR.:  And we're still probably talking less than a thousand.  Again, when you look at monitoring studies, you're usually -- Dr. Blando knows.  If you had a hundred samples, you'd be singing a happy song.  I'm being a little colloquial and down in the country, but it is not unusual to have small samples when you're talking about monitoring studies.
                  DR. KENNETH PORTIER:  Let's take a break.
                  DR. HENRY ANDERSON:  Absolutely.  Oh, I just think it's hard, I know, from the public to get a sense of how data rich is the database you have to work with.
                  DR. KENNETH PORTIER:  Thank you, Dr. Anderson.  I'm dying for a cup of coffee here, so we're going to take a break.
                  DR. HENRY ANDERSON:  Oh, go take a break.
                  DR. KENNETH PORTIER:  3:45 we'll reconvene, and we can continue with your comments then, Dr. Anderson.
                  DR. HENRY ANDERSON:  That's okay.  I'm done.
                  MS. TAMUE GIBSON:  Committee members, we do have tea and coffee in our administrative breakout room from this morning.
                  
                                    (BREAK)
                  
                               CHARGE QUESTION 3 
                  
                  MS. TAMUE L. GIBSON:  All right.  We are ready to begin, and I'll turn it over to our Chair, Dr. Portier.
                  DR. KENNETH PORTIER:  I've asked EPA to read Charge Question 3 on consumer exposure assessment to kind of summarize, if they could, the four paragraphs of lead-in information.  If the panel can read that. 
                  DR. KATHERINE ANITOLE:  Okay, Dr. Portier.  I'm a little caught off-guard.  I haven't summarized these four paragraphs.  I'm just going to read the first sentence of each one and then I'll get into the questions themselves.
                  Question 3:  Consumer Exposure Assessment.  Section 2.3.2 of the Draft Risk Evaluation.  EPA evaluated acute inhalation exposure to consumers for nine product use scenarios.  
                  EPA evaluated exposure to consumers in residential settings following acute exposures and considered both users of a product and bystanders within the residence where the product was used.
                  Three models were used to evaluate acute inhalation exposure depending on the condition of use evaluated.  
                  Product-specific consumer monitoring information was not identified during the systematic review process; therefore, model inputs related to consumer use patterns (duration of use, mass of product used, room of use, and similar inputs) are based on survey data found in literature.
                  Question 3.1:  Please comment on the approaches, models, exposure or use information (for example, information on duration, number of user events, amount used) and estimates for the nine consumer uses evaluated in this draft risk evaluation.
                  Question 3.2:  Please provide any specific suggestions or recommendations for alternative approaches, models, exposure or use information (for example, information on duration, number of user events, amount used) that could be considered by EPA in developing and/or refining the exposure assumptions and estimates for the nine consumer uses evaluated for this draft risk evaluation.
                  And Question 3.3:  Dermal exposure was evaluated using a permeability method within CEM based on the availability of a permeability coefficient found within the literature and a study by NIOSH 2017.  The permeability method within CEM does not consider evaporation when estimating exposure, which is the primary basis for EPA evaluating dermal exposure, only for consumer uses where there is a continuous supply of product against the skin, during use, or a barrier prohibiting evaporation.  Please comment on the chosen approach and provide any suggestions or recommendations for alternative approaches, dermal methods, models or other information which may guide EPA in developing and refining the dermal exposure estimates.
                  DR. KENNETH PORTIER:  Thank you.  And we've asked Dr. Bennett to lead this discussion. 
                  DR. STEVEN BENNETT:  Thank you.  I think I will introduce a few concepts, then let the others take over.  I didn't have an opportunity to gather all the questions from the other panelists.
                  I think one thing I'll make a comment, and this will probably surprise others.  I think was a pretty good overall piece.  And I think one of the things, as I was going through a couple areas, I addressed -- a couple concerns I raised where I was thinking about as going through this, as I raised those issues, when I went to the end, to the discussion of the uncertainty, et cetera, they had addressed those questions.  So I want to compliment the Agency on that part.  So when there's deficiencies, if you appropriately address it, the panel usually -- it helps us out a lot.
                  A couple areas similar to what I raised earlier with respect to the consumer and the industrial/commercial uses, probably a good chance to reevaluate the uses.  I think in many cases, the uses that have been identified in some case were based off the earlier risk evaluation, and that's in 2017.  
                  I did some spot checking on a few of the products that were listed in there, and I was not able to find them on the manufacturer's website now.  Some were broken links, some were searching.  But I'm not sure if that's fully accurate.  I would suggest the Agency may want to double-check those to make sure that those are still in the marketplace.  Others were definitely there, as well.  But I do want to make sure that is fully reflected.
                  Another issue that I articulated last time with utilization of the Westat Study Survey.  I still have concerns with that, dealing with that is 30-year-old data.  But I also recognize that is one of the best -- by far the best survey that is out there.  Certainly great for suggestion just broadly, figure out how to get that updated to some of today's use and practices.  I think that's both an encouragement to the Agency, but also to the recognized impacted industries, et cetera, like ourselves.  So it's certainly an area that should be looked into.
                  One area that I thought -- the incorporation of the inhalation within the consumer space, I was a little puzzled by why the vapor portion of the inhalation was incorporated into the consumer.  I would sort of think that's more applicable to a general population exposure, rather than trying to trace it down back to an individual, the consumer space.  
                  So I think it might be useful, a better explanation as to why that was incorporated under the consumer space.  Granted, the general population is excluded from the risk evaluation or is put elsewhere.  But I think that there should be a better explanation of that piece, of why that approach was taken.
                  One area within one of the identifications.  I think it was the electronic degreaser and the specialized electronics utilization for cleaning program, VCR versus today.  I'm not sure that's fully reflective with respect to consumer electronics today.  You certainly can clean many consumer electronics now with water or other solvents rather than a typical chlorinated or halogenated degreaser.  So I think that might not be fully descriptive of that particular use.  I think that's a piece.
                  And I wanted to raise a couple other questions of looking at the population exposed and susceptible subpopulations.  I'm not sure if that was adequately addressed within the consumer piece.  So I think that's an area that may want to be looked at and make sure that's fully put in there as well.  And that's all I have for right now.  I'll open up to the next panelist.
                  DR. KENNETH PORTIER:  Dr. Barton.
                  DR. CHARLES BARTON:  Thank you.  If the modeling assumptions are made that provide inputs to the exposures and a justification for choosing these would assist in the reader understanding as well as transparency, I'd encourage this.  And this also goes to Charge Question 2 as well.  
                  Another issue, a question that I had, was regarding the permeability that was used in the consumer exposure model.  I think it would be good if it was expressed why this was chosen, considering that vaporizing isn't taken into consideration in that, whereas the fraction that is absorbed was looked at in the worker exposure.  So why the differences in models that were used, at least mention it and why it was chosen.  Because right now I have to assume the reasons why.  And that's all that I have.  Thank you.
                  DR. KENNETH PORTIER:  Dr. Davies.
                  DR. HOLLY DAVIES:  Thank you.  I'm not a big model person, but the models and inputs seemed reasonable.  And I think we talked earlier about the size of the houses and such, and of course houses are all different.  
                  I can't remember the phrase you used in there about them being the normal inputs or the reasonable inputs, or whatever it was, it was a good phrase.  Those are the inputs we all use.  Those are the assumptions that are commonly used, or something like that.  It was a good assumption, that is what we use for housing size and we have to use those assumptions of American house size.
                  I had missed that children could be bystanders.  I'd actually put down originally that children should be bystanders.  And Katherine in her presentation this morning said that they could be.  So I looked and I saw on Page 112 just a small line about bystanders could be any age, so I was really glad that you pointed that out this morning and that I saw that.  Because in all the other explanations, I hadn't seen that. 
                  I looked for data and couldn't find it in the places that I normally look for data on chemicals in products.  I looked in the Danish EPA that has data and I looked for state data and I couldn't find anything else.  So again -- and the Westat study is old but does seem like it's a good study.  And I couldn't find any other data.
                  Not surprisingly, I looked up the criteria for the data quality evaluation for this kind of study, and there is no criteria for age.  So it was not ranked for being too old.  Some of the other types of studies have criteria for age. 
                  And I also liked her explanation as to why it could still be relevant, but also reasons why to take it with a grain of salt.  So I think that was my comments. 
                  DR. KENNETH PORTIER:  Dr. Kissel.
                  DR. JOHN KISSEL:  Okay.  For 3.1, my primary comment is that I couldn't really figure out what the approach was for the dermal -- I asked this qualifying question earlier, but I couldn't follow the calculation.  So that's my primary comment, that whatever is being done isn't transparent.
                  Are we doing these together or are we going one at a time on this? 
                  DR. KENNETH PORTIER:  Do all three together. 
                  DR. JOHN KISSEL:  So with respect to -- and this question keeps coming up -- permeability approach versus fraction absorbed, you know, ultimately you're trying to make sense of the data and you should be running both approaches and getting to the same place.  
                  With respect to what other models might be used, for 1,4-Dioxane you were using the Kasting and Miller/ Frasch version, and there's no reason why you couldn't be doing that here also.  I didn't like the way you interpreted it last time, but the modeling -- this is a volatile compound and you've got competition between uptake and volatilization, and that model was intended to deal with that situation.  And ultimately you should be able to take those models and predict Frasch's low fraction absorbed, if the world actually makes sense, if we actually understand what we're doing.  
                  So I'm a little put off by the either/or approach to modeling that I'm hearing from both EPA and the other panelists.  Because I think ultimately the goal is to get the physics right and then all those models will collapse together and be the same thing. 
                  With respect to -- and I think I kind of said this already, but the permeability coefficient from NIOSH is okay as far as it goes, except that it's contradicted by the Frasch data.  And once again that's getting back to this big picture.  If you only pick one approach and it's contradicted by another approach, one of them is off the mark and the interest ought to be in making sense of this issue.  And we have more coming down the pike.  NMP and methylene chloride and TCE are similar kinds of compounds that these same issues are going to come up, and at some point I hope EPA evolves to making a coherent whole of all of those compounds.  So you can use the same approach for every one of them and produce a result that seems to make sense.
                  DR. HOLLY DAVIES:  Can I have one thing I forgot to say?
                  DR. KENNETH PORTIER:  Yes.  Identify yourself. 
                  DR. HOLLY DAVIES:  This is Holly Davies, and I said this to John Kissel earlier, and I should say it for everybody.  Because I don't know the dermal models -- and I had asked John this, because it doesn't make sense to me why they couldn't use the same dermal exposure assessment for the consumer as they use for occupational.  But there's a sentence in there about we couldn't use the same ones, so we're going to use different ones.  And that was just a question that I had, because they're all people. 
                  DR. KENNETH PORTIER:  Dr. Bennett.
                  DR. STEVEN BENNETT:  I think I'll add one recommendation.  There was an earlier question, a comment asking about the occlusion -- the occluded uses, recommend that an expanded discussion of that or incorporation of uncertainty associated with use of the gloves or other clothing et cetera that may occlude the solvents, or it may permeate the membrane and then be stuck to the user's skin.  That's certainly a discussion that probably should be looked at, both within the consumer space and possibly the occupational.
                  DR. KENNETH PORTIER:  Dr. Anderson, Dr. Cory-Slechta, Dr. Pessah.  Do any of you have comments to add?
                  DR. DEBORAH CORY-SLECHTA:  I don't. 
                  DR. KENNETH PORTIER:  I'm glad to hear you've joined us, Dr. Slechta.
                  DR. DEBORAH CORY-SLECHTA:  I've been sitting here for a long time. 
                  DR. HENRY ANDERSON:  My only comment, and it's probably a bit off the mark, but that has to do again with Holly's talking about mass balance.  I really don't know how much product is in these, but for having used spray cleaners and spot cleaners and degreasers on my motorcycle and things like that, pretty much it's gone when you're done.  
                  And the question is where does it go and is it an environmental problem.  And that's partially on what's the volume of the material that's actually being distributed in small containers and used indoors.  And I suspect when it's kind of left over there's a bunch of gunk with it that's poured out on the ground in the backyard and ends up in -- or worse yet down the sewer and then it's out in the lake as a sheen.  So I'm not sure what other exposures remain if the work is done indoors.
                  DR. KENNETH PORTIER:  Dr. Kissel.
                  DR. JOHN KISSEL:  I just wanted to add that Holly's comment and Steve's comment were basically the same thing.  If you read 3.3, it's basically what EPA has done is assumed that these are -- that the consumers are having occluded exposures, that they have their finger stuck against a coin or something and it can't evaporate so it goes into the skin, however that long lasts.  You're doing permeability into the skin for however long that lasts and you don't worry about evaporation.  So, again, they're subsets of the same model.
                  DR. KENNETH PORTIER:  Dr. Doucette.
                  DR. WILLIAM DOUCETTE:  I didn't read this section as carefully as the ones I was responsible for, but I did have a question that either the panel members or maybe the folks from EPA can answer.  I didn't see it in the first couple of paragraphs, but I got involved in a study we were looking at vapor intrusion of chlorinated solvents and we were finding homes that had elevated concentrations that shouldn't have happened based on where they thought the plume was.  
                  So we went in and did room-to-room measurements and looked for consumer products that actually were omitting these chlorinated solvents.  And what we found is we were looking primarily at chlorinated solvents, not the brominated.  But basically the spray cans that are used for degreasers or spot cleaners.  
                  And what we found was that we measured the rate that would come out during use and then we also looked at what happened during storage.  And they were a continuous source.  And the spray cans that were nearly 100 percent of PCE, for example, were an unlimited source for as long as the can actually contained that substance.  And we found that in rooms, that actually exceeded the cleanup standards that the Air Force had for indoor air, just coming from that spray can.
                  So my question was do you consider -- just once you've stored that product, and it's been used, it continues to release constantly.  And is that something that you looked at?  Hopefully I made that clear.
                  DR. KENNETH PORTIER:  Dr. Barone.
                  DR. STANLEY BARONE, JR.:  We are looking at degreasers, spray degreasers.  I'm not sure I'm aware of that particular data.
                  DR. WILLIAM DOUCETTE:  You're using only during the use.
                  DR. STANLEY BARONE, JR.  Right.  I'm not aware that we have that kind of data.   
                  DR. WILLIAM DOUCETTE:  And there are some flux data -- not for this particular compound, but I could send you a couple that where actually -- that measured the flux.
                  DR. STANLEY BARONE, JR.  That would be very helpful.  Again, it changes some of our calculations if we actually had that kind of source data.
                  DR. WILLIAM DOUCETTE:  Constant source.  Thank you.
                  DR. KENNETH PORTIER:  Dr. Kissel.
                  DR. JOHN KISSEL:  I would just like to say that the TEAM study demonstrated that in the early 1980s; that the solvent concentration inside homes in Bayonne, New Jersey was enormously higher than the outdoor air in a refinery town because of consumer products. 
                  DR. KENNETH PORTIER:  Any additional comments?  Dr. Bennett, I would say when I read this section and I went back into my ceramics hobby area, and I found a bottle of mold release with            1-Bromopropane in it.  It's been there for at least 15 years that I know of, because I inherited it from my father-in-law.  And now I'm not going to use it.  I don't know what to do with that.
                  DR. SHERI BLYSTONE:  It's releasing anyway.
                  DR. WILLIAM DOUCETTE:  You've been exposed for the last 15 years.  
                  DR. KENNETH PORTIER:  It's still got juice, I think.  I'm not about to test it.
                  DR. HENRY ANDERSON:  Take it to a paint collector.
                  DR. KENNETH PORTIER:  Ms. Rudel, you wanted to comment. 
                  MS. RUTHANN RUDEL:  Just really quick.  There's one sentence about insulation, where it says we used a certain brand of Polyiso board, I guess as the basis for the model.  And I had been looking to try to get through the supplemental and see if there's more information.  But I was curious, like, why you used that, what information you had about the off-gassing of 1-Bromopropane from that particular product and whether -- that seemed like a board that's already -- it's already a hard piece of insulation when you install it, but is 1-Bromopropane used in the spray foam stuff that's installed on site that's blown in or whatever? 
                  DR. KENNETH PORTIER:  Now we can turn to EPA.  Any comments or clarifications? 
                  DR. STANLEY BARONE, JR.  One of the things you'll see in our assessments, and the puns intended, is we recycle our models.  Where we have developed a model for spray foam insulation for a semi valve tool, a flame retardant.  We'll basically build upon that model with additional modules or additional information.  
                  A lot of the times, the model starts out for a particular chemical purpose, but gets expanded for other chemicals -- related chemicals, and then other chemical classes.
                  So, again, it's about validation and review of the domain, and the P-chem properties, of course, are driving a lot of the issues of is this covalently bound, is this emissive, or not.  So there's a lot of issues there, but I'll just say that we're trying to use these models to the maximum extent possible.  Questions? 
                  MS. RUTHANN RUDEL:  I was just curious about my specific question though.  Is 1-Bromopropane used in spray foam insulation that's installed -- that's blown into people's houses?  It's like installed on site rather than put in as a board. 
                  MR KEVIN VUILLEUMIER:  This is Kevin Vuilleumier again.  For the product that we looked at, it's actually a board.  It's a predesigned board.  We don't have information, based on our product review actually, to determine whether or not that showed that was in the spray -- I think you're talking about the spray foams that are just a bottle that sprays out and then expands.  
                  That wasn't in the consumer uses that I know of, that we tracked down for products.  But we aren't certain -- I don't know an answer if we do or do not have it in some of those products, but we didn't any in the consumer uses that we did identify, that I know of so far.
                  So it's in the hardboard, which is what we analyzed.  And then for spray foam, we can certainly look back in the products and see if there are new products or products that we missed, that is a spray foam as you're referring to, that expands, and see if that contains 1-BP.  But for right now, the conditions of use that we looked at, no products were identified as a spray foam that contained that. 
                  DR. KENNETH PORTIER:  So a crack and crevice foam would be a consumer item.  But what about an occupational?  Because the other one she's talking about is like a two-part spray foam, where the guy's in the full spacesuit and he's in your attic spraying this stuff all over the place.
                  DR. STANLEY BARONE, JR.  We actually have a model for that, for isocyanates, again for the expansive foam.  So, again, to Kevin's point, we haven't yet identified that in those spray foam products.  It's just in hardboard products right now.  Again, it would be an additional effort if we found that, but we have models to actually do that. 
                  DR. KENNETH PORTIER:  Dr. Cobb.
                  DR. GEORGE COBB, III:  I think some of the confusion is probably arising from I think page 121 where you talk about a model for emissions from spray polyurethane foam, and whether that's -- I think that's where the question comes.  And I actually had that question, too, so -- 
                  DR. STANLEY BARONE, JR.  I think that actually gets back to the issue that I was trying to raise.  We're using models that were developed for one particular purpose, and using them for additional functional uses, related but different. 
                  I have a couple other clarifying comments, if I may, Chair.  To Dr. Kissel's question earlier about which module within the model we were using -- and I believe this was the MCM model.  It's Peter 2B, not Peter 1B.  There was a typo for that module, and that's why you weren't following the proper breadcrumbs.  We had a typo.
                  For, I think, Dr. Johnson's question about the high and low intensity issue, the medium intensity is correct, but the high and low intensity values got switched.  And it's correct, I think, in the supplemental file.  The numbers are correct in the supplemental file, but in the text of the document they got flipped. 
                  Also the earlier question Dr. Voorhees asked, and then there was follow-up about the housing stock.  Housing stock and the size of houses has changed over time.  We have updated information from the Exposure Factors Handbook.  That Exposure Factors Handbook, again, comes up with -- from the National Building Housing Survey, the average size of rooms, how many rooms in a house, whether they have basements or don't have basements.  
                  So there is, for our model inputs, it does come from a national survey.  And again, as you may or may not be aware, when you do a national survey you have to adhere to rules and guidance about national surveys and information collection rule.  And it's not a trivial exercise to recapitulate a national survey, particularly if you have lots and lots of respondents, potential respondents.  That requires public comment rulemaking and OMB review. 
                  DR. KENNETH PORTIER:  Ms. Rudel.
                  MS. RUTHANN RUDEL:  Ruthann Rudel.  So I guess I have two really specific suggestions for the report, on the subject of home insulation.  And if you've done this in the supplemental or I've missed it, then I'm sorry.  But I guess I still don't know whether 1-Bromopropane is in insulation or not.  And I heard you say we didn't identify that it was, except for these boards.  
                  I think it would be -- it's useful, again, for risk managers or the public to know what you don't know, I suppose, which is we don't know that it isn't, I guess is what you're saying.  And to just be clear, like here's all these different kinds of spray foam insulation that people might encounter.
                  One of them is a company that comes in the back of their truck and they blow stuff into your attic and they're wearing hazmat suits.  And we don't know, I guess, if there's 1-Bromopropane in your house after they do that or not.  
                  And then there's a can of stuff that you can buy and do your crack and crevice treatment.  And we don't know if 1-Bromopropane is in that.  And then there's insulation boards. 
                  Then my second suggestion is to do an exposure scenario where the boards are put in a space that someone is living in, like if they have a basement apartment or whatever it is, rather than that it's in a crawlspace or attic where nobody is living.
                  DR. STANLEY BARONE, JR.  This actually gets to Dr. Bennett's question or comment earlier also.  We will be, post peer review, updating our analysis.  We'll be looking at new TRI data as the next cycle comes in.  We'll be making sure that we've captured, to the best of our ability, the updates of the product uses, and market uses, and that gets included.  And especially in context of where we have unreasonable risk findings.  So there will be additional review.
                  To your point, it really gets at we've identified to date what's in the products, what's in the market.  The market changes, and I'll use the example in the commercial sector with dry cleaning.  When we started the assessment, the number of dry cleaners were a certain number.  
                  We found out from the producer/manufacturers of 1-Bromopropane that it was licensed to dry cleaners of a specific type of machine.  And so the original estimates of dry cleaners from Dr. Blando and others was way smaller than the total number of dry cleaners in this country.  It turned out to be less than 20 dry cleaners in the entire country were using 1-BP.
                  So, again, today we may have no great stuff spray foam with 1-BP.  It may not occur in the current product formulation.  That does not mean tomorrow it couldn't replace another solvent.  So this was brought up at our previous peer review, is there are always changes in formulations in the market; and, depending upon selection of what's the best propellant, what's the best for whatever purposes, things change. 
                  So we are trying to anticipate.  We're looking at the uses.  We're looking across our assessments at these conditions of use and doing the best predictions that we possibly can when we formulate the conditions of use, which we took comment on previously -- and again, a year ago to now, things could change.  But we're still trying to anticipate all that in our assessments. 
                  So we appreciate your comments.  And if you have any data, we really would appreciate data that would help inform or help expand that particular analysis for that condition of use.
                  DR. KENNETH PORTIER:  Dr. Gilbert.
                  DR. KATHLEEN GILBERT:  I was interested in the exact same question, as far as 1-BP in spray insulation.  And I just Googled it and I found several that you can buy commercially that have 1-BP in it.
                  Then I also looked at the insulation, the boring insulation, which is made by -- Dow is one of the biggest proponents of that.  And they very proudly say on their website that THERMAX, which is their version of Polyiso, is used in like 70 percent of commercial and residential housing now, for wall and attic insulation.  So it sounds more extensive.
                  You can find companies that will sell you insulated packaging for shipping that are lined with Polyiso.  So, to me, the insulation issue seems to be sort of underrepresented in this document. 
                  DR. KENNETH PORTIER:  Good point to note.  
                  Mr. Kaufman.
                  MR. ALAN KAUFMAN:  Thank you.  I was just going to make a comment on the blow-in insulation, that the ones that I'm familiar with are not polyurethanes.  It used to be cellulose, blow-in cellulose, and that's fallen out of favor because obviously it's flammable.  Most of them are glass fibers now.  And the reason for the hazmat suits is because it's a mechanical irritant.  You don't want to be breathing it or get it on your skin.  But I wouldn't expect it to contain 1-BP at all.
                  DR. SHERI BLYSTONE:  Excuse me, this is Sheri Blystone.  Just to comment on that, polyurethanes are absolutely used in spray foam insulation.  The blowing agent I don't know is       1-BP.  I think what we've heard there is no evidence that that is the blowing agent.  I worked for a company that did that.  I don't recall us using      1-Bromopropane.  So I think it's fairly -- but it is absolutely a use, spray foam.
                  DR. KENNETH PORTIER:  Dr. Bennett, you got all this? 
                  DR. STEVEN BENNETT:  I hope so.
                  DR. KENNETH PORTIER:  Good.  Any additional comments from the panel, from panelists on the phone?  I'm ready to move on to Question 4.  Not hearing any, I will bring up Question 4 and let's move forward. 
                  
                               CHARGE QUESTION 4 
                  
                  DR. KATHERINE ANITOLE:  Question 4, Environmental Hazard and Risk Characterization.  Section 3.1 and 4.1 of the Draft Risk Characterization.  Reasonably available information indicates that 1-BP exhibits a moderate environmental hazard to aquatic species.  A screening level analysis of potential risk to aquatic species indicates that expected environmental concentrations are below hazard thresholds for aquatic species.  In addition, a qualitative consideration of physical chemical properties and the conditions of use in this draft risk evaluation indicate that risks to sediment-dwelling invertebrate species and terrestrial species are not expected.
                  Question 4.1:  Only a few environmental test data endpoints, including ECHA, are available in the public domain for 1-BP.  Most are from the ECHA website.  EPA attempted to obtain the full ECHA studies with no success.  Since the studies were in French and Japanese, and no U.S. sponsor, EPA decided not to make further attempts to find the studies.
                  Given that the ECHA environmental test data results are in the public domain, EPA decided to use the environmental data.  Please comment on the reasonableness of this approach for the environmental hazard assessment of 1-BP.
                  Question 4.2:  EPA determined that there are no environmental risks based on a screening level assessment of risks using environmental hazard data, TRI exposure data, fate information and physical chemical properties.  Please comment on the approach used in the screening level assessment.  Are there other data that EPA could have considered?  If so, please provide specific data and references.
                  DR. KENNETH PORTIER:  Okay.  Dr. Doucette, before we jump into this question, I just wanted to let the panel know we're not going to move on to Question 5 if this question -- which I think is going to get some pretty good discussion -- goes short.  Sometimes the panel is ready and gung-ho, but I don't think we're ready to jump into the human head.  We want to be fresh on that. 
                  So we're going to take the environmental hazard and risk characterization and we're going to work it very well, and then that's going to be the product for today.  
                  Dr. Doucette.
                  DR. WILLIAM DOUCETTE:  This is Bill Doucette.  So what I've tried to do is just kind of summarize things in bullets, to kind of facilitate the discussion and I'll provide more detailed comments in the report.  
                  First of all, I'd just like to thank the folks from EPA.  I thought this was a much more refined document than the PV29 that I was first involved in, and I really appreciate that.  I also appreciated the uncertainties and limitations slides each time, because I think that helps me put things in perspective also, and it was fairly well documented. 
                  I'm still having a little bit of difficulty understanding why spills to subsurface aren't considered a potential release to the environment when stormwater releases to surface waters were considered.  And in talking to my colleague, who's been involved in a lot of dry cleaning facilities -- granted, there's not as many as there are for the chlorinated solvents -- he was aware of leaks and again this issue of vapor intrusion into homes near the mom and pop dry cleaning facilities, and the whole idea of soil vapor movement and things like that.  I'm still struggling with why we don't consider that as a potential source, even though I understand that regulatory we're going to deal with it as a toxic air release.
                  Not to make a mountain out of a molehill, but I was involved in a TCE contaminated site where we actually looked at the risk of TCE vapors to burrowing animals as an ecological risk point.
                  So, again, I'm struggling with that, but I'm not going to belabor that anymore, other than I still think there is that potential environmental route of exposure that might be considered. 
                  Also the idea of looking at more data rich solvents like TCE, PCE, benzene, that are very similar in chemical properties, maybe not in terms of persistence; but to help put things in perspective I think would be nice.  And as a matter of fact, I kind of wish we would have looked at several of those solvents together if we're going to do another packet of compounds.
                  I mentioned the consumer products as a continual source of material into the indoor environment, not necessarily related to the environment that we're supposed to be talking about.  But if that's a source, it's eventually going to get out.
                  These are general comments.  Avoid -- I think -- who brought this up as far as the low, medium and high solubility sorts of things?  I think it was over there.  And it was much better than last time, because there was also inconsistencies between the physical, chemical properties.  And it's always difficult to describe something as high, medium or low unless you actually put that value behind it, and maybe not even say high, medium or low.  Just say this is the solubility, and then you get rid of all those sorts of potential inconsistencies.  Just a thought. 
                  When you talk about something being highly volatile, I go over this with my student in my environmental health or fate courses, it depends on the phase.  Vapor pressure is a great descriptor if it's spilled on an inert surface or in a beaker.  If it's not so good if it's spilled onto soil or activated carbon and all those sorts of examples.  So it's not so much that I worry about the overall assessment, is that at least the terminology should be a little bit more precise all the way through the documents.  It just makes things -- you're less open to criticism if you do that. 
                  So, again, volatility depends on the phase, and it also depends on the environmental conditions.  And all the parameters that are given in terms of physical chemical properties, or almost all of them except the persistent properties, are equilibrium distribution, and they say nothing about rate.  So there are times where we refer to volatility when we're really talking about decision between two phases.  And I think it's just important to be correct. 
                  There were also several assumptions talking about zero concentrations in a phase.  There's never zero concentrations in a phase.  It may be small.  But even if we look at the last go-around with this that you did in 2016, the Fugacity Level III model was presented, and there are measurable percentages in soil and sediment depending on the release scenario.
                  So, to say that there is zero in a particular phase, maybe from a practical sense that might be true; but really it depends on the inputs relative to the outputs.  
                  So, again, it's more of a terminology sort of thing than anything else.  And partition coefficients is still equilibrium.  And you can't use an equilibrium partition coefficient to describe whether or not something is going to be there in no concentration, because it's always got to be there if you're assuming equilibrium.
                  The other thing that -- again, related -- and these are very generic, and I'll have some specific ones, and I'll try not to take up too much time.  But there was a statement made that we're not going to consider soil an issue to terrestrial organisms because it has a low Koc.  And I think that's a reasonable assumption to make, but you need to keep in mind that even a Koc of 40, which is relatively low in the grand scheme of things, means there's 40 times more of the compound in the organic matter associated with that soil than there is in the water.  And so, again, it comes back to properly using the terminology of equilibrium partition coefficients with respect to what I'm trying to say.
                  There was also a statement in there about dry soil sorption.  If the soil is dry, the assumption was -- since this is -- if I remember correctly, the way it was quoted, the idea was it's reasonably soluble in water, and if the soil is dry, it's not going to sorb to the soil because there's no water there.  
                  Well, that's exactly the opposite for most volatile solvents, is that there's actually more sorption to dry soil than there is wet soil.  So there's just some things like that, and I'll provide references, that I think just the accuracy could be a little bit better.
                  As far as the things specifically to 4.1, the environmental test data endpoints we were supposed to look at, there was really only one study that was highly evaluated, and that was the fish study -- which was interesting because I actually worked at that lab for Larry Brook and Dan Kolub (phonetic).  I guess it was probably five years before they actually did this study.  So I'm familiar with that group and they do wonderful work.  But I also remembered -- so I decided to go back and look at that.  
                  So this was the key study.  Everything else was really extrapolated from that.  There were ECHA studies that talked about toxicity to daphnia and other organisms and algae, and I talked to Dan a little bit over this.  But this was the key study.  And the reason that I could justify using the ECHA data was because it all seemed to intercorrelate well with this result.  
                  So then I wanted to go back and look at that fish study.  And I'm not a fish toxicologist, I'm a chemist.  But it was interesting.  The first thing I did, of course, as a chemist is you look at the analytical data and how they did that, and would I consider that a key study or a study of high quality when I looked at that.  And the only information they provided within that document, that long document with multiple chemicals in there, is either they used gas chromatography by direct injection or liquid extraction, or they used HPLC.  
                  For 1-Bromopropane they should be using head space -- you know, there's a lot of techniques.  So that makes me doubt a little bit the quality -- even though I was involved in some of the earlier analysis, it makes me doubt the quality of the exposure concentrations for that particular fish study.
                  Now, granted, chances are that these concentrations that they measured then were probably actually higher to the organism than what they reported.  So that may be an okay thing in terms of being conservative or not.  But the criteria that was used to qualify that as a high-quality study, as an analytical chemist, I wouldn't view it that way.  
                  So it's just something to -- again, that you've got to work with the data that you have.  But that stuck out to me.  And I don't know, hopefully some of the other panel members will talk about the quality of the study.  But everything was based really -- that was the grounding piece of information that allowed me to use the ECHA data.  I don't know how everybody's going to feel about that.  Hopefully we'll get some feedback.
                  I do agree that there's relatively low persistence in the environment based on -- although it's not readily biodegradable, there is evidence of biodegradation aerobically.  It is one of those compounds that should undergo hydrolysis and photolysis once it reaches the atmosphere.  
                  There was a table that was given in the summary looking at what was -- in terms of physical chemical properties, what was considered high quality data and what was considered low quality data.  And I found it interesting that all the EPI Suite estimates were considered high quality, and there was one -- I think it was on the hydrolysis, which I think was 26 days, that referenced a classic publication, where most of the QSARs were developed for hydrolysis that was given low quality, and I wasn't sure exactly why.
                  I favor estimated values over what I think was an experimental value.  And there are reasons for that.  It would have been nice to have a little bit more information on that.  Then also within the EPI Suite estimation routines, there are multiple methods usually within all those properties.  And it would have been easier for me if that would have been specified, whether I used an octanol water partition coefficient to estimate Koc or I used the molecular connectivity indice to estimate that, just so I know.  And that gives me also a better feel for where I think the air associated with that is going to be.  And those were considered key and high properties for that.  So it's very important to be transparent in how that was done, at least in my opinion. 
                  The other thing that came up from several members, including our chairman, was the assessment factors.  Again, hopefully somebody else will address this that's a toxicologist, but the assessment or uncertainty factors, whether one, five, ten.  I found a couple papers that for pharmaceuticals, obviously not related, that used 100.  So I think for the most part it was fairly well documented, but I missed it in a couple places.  It would be nice if I got hit over the head with that a little bit better.
                  The TRI data, we'll have some discussions whether that's conservative or realistic.  It's just something to think of.  
                  And then let me see.  I've got to scroll down to the next one.  So that was kind of addressing 4.1.  Again, the fish data and the other data, hopefully other panel members will discuss.  
                  In terms of 4.2, there really isn't any measured environmental data to speak of.  Several of the key properties that were used to estimate fate were obtained from book chapters, and not the original reference.  Merck index was used for most properties, not what I would consider key references.  
                  I'd come into this after the PV29 thinking ah, there's going to be tons of environmental data for this particular compound.  And there doesn't seem to be.  Most of the key properties were actually estimated.  It's that minimum data set question to feel good about it.  But anyways, that's just something I noticed.
                  There were a couple of cases, for example, the biodegradation data, and the hydrolysis data, and even the photolysis data, where data was pulled from a reference.  Since I used EPI Suite for some of the key properties, it would have been nice to actually have both the experimental data and the EPI Suite data side by side to see how well they corresponded.  Because there are -- particularly for hydrolysis and biodegradation, that would have been nice to put together.  It would have helped me feel a little bit better about the experimental data, too.
                  Again, I brought this up, but the idea of analog comparison.  Again, I don't know that -- well, I know that TCE and PCE and benzene are not the same as 1-Bromopropane, but they are very similar in physical chemical properties, and they're a lot more data rich.  And it might help just to kind of put things in perspective for the -- certainly that's what I do when I look at the data for 1-Bromopropane, is I want to compare that relative to compounds that I know have more data and I have more experience with. 
                  So, with that, I'd like to open it up to my other discussants.
                  DR. KENNETH PORTIER:  Dr. Blystone.
                  DR. SHERI BLYSTONE:  I sort of volunteered for this question, thinking naively that I'd be able to get some value out of ECHA database that nobody else could find.  But that's not true.  But I will take this opportunity to -- so that was really my interest in this.  
                  I think this runs squarely into the problem that we had, even with PV29, that there are studies out there, there's commercial values associated with those studies; and until we resolve that question, you're not going to get the full studies, if there's no commercial incentive to provide those to the Agency.  There was certainly no legal requirement to provide them.  
                  Imagine if there was a way to get that data that didn't adversely impact the commercial value of those, that perhaps that would be an opportunity.  Other than that, I agree.  I don't have anything to add to what Bill said.
                  DR. KENNETH PORTIER:  Dr. Cobb.
                  DR. GEORGE COBB, III:  Thank you.  I agree with much of what's been said here.  I will maybe start the conversation about the adjustment factors.  To me it seems like an adjustment factor of ten for a chronic value that's based on one acute value and an acute to chronic ratio of ten.  Probably too low because you've got uncertainty -- excuse me, an adjustment factor of five for a known value for the acute.  So it doesn't seem to me that you can have just 2x more than that for converted value.  But I'm going to, again, leave that to the toxicologist to speak to more later.
                  Also the epi study that was used to develop the hazard quotient has 95 percent confidence in that report -- or at least in that summary.  It also has a NOEC, and neither of those are presented in this assessment.  Again, perhaps in the absence of any other data, you would err on the side of conservativism and go with one of those lower bounds.  The highest toxicity was predicted to be 17.7 milligrams per liter, and the lowest toxicity of 33.3.  That's the 95 percent confidence bound.  And then the NOEC predicted to be 1.77.  And those look like factors of ten.  But that's probably where the doses were set at.
                  Sorry I'm reading through these.  I'll echo something that Dr. Doucette said about the volatilization from soil or the contamination of soil.  I think that's possibly a place where we definitely need some clarification and addressing in the risk assessment.  I know you haven't had any time, but the .93 milligrams per liter for the algae is still not addressed, how that was determined not to be used in the assessment. 
                  I maybe can talk to Bill about this a little bit more afterwards, but the degradation   half-life of 1-BP is estimated to be one to twelve days in the atmosphere, which maybe sets the lower bound of exposure as what's being released rather than that being the exposures.  So there's probably a way to model that to what the exposure would be if there was a continuous release.
                  Also as far as the HAP assessment for atmospheric exposure, data from that assessment should be considered here to assess what people are actually being exposed to, because we're being exposed to -- especially workers are being exposed to something at work and then they're also being exposed at home.  Or as they're driving around their neighborhoods, those kinds of things should be considered or we're not protecting those populations. 
                  Finally, maybe I should ask this as a question.  It seems like if these compounds are used in insulations, and they are used in cleaning computer products, that schools and exposures in schools are important for children.  And I'll leave it at that. 
                  DR. KENNETH PORTIER:  Dr. Schlenk.
                  DR. DANIEL SCHLENK:  Thanks.  So I'll just kind of go through 4.1 and 4.2 separately.  I think Bill did a great job sort of describing the fate issues.  I tend to agree, and almost verbatim said the same sort of things. 
                  With 4.1, the ECHA documentation, it seemed to me the question was is it okay to use this data.  And I would say definitely.  Given the paucity of data that we have for any of the other compounds, I'd say use whatever you can get, within reason obviously.  But particularly in this approach, I think it's definitely warranted.  Particularly at least in a weight of evidence approach.  
                  In fact, we talked a little bit about the exclusion of so many articles and references.  This is a great example of why maybe we shouldn't exclude so much, and that they can be used in a weight of evidence sort of capacity.  Because there's definitely some papers that I found when I did a CiteFinder search that were not included even though the number was there.  And I'll talk about that here in the effects side of things in just a bit.
                  I've always been puzzled by the exclusion aspects of literature.  I know there's criteria that you have to use.  But I think as an academic, I think all the literature has some value, even though it may not meet some standards.  But this is a great example of a language requirement that could be easily remedied.  
                  In fact, I've been on like 30 of these things, and I know we've had Canadian scientists as ad hoc members.  If the French study was that important and data is so hard to come by, it would seem to be that would be a no-brainer to bring some of the French, you know, Canadian guys on board.  Even if it were for that one study, they could still evaluate the same things that Bill and I are actually doing.  Japan, obviously, is a bit of an issue, but I do know we've brought people from overseas, at least on FIFRA panels.
                  
                  UNIDENTIFIED MALE:  There's also Google Translator.
                  
                  DR. DANIEL SCHLENK:  Yeah, there's that too.  Google Translator works well.  Dear translator works well, too, I just found out.  So, anyway, that's my 4.1 comments. 
                  4.2, I kind of split this into exposure and effects sort of components.  First of all, a conceptual model.  Again, this was the same thing that came up with some of the other compounds a month or two ago.  But if this compound is an insulation and it's at five percent, that seems like a pretty high concentration.  It would seem to me that landfill would be another significant environmental source, particularly for leachate.  
                  If the compound has a log Kow of 2 and it does have that water solubility component, there's definitely potential movement into waterways from that.  Especially given the stormwater episode for the TRI, I would think that landfill would be another sort of source that was there. 
                  Again, I know you don't have data for this, but it would be something to include in at least the uncertainty analysis at the end.
                  Based again upon Bill's comments, a log Kow of 2 still represents 100-fold more likelihood that this compound is going to be in inorganic media versus water.  So even though the VCF is predicted to be low, I think you still are going to see issues related to that, as well as size.  I mean, that's the other thing when I teaching ADME, is it's not just -- the felicity it's not just charge.  It's size and it's a small molecule.  So again, at least partitioning into organisms, you might be a little bit underestimated there. 
                  I'm not convinced these compounds don't partition into sediment.  But again, we were told not to recommend research studies, but to identify research gaps, and this is definitely a research gap, that at least a registrant or a manufacturer should be able to remedy, perhaps, if that were possible.  I would sort of estimate, again, it's pretty well biodegraded.  But again, there's some uncertainty there with that.  
                  So the good news is -- I would say, you could look at this also half full.  The good news, having been here three years ago, there was zero on environmental relevance, which was pretty boring for me.  
                  But the good news is I really appreciated the water exposure that was provided by TRI, so that was obviously the good news.  I'd recognize it did represent a worst-case scenario with a range of values really dependent upon water volume, at least in the E-FAST model.  And I talked a little to Dave about this as well.  It seems that that seems to drive the concentrations that are actually predicted.  
                  I guess my only concern with that, kind of boils down to what Bill mentioned earlier, is the nomenclature and the upper end number.  I'm more comfortable with a percentile number, that if you're going to say this is the 90th or 10th percentile, whichever way you're using it in that regard of exposure, this gives a little more credence to the number instead of saying that.  I like that you presented the range of values that was there.  
                  So that's the good news.  The bad news, obviously, was that you have one facility and that value is based on stormwater, which probably -- I'm not sure that represents a real world sort of exposure scenario.  I'm guessing most of this would come as a point source, input from a facility.  
                  So that's the bad news, is you've only got the one point.  And I recognize it's the only data you've got.  One issue, one thing I would recommend, it sort of goes to Bill's issue, is there's a ton of halopropane concentrations that are measured in MPDS data that's out there.  I don't know what it takes to go into that.  I just emailed my colleague at L.A. County San, that they measured nine different halopropane compounds.  
                  So, using those values, you might be able to at least get at an estimate of what's there, at least as an exposure side of things.  Again, just something to throw out there as a possibility to enhance your data set.  And again, these compounds have very similar physical chemical properties, and it would be nice to -- I think you can make that comparison.
                  On the effects side of things, and this was mentioned before, I honestly think that the methods were appropriate given the data that you guys had.  Using the safety factors and AFs, that's a standard practice using the ten and ten, and the two and five.  I understand that is the case.  
                  Again, it's a tremendous amount of uncertainty, and this will deal with Question 6 or 7, whatever it is, from the risk characterization side of things.  I just thought that needed to be a little bit more described, that this is really, really uncertain, that you're basing this on one value, and there's some huge data gaps there, particularly when you don't have any reproductive or developmental data here.  And again, these are huge data gaps that need to be filled by somebody.
                  The other thing that I found lacking -- and it kind of goes back to the CiteFinder search, there's some pretty solid papers out there on QSAR AOP-based methods that could have predicted at least more acute data that could have been utilized in the assessment.  Again, given the investment the Agency has made into adverse outcome pathways, I'm shocked that that hasn't been utilized as much, even -- not only for the eco side, but also for the human health side, talking about some of the mechanistic tox data.  
                  I do recognize the MOA is in there, but the AOP stuff still is lacking.  And I'd mentioned that three years ago too, as well.  So I'm happy that you included the water stuff.  I'm still a little bit disappointed the AOP stuff is not in there.  I'll provide those references, as well, for the AOP narcotic agent acute data.  
                  Again, yeah, given the mammalian toxicity data where we do see reproduction and developmental neurodevelopment being sort of targeted, it would seem to me as zebrafish assay would be a really, really good thing to run on this, at least for the eco side, as well as maybe the mammalian side just to confirm.  I recognize we're not supposed to recommend research, but those are things that would be good to have. 
                  DR. KENNETH PORTIER:  Dr. Twiss.
                  DR. MICHAEL TWISS:  Thank you.  In response to Dr. Schlenk's comment about the French language, we have someone on the panel who speaks French. I did my doctorate in French.  
                  And so I'm going to answer in response to these specific questions where I felt most comfortable speaking about.  And with respect to 4.1, it becomes a question of trust, trust in science.  We all agree that the EPA must use the best science available to support its conclusions.  And the application of a robust critical review of the available aquatic toxicity studies to select and use only those of high quality is an example of this approach.  And that gives the public confidence that the EPA is indeed using the best science available. 
                  An example of that, they took 147 studies and just pulled it into one.  However, neglecting studies due to them being written in a foreign language and not having a U.S. sponsor, I believe is illogical.  And as was mentioned this morning by the woman from the University of California San Francisco, it's a non-empirical based scoring system.  So it just doesn't fit the way that US EPA wants to be selecting these studies.
                  So those documents in the ECHA database should be examined fully to see if they're usable, especially given the fact that so few acceptable studies are usable.  One out of 147.  And the fact that data from the ECHA set of studies provide results in support of the single study on aquatic toxicology that was deemed of high quality, yet remembering the comments made by Dr. Doucette, to justify the use of ECHA data derived from studies that were not examined in detail is not a powerful approach, and therefore seen as an unreasonable approach.   
                  Similar statements made in the public this morning, and in the submitted written comments that were received, and I'm pointing outlined the Environmental Defense Fund, Earthjustice and Dr. Woodruff, as I mentioned from the University of California San Francisco, support that view.  Thus, the public confidence in this approach is rightly not very high.  So there has to be some effort to examine those data.  You just can't blow them off, especially since, this was mentioned earlier, there's such a paucity of data.
                  With respect to the second question, 4.2, the uses of 1-BP have destined to atmospheric waste disposal fallen or during use rather than entering surface or groundwater by any other means other than accidents or leakage and spills.  
                  Although the solubility of water is moderate, two factors prevent it from accumulating to its maximum solubility in water, notably its high volatility and low partition coefficient to organic matter, both dissolved and particulate.  And also, I'm aware of the criticisms of considering just that by Dr. Doucette.
                  The estimate of surface water concentrations in those two rivers in Texas, where the Nueces River and some unnamed water body is conservative, and that did not assume any degradation in the water, assumed low stream flow discharge, and assumes a total annual discharge occurred in one day.  So the estimated concentration of surface water under these conditions would be very low, at least considered to be 70 parts per billion.  
                  And I think that's kind of what it would be.  I don't think it should remain high because of it's going to want to leave and get into the atmosphere.  
                  Looking at some other references, note that the USEPA Great Lakes National Program Office supports the Great Lakes Fish Monitoring Surveillance Program.  And they measure legacy organic contaminants and chemicals of emerging concerns in fish flesh in the Great Lakes.  And the Great Lakes is a region that's suffered from environmental discharges from the population of 35 million people living in the basin.  So anything that bioaccumulates should be detected in fish.  
                  And the PI of that program, Dr. Tom Holsen, I spoke to on the 27th of September, and he says it's not something that they measure in fish flesh or monitor it.  And they would know because they scan for a screen of compounds.  So that's somewhat heartening.  That's all.
                  DR. KENNETH PORTIER:  I have a question for those of you who've already commented.  Earlier this morning there was sort of a discussion about the ECHA data that brought up the issue of the green algae.  And on page 138, the report says, "Secondly, one additional study with green algae was identified in the ECHA database which was not presented in the June 2018 problem formulation.  These updates do not affect the preliminary conclusions of the screening level analysis presented in the June 2018 problem formulation.  So the preliminary conclusions that additional analysis of environmental hazard remains unchanged."
                  And I wanted you guys to comment on that, that incorporating the -- it sounded like the algae were much more responsive to this chemical than the fish were, but that was disregarded.  Any comments?  Dr. Cobb?  Or Dan?  Dr. Schlenk.
                  DR. DANIEL SCHLENK:  One comment you can have on this is that -- I mean it's not very common to include a chronic algal value that's there.  So they're actually including a chronic fish value that essentially kind of says the same type of thing.
                  So if you've got the chronic fish value that's not showing any effect, then oftentimes that can -- the conservatism that's present in that would sort of provide enough safety to sort of oversee the algal effects.  But that's how I kind of look at that.  I don't know if anybody else has something to add.
                  DR. MICHAEL TWISS:  I'd have a hard time with considering chronic exposure in algae, considering they kind of reproduce rapidly and they will evolve in a flask over a week.  And plus these ECHA data aren't -- we don't have the studies available to us to criticize whether or not they were done properly.  So I don't think we can discuss the number. 
                  DR. DANIEL SCHLENK:  That's a good point, actually.  Because I think they were actually using it to confirm the Geiger study.  So that was kind of the primary use of that.  Wasn't necessarily to set it as its own standard, which is another reason I guess you wouldn't use the algal study.  We don't usually do chronic algal studies, so that's just not a -- 
                  DR. KENNETH PORTIER:  Okay.  Just wanted to kind of get that on record because I did hear that kind of as a criticism this morning; and I wanted to get the feel for it.  
                  Is there anybody else on the panel who wants to comment?  Dr. Doucette, you wanted to keep going?  I think we've got a clarifying remark from EPA.  Do you want to jump in at this point? 
                  MR. GARRETT JEWETT:  I just wanted to clarify quickly what you were saying.  Hearing those words back at me, I realize this may be not as clear as it could have been.  So regarding that algae study, we did incorporate the algae study into the assessment.  
                  What I was trying, I guess failing, to say, was that since the problem formulation where this screening level analysis was initially proposed, since the publication of that, another algae study was identified in the ECHA database.  So that was just mentioning that if you're reading the problem formulation and you saw this screening level assessment, and then you jumped over to the risk evaluation and you saw, oh look, there's another algae study in there.  Where did that come from?  It was just clarifying that that's where that additional endpoint came from. 
                  DR. KENNETH PORTIER:  I didn't read it like that at all; but thank you. 
                  Dr. Doucette.
                  DR. WILLIAM DOUCETTE:  Just a little bit of follow-up, I guess, to several comments that were made and the whole idea of water concentrations, bioaccumulation and those sorts of things.  I totally agree that it's relatively low.  But again, that log Kow of 2 is a factor of 100.  And oftentimes you just look at Koc and Kow as being equivalent. 
                  The other thing that's important to remember in all these things is it really depends on the rate of input relative to the rate of output.  And I've been on -- one, two, three different sites -- and again this is TCE not 1-Bromopropane, where we were actually worried about things in surface water, not because it persisted very long, is because the input was greater than the volatilization rate.  
                  It was a relatively stagnant pond.  It had a lot of biological material that actually served to help restrict the volatilization, and the concentrations were close enough that we were actually asked to go and measure them because they were close enough to suspected toxicity values.
                  So I think that it's nice to look at some of these partitioning values and make general references to them, high or low or medium, but you have to be careful.  It's really a dynamic system and it's kinetics, which we don't get from that.  So there are cases where the input is greater than the output, at least for reasonable periods of time. 
                  We also showed for a TCE cite that it bioaccumulated into earthworms with C14 labeled.  Even though -- again, TCE is not the same as this compound, but it's similar enough that I wouldn't be surprised under the right circumstances that we could see that. 
                  So, again, it's just being careful -- in the terminology that is used to give blanket statements that there's nothing there, there's no risk or no chance of environmental concentration, it really depends on the situation and input versus output.  Sorry for getting on my bandwagon again. 
                  DR. KENNETH PORTIER:  I think, Dr. Schlenk, you wanted to comment.
                  DR. DANIEL SCHLENK:  Yes, I couldn't agree more in terms of your comments, Bill.  I guess the question is here -- I mean these are things that I think need to go into the uncertainty analysis.  Because obviously the data that they have is so limited.  It's really, really tough to do that.  
                  But the movement question is definitely something I think needs to be addressed, at least for those log Kow(2) compounds, because they're definitely absorbable and movable.  But again I think this is something that should go into the uncertainty aspects that are there, for future studies or whatever. 
                  DR. KENNETH PORTIER:  Mr. Kaufman, is your flag up for a comment or just left up?  I thought it was a legacy.
                  MR. ALAN KAUFMAN:  Sorry.
                  DR. GEORGE COBB, III:  Since we're mentioning uncertainties, it's my preference or perspective that those uncertainties should be included in the adjustment factors or the uncertainty factors earlier, rather than just spoken to as uncertainties later.  And towards the comments about the algae, that was me that started that.  I agree with the agency, that even if you put them in, you would not get to a higher quotient over 1.  So we're totally in agreement there. 
                  I guess my point was more towards future assessments.  If we're not trying to protect algae or if that chronic assessment's not going to be something germane to protecting algae, just state it up front and be done with it; rather than including it and then saying we're not going to assess it because of not protecting fish. 
                  Then, finally, to wrap back to something Dan said about what other data could or should be used, I think some data should be gathered by the trade organizations or the industries to inform these exposure assessments.  And that's really the crux of the issue, is those data aren't available. 
                  DR. KENNETH PORTIER:  I want to reiterate something that Dan said on the adjustment factors.  I know he and I talked about it and he says, well, these are standard and well understood, but they're not well documented in the report.  Because I looked at him.  I said, "Where did that number come from?  I couldn't find it."  And I kind of came to the same conclusion, well, that's just a number they used.  And I think that's what you said, this is the number we used.  It would be nice if that were a reference to where the heck that number came from.  Dan.
                  DR. DANIEL SCHLENK:  Just one other thing I forgot to mention with Bill's comments.  There is a term, in terms of the input that we use for pharmaceutical input, and it's called pseudo-persistence.  And that's exactly -- I mean obviously it depends on the loading.  And given the volume of these, I'm not sure what the loading is that allows pseudo-persistence, but that's a terminology issue that seems to be coming up more and more, particularly with compounds that aren't supposed to be absorbed and aren't supposed to be persistent, but they actually are because of the loading issue.  So I just wanted to get that on the record. 
                  DR. KENNETH PORTIER:  Dr. Johnson.
                  DR. MARK JOHNSON:  I heard others say this and I just want to reiterate, I think it would be useful to do some in silico projections.  I think EPI Suite has got a model to do this, right, where you can compare your refined value, using the adjustment factors, to what the QSAR would predict.  If they corroborate, then that would be very nice if that were to happen. 
                  Just a quick comment.  There's a paragraph -- actually it's a sentence on page 141 -- that talks about concentrations.  If we could normalize that for units, that would make it a whole lot easier for interpretation.  
                  In other words, we always do milligram per kilogram per day for human health.  It would be good if we kind of did that and just used exponents just to be able to understand exactly what was done in these calculations.
                  DR. KENNETH PORTIER:  Any additional comments?  Committee members on the phone, any comments?  Are they still there? 
                  Seeing no comments, I'll turn it over to EPA for any clarifying questions or comments. 
                  DR. STANLEY BARONE, JR.  I have two process issues that I wanted to clarify on, and one clarifying question.  The first process is I heard Dr. Blystone talk about the data issues.  And then I think there's some -- maybe some misunderstandings about what the data is and who owns the data.  So I wanted to talk a little bit about that, because it's not just this case, but it's going to be other cases.
                  In FIFRA, we have 10(g), which is a particular element of FIFRA where data owners can get remuneration for the value of their data, others can use their data in a submission package.  We don't have that under TSCA. So that's one issue, background issue.   
                  Then the other issue is the ECHA database.  The data, robust summaries, is what's provided to ECHA.  ECHA doesn't own that data.  The companies or consortia own the data.  And so if there's not a U.S. sponsor or a U.S. company in that consortia or owning that data, we don't have direct access to get that data.  
                  So it's not like we just oh, go take a French translation of a study or a Japanese translation of a study and translate it.  It's not like it was published in French or Japanese and we can just translate it at will.  
                  So there's a couple of process issues we're still trying to work through.  And even with our colleagues across the pond and in communication with their registrants, encouraging conversations to proffer this data, offer this data to us, to reduce uncertainties and provide more transparency.  There's some things we have not worked through, because of the way -- we have different statutory authorities.   And some comments -- oh, just sue or just use all your TSCA authorities.  Well, if it's not a U.S. company, it's not a U.S. entity.  There are some real challenges here.  So it's not as simple.
                  Then we're trying to do a risk assessment on a schedule, three-year schedule.  And again, this is a little more protracted because this is a do-over.  We started this back in 2014.  And then TSCA happened, and then we did it over.  So we've had more time on this particular one.  So I just wanted to lay that out as some background. 
                  There was also comments made about, oh, well, we can just have a non-U.S. citizen or a foreign speaker as part of this panel.  Well, actually, we now have regulations that limit peer reviewers on our panels or have limitations to our FACA committee composition.  You have to be a U.S. citizen or a Green Card holder, and there are some very rare exceptions where we can have foreign participants in a federal advisory panel.
                  So it's not so simple.  And as was pointed out, it's great when we have foreign speakers, U.S. citizens, on the panel who can actually read and interpret and advise, but that's not -- you just can't go get the world experts like we used to do.  And the DFOs know about this and know the limitations that this has caused in the constitution and makeup of ad hocs and of standing panels.
                  And that's a relatively recent development.  Some of you who've served on federal advisory panels know that it hasn't always been that way.
                  So I just wanted to put some of this process stuff out there for you to understand.  It's not because we're not trying.  There are process issues and regulatory issues that we have to work with and work around. 
                  Now I'm going to shift to the science, and I actually have a clarifying question.  And this goes back to discussion about mode of action and AOPs, and I heard Dr. Schlenk speak to the Agency's investment in AOPs.  We don't have a lot of quantitative AOPs, very few in fact.  But your guidance on which AOP you were actually referring to in your comments, and how you think that might be applied, would be very, very helpful, to bolster this. 
                  And again, this kind of gets at Dr. Doucette's comments about reading across and looking at similar compounds and building the evidence with other compounds.  And again, that's why the Agency and other agencies are investing in AOPs.
                  So more constructive comments on that will help us, not just for this chemical but for other chemicals.  So I appreciate that discussion.  We do.
                  DR. KENNETH PORTIER:  And especially going to be good for tomorrow's discussion as well.  
                  Dr. Blystone.
                  DR. SHERI BLYSTONE:  Just following up a little bit more on the ECHA database and the robust summaries.  I mean, that is publicly available information.  The robust summaries were designed to provide information without giving away the commercial value in the study itself.  
                  So I think the question to the panel is, it is publicly available information that, in this case, EPA was unable to get the studies for.  Is the way that they used that appropriate, or should they just ignore that publicly available information?  
                  And to me that's the heart of the question here.  It's like it is there, there is quite a bit of information in the robust summary to allow you to evaluate the type of data that it is.  It isn't enough to do a data quality evaluation.  You can't put it through systematic review.  But is it something that they should consider?  And if they did -- and as they did consider it, is that an appropriate use of it? 
                  For me, I think it would be -- we're talking about why you throw literature out the window.  This is information.  It is relevant.  So, for me, it seemed like it was an appropriate use of that publicly available information to further inform this assessment. 
                  DR. KENNETH PORTIER:  Any additional comments?  I think the panel is worn out.  This has been a long day, and tomorrow's possibly going to be even longer.  At this point I'm going to kind of call the end of the meeting, turning it over to the DFO if she has any additional comments, especially when we begin tomorrow morning.
                  MS. TAMUE L. GIBSON:  The only additional comment I have is we'll see you at 9 a.m. in Ballroom C.  So we'll see you then.  Thank you. 
                  
                  (MEETING ADJOURNED FOR THE DAY)
                        OPENING OF THE MEETING - DAY 2
                  
                  MS. TAMUE GIBSON:  Good morning and welcome, everyone.  I would like to thank you all for participating in today's meeting for the risk evaluation for 1-bromopropane.  At this time, I will like to reconvene the meeting of the TSCA SACC.  As a reminder, I would like to indicate that the meeting materials are located on regulations.gov and the docket number and the website are noted on the meeting agenda.  I would like to now turn the meeting over to our chair, Dr. Portier.
                  
                        FOLLOW UP FROM THE PREVIOUS DAY
                  
                  DR. KENNETH PORTIER:  Thank you.  Good morning, everyone.  We're not going to introduce all the committee members because, I think, we know each other, and we did that yesterday.  There's less of a public here and all of that public was here yesterday, so they know our names.  
                  I want to finish up with a few comments on Question 4 before we jump into Question 5.  I was talking to Dr. Kissel this morning, and he indicated that something in the discussion at the end of the day fired off some neurons.  John, it's not quite there, yet so I'm trying to stretch this out while his computer comes up.
                  DR. JOHN KISSEL:  I think I have it in my brain.  Well, I was going to present a slide which I don't have ready.  The discussion yesterday when Bill mentioned indoor air quality, it strikes me that for VOCs in the context of Lautenberg that there should be a general public consumer pathway for indoor air which is not -- that we've been arguing about HAPs and whether we'll ever get to that or not, but HAPs is a different issue and that's outdoor air.  The team study showed pretty clearly that indoor air -- for a typical person who doesn't live immediately adjacent to a production facility or a waste disposal facility, indoor air is a much larger source of VOC exposure than is outdoor air.  And it's because products are just generally in commerce and you take them into your house if you have aerosol cans, paints, all manner of things.  Wite-Out used to be -- the solvent in Wite-Out was TCE.  
                  There were lots of ways that you could be exposed to VOCs from just buying things that were in commerce which I think is a large part of the inspiration for the Lautenberg Act.  Those data from Bayonne, New Jersey, which showed that so clearly, that's Frank Lautenberg's home state.  I think he was aware of that data.  
                  I think this notion that there should be an overall accounting -- if you're going to sell chemicals in mass quantities in society, then you want to know, what are the consequences of that?  And one of the consequences is that everybody's going to have trace levels.  And we have heard that there is maybe a biomarker.  It has an asterisk on it because it's not specific, but there is reason to believe that everybody in society is getting some exposure to this compound.  
                  We know that it's in some common products like insulation board.  And so, if it's in houses, then everybody's breathing some of it all the time, and those exposures may or may not be of toxicological significance.  But they are of significance in the context of the Lautenberg Act in understanding because basically the idea is that somebody should be accounting for chemicals in commerce and we should have some concept of to what extent we're being exposed to those things.  
                  So, I think, for this VOC and for other VOCs, that should be part of the overall matrix.  The schematic of how people get exposed should include for things that might show up in common commercial products.  Indoor air should be in the mix because, I think, that's what the Lautenberg Act is all about or, I think, that's part of the inspiration for it.  And so, I think, we're missing something here, and the HAP argument is a little bit of a diversion away from where a typical American consumer might be actually exposed to this compound.
                  DR. KENNETH PORTIER:  Thank you, John, and you're going to provide this to the lead to add.  Is there anything else on Question 4?  Anybody wanted to -- okay, well, we're going to then proceed to Question 5.  Oh, I'm sorry.  
                  MR. MARK HARTMAN:  Chair, so I heard the comments, and I think they're definitely an extension of what we discussed yesterday.  I'm not understanding if there's a recommendation there.  Is the recommendation to -- because we've already spoken to the issue of assessing consumer products and assessing insulation in the indoor environment, so I'm not sure what the recommendation is.  I'm paraphrasing.  I'm not trying to put words in your mouth, but is the recommendation about aggregate exposure assessment?
                  DR. JOHN KISSEL:  Well, I am an advocate for doing aggregate exposure assessment.  I think that maybe what I'm arguing for is an attempt to close the circle a bit about the -- so you've done modeling with CEM that says we expect these levels are showing up in housing.  The question is would those levels explain the urinary metabolite that's apparently widespread in the population.  Can you explain what those exposures might be?  So that biomarker might be coming from someplace else because it might be a different brominated compound, but this a 25 million pound a year chemical, and it would take a lot of other things to account for -- to be a similar kind of mass in general play in society. 
                  And going a step beyond the -- this is just my general -- somebody mentioned -- I think Holly mentioned the merit of mass balances yesterday.  In exposure science, the object shouldn't just be to generate an estimate and say, okay, we don't have a problem.  It ought to be to generate an estimate, and then test the estimate against biomarker data and say, oh, not only do we not have a problem, we actually understand what's going on.  And, I think, this exercise falls short of that.  Just generating a number and saying that's what's out there needs to be proofed against whatever biomarker data you can find.
                  DR. KENNETH PORTIER:  This is Ken Portier.  Also what I was hearing was that what you've done is a condition of use, and you've looked at indoor air for one particular condition of use.  Then what I'm hearing from the toxicologist to say, but that may not explain the whole internal air issue.  Because of the other consumer uses, multiple consumer uses that utilize this chemical may provide less or as much exposure in the indoor air as that insulation condition of use.  Does that make sense, John?
                  DR. JOHN KISSEL:  Yeah, I think the foam board was one thing that we think is there, but there was still this issue of propellant in spray cans and how widespread is that.  Maybe it's not widespread, but there was an awful lot of, well, we don't really know.  You can make projections, and then you could take that to, well, okay, if people are breathing this much, then this is what should be coming out in their urine and is that a large number or small number compared to that biomarker that happens to be in most of the people in the United States?  And so, do we really understand what's going on?
                  I don't recall -- and I'm running behind today -- but I was going to look up 1-bromopropane as an indoor air pollutant, and I don't remember seeing it in some of the conventional compendiums.  But that could be not because it's not there but because it doesn't fall in the GC scan that people are routinely doing.  There's lots of reasons why things show up in the list of common indoor pollutants that aren't necessarily because they're actually the things we ought to be most concerned about.  So, I think, the exercise under the auspices of Lautenberg ought to be actually not just kind of check off a box and say, okay, we assessed this, and we didn't get high risk; it ought to be, do we really understand what is going with this compound?  Where it's going?  What is the life cycle assessment?  What is the life cycle of this compound?  And how much of it is out there?  How much of it is passing through human beings on its way into the atmosphere where it ultimately breaks down with sunlight?
                  DR. KENNETH PORTIER:  Yes, Dr. Blando.  Oh, I'm sorry.
                  DR. SHEELA SATHYANARAYANA:  I just wanted to say that I completely agree with John in reading these documents and listening to the discussion yesterday and just knowing the data that within NHANES and then with NCS, you have over 90 to 99 percent exposure and you know that that metabolite is in the metabolic pathway.  And then that's even further confirmed by the occupational studies where you see correlation with the exposures and urinary biomarker.  You could do better modeling of general exposure scenarios to better understand the risk.  And I felt like it was just written off in the discussion.
                  DR. KENNETH PORTIER:  That was Dr. Sathyanarayana, right?  Your name.
                  DR. SHEELA SATHYANARAYANA:  Oh, I'm sorry.  I didn't even recognize.
                  DR. KENNETH PORTIER:  Just remind the committee to say -- at least I tried this time.  I didn't say Dr. Sheela.
                  DR. SHEELA SATHYANARAYANA:  I haven't had a lot of coffee.  It's Sheela Sathyanarayana.
                  DR. JAMES BLANDO:  Sure, this is Jim Blando.  I would agree that, certainly, we want to understand all the exposures and, in particular, the consumer product exposures.  But I do want to just point out a concern that I had when I read the documents is the biomarker that folks are referring to, I don't believe, is specific to bromopropane.  So you can really have a lot of misclassification error of exposures and, I think, that's something that really just needs to be very well understood.  Perhaps I'm not understanding it, but my understanding is that biomarker's not specific.  So we would definitely need to understand what other compounds would be related to the presence of that biomarker.  
                  Certainly, if you have workers that are working with bromopropane, you can easily make the attribution to that specific biomarker.  But, in the general population, if people are exposed to multiple chemicals, and this biomarker is not specific, you could very easily misclassify their exposures.  So I just think that's something that really has to be carefully thought out.
                  DR. SHEELA SATHYANARAYANA:  Can I just respond to that really quickly?  So I do think you can do that kind of modeling because you do know that it's a biomarker of other compounds, and you know those exposure levels.  And I feel like that level of modeling is not so advanced.  I mean, it's something that people do all the time.  So I do feel -- I understand that it's not a specific biomarker, but you still can do the advanced modeling with the other compounds.
                  DR. KENNETH PORTIER:  Dr. Gilbert, I think, and then Dr. Schlenk.
                  DR. KATHLEEN GILBERT:  I agree with what's being said.  So it may not be a specific marker, but in the document, which is going to read by many different sorts of folks, it says in there, all it says is, it may not be a specific marker.  And I followed it.  
                  At one point, they don't have any references or explanations, and, at one point, they reference one of the supplemental documents.  So I looked in there, and it just said, basically, look, it's so widespread and we don't think 1-BP is that widespread, so this must be wrong.  It doesn't have any explanation.  It doesn't have any sort of rationale for that statement.  So, to me, if you're going to say that it's nonspecific when it's such a big issue as far as the public being exposed to it, then you'd need something to back that up as opposed to just saying it may not be specific.
                  DR. KENNETH PORTIER:  Dr. Schlenk.
                  DR. DANIEL SCHLENK:  We just looked up a paper.  In 1966, there is a metabolism paper of halopropanes -- iodopropane, 1-iodopropane, 1-chloropropane, nitropropane -- any one carbon propane is going to give you that metabolite.  I mean, it's just a basic metabolic biotransformation concept.  And it's not even going to be enzymatic; it's actually it's a substitution reaction for glutathione on those electronegative halo groups that are on that molecule.  
                  And, again, we said this three years ago; all you've got to do is look at drinking water.  There are halopropanes in drinking water without a doubt.  There's one 2-halopropanes.  There's at least nine of them that are monitored from wastewater discharges.  So you know that they're formed by chlorinated byproducts.  I don't know.  It just seems to me that that's a fairly obvious one plus one equals two kind of situation where it's not specific.  I mean, chemistry is chemistry, I guess.  
                  DR. KENNETH PORTIER:  Can you provide that reference to Dr. Blando?  Dr. Blando.
                  DR. JOHN KISSEL:  It should be possible.  One of the things that gets monitored pretty well is drinking water, and you ought to be able to do a mass balance on, how much of that stuff is drinking water, and how much of that biomarker could be accounted for that, and is there a missing piece or is there not?  
                  In terms of -- otherwise, what is the annual production volume of those compounds?  This is a huge 25 million pounds a year are coming in.  This is a huge amount of material.  There aren't that many compounds out there that are produced in that kind of mass annual input into society.  And it might very well be that the drinking water would swamp, but, okay, work out the numbers and show me that.  
                  DR. DANIEL SCHLENK:  Yeah, I mean, I can't do the mass balance, but, based on Bill's data the other day saying that you find chloropropanes in other things too, I'm just guessing drinking water's one possibility.  I've never done that, but I think it's a compound that is fairly prevalent.  And, again, that's just one.  And to try to identify every single halopropane that's present out there and try to do a mass balance, I think, is a little bit difficult to do.  I don't know.  It could be done, maybe.  I don't know.  But to try to identify a source of every halopropane and every -- and not even halopropane, even a cyanopropane is going to be on there.  Any nitrile propane is a prominent metabolite through that pathway.  
                  DR. KENNETH PORTIER:  Dr. Blando.
                  DR. JAMES BLANDO:  This is Jim Blando.  I agree that I think it'd be great to do the modeling and really understand this issue.  I'm certainly not at all trying to discount that.  I'm not sure that 25 million pounds a year is a high-volume production -- an HPV, high-volume production or H -- whatever the acronym is, the High-Production Volume chemical.
                  But, from a public health intervention standpoint, the problem is, if you miss a tribute, a source of a pollutant, you may end up producing an intervention that doesn't get you to your goal.  If you have this nonspecific biomarker that's coming from chlorination disinfection byproducts, for example, and you spend a lot of money to control Wite-Out or to control consumer products, you may find that there's no change in the biomarker concentration at all because you're trying to control the wrong thing.  So I would agree with Dr. Gilbert that this seems to be an important issue, especially things that people have cited, so perhaps -- although I know we're not making research recommendations but perhaps -- that's something that would be worthwhile to really try to better understand, so we don't end up with an intervention strategy that's not going to work.
                  DR. KENNETH PORTIER:  To use Dr. Schlenk's terminology, we've identified another data gap that somebody should fill but not this particular group at EPA.  Dr. Doucette.
                  DR. WILLIAM DOUCETTE:  I think this was all my fault anyway with the example of the spray can, and I don't apologize.  It's just something that came up, and it worked its way around.
                  But that might be a relatively simple thing to do is look at the -- if I have a can identify -- and I don't know how widespread this is relative to some of the other splat solvent degreasers that are used in commercial products, but it might be as simple as looking at how widespread that is.  Put that in a house, assume that it all goes out, and I can provide some rate information, and I know what -- you get measurable concentrations in the air, and is that important relative to the insulation scenario or anything else?  
                  I guess that's kind of where I was going with that yesterday, and I didn't maybe explain that as well as I should.  And that might be another way to avoid having to worry about the -- at least it would give you another common scenario of potential exposure in indoor air, independent of the markers and all the other issues that are related to that.
                  DR. KENNETH PORTIER:  Dr. Blystone.
                  DR. SHERI BLYSTONE:  Yeah, just a couple of clarifying comments.  I agree with Dr. Blando.  Twenty-five million pounds -- although, I think this is actually more in the 10 to 15 -- while it's over the one million pound for high production volume, amongst high production volume chemicals, it's not that high.  And on the drinking water side, it's, potentially, the disinfection process that is generating these compounds.  So that has nothing to do with production volume of anything; it's part of the waste treatment process.
                  DR. KENNETH PORTIER:  Dr. Cobb.  Flags down, flags up.
                  DR. GEORGE COBB:  So, to circle back, we're dancing around the issue of we don't have the data we need.  And the data gap can easily be filled by simply monitoring bromopropane indoors in several locations and be done with it.  And that's what the industries producing and using these compounds should do.
                  DR. HENRY ANDERSON:  This is Henry Anderson.  I view this as there needs to be a word of caution that you have this data.  We really don't have all the details about it, but it is of sufficient concern that there may be a consumer or a general population exposure that should be investigated further.  At best, these measurements, now, are kind of a qualitative rather than a quantitative indicator of exposure.  And I think it's enough to not just dismiss it, which I think is kind of done in the document, but rather identify it as a potential concern that needs to be evaluated.  So it's just kind of recognize the issue, not just dismiss it.
                  DR. KENNETH PORTIER:  And I think that's the recommendation that's coming out.  Thank you, Dr. Anderson.  Dr. Blando, did you capture that?  
                  DR. JAMES BLANDO:  Yes.
                  DR. KENNETH PORTIER:  Okay.  Now, can we go on to Question 5?  I didn't mean for this to -- that was an interesting side discussion.
                  MR. MARK HARTMAN:  I just wanted to clarification on the recommendations, and I think this, Chair, I think this conversation was very helpful.  And I think there are actually some actionable things that we can do.  Thank you.
                  DR. KENNETH PORTIER:  Well, I think you captured the kind of disagreement as to the level of importance that comes into the committee.  I didn't expect that, but it's good that we've captured that.  Let's carry that into the human health.  I've asked Dr. Anitole to make kind of a shortened presentation, and we're going to take kind of one question at a time today because we have different leads on different questions.  So we're going to kind of take a measured approach to today's questions.
                  
                               CHARGE QUESTION 5
                  
                  DR. KATHERINE ANITOLE:   Thank you.  So Question 5.   Human Health Hazard and Dose-Response Assessments.  This question pertains to Section 3.2 of the Draft Risk Evaluation.  Summarizing this slide -- EPA considered the adverse human health effects for 1-BP across organ systems.  For hazard identification and dose response, EPA reviewed the evidence for 1-BP toxicity and selected liver and kidney toxicity, reproductive and developmental toxicity, neurotoxicity, and cancer.
                  EPA selected endpoints supported by the weight-of-the-scientific evidence for non-cancer and cancer following acute and chronic exposure scenarios for workers and acute exposure scenarios for consumers.  EPA determined that using developmental toxicity and neurotoxicity endpoints for dose-response calculations would be protective of the most sensitive life stages including the developing fetus for non-cancer points of departures and risk estimates.  EPA used these endpoints to calculate points of departure to assess non-cancer risks associated with chronic occupational inhalation exposures.
                  Question 5.1:  As part of the review, please comment on the choice of these endpoints as PODs for assessing risks in humans associated with acute and chronic inhalation exposures to 1-BP.  Specifically, are there other data that EPA could have considered for the hazard identification and dose response associated with acute inhalation exposures?  If so, please provide specific data and references.  Are there other data that EPA could have considered for the hazard identification and dose response associated with chronic occupational inhalation exposures?  If so, please provide specific data and references.
                  DR. KENNETH PORTIER:  Thank you.  Dr. Barton.
                  DR. CHARLES BARTON:  Good morning.  Thank you.  This draft is fairy mature.  It's already gone through several drafts, and it really is seen here because it's much more advanced than some of the other chemicals that we've looked at.  
                  Well, the non-cancer endpoint selected for points of departure for both acute and chronic exposure scenarios were derived from three studies: WIL Research (2001), Ichihara (2000), and Honma (2003).  All three studies scored high in data evaluation, followed OECD guidance, and good laboratory practice; and this is appropriate.  
                  With that said, typically data from the acute inhalation toxicity study are used to examine hazard and dose response from an acute inhalation exposure.  Such data were not submitted to the committee for review; however, three key and two supporting studies have been submitted to ECHA, however, not enough detail provided in the submissions to ascertain a point of departure.  The point of departure derived by EPA from short-term exposure to the F0 generation in the two-generation inhalation reprotox study was based upon a decrease in live litter size.  It was assumed that this decrease in size of live litters was due to effects on both the male and female rats.  This approach is not generally accepted or considered appropriate for deriving a point of departure for acute inhalation toxicity.  However, it may be acceptable in lieu of data from better studies if the assumption is true that the decrease in live litter size is actually due to exposure between ovulation and implantation in males and females in the F0.  
                  Both post-implantation loss and decreased live litter size are being used for both acute and chronic exposures.  The argument was already made by EPA that the live litter size decrease was from a short-term exposure.  The rationale for how it is now being used to characterize chronic exposure needs to be explained.  It's understandable, but it needs to be explained, so everybody can understand in following.  
                  Details for how the assumptions were derived and calculating dermal HED from an inhalation HEC needs to be provided, and the validity in these assumptions need to also be provided.  It was noted that regarding the BMD calculations supplemental information, pages 22 - 26, it seems to be a MAR calculation for post-implantation risk -- is highly uncertain due to removing the high dose which has a lower trend leaving three points to include the control.  All of which are highly variable where there's post-implantation loss and controls.  This suggested using reduction in litter size predominately.  However, it doesn't seem to matter much as results from both result in similar margin of exposures.  It's further unclear why the EPA used the standard uncertainty factor for animal-to-human extrapolations when they already apply to dose of metric adjustment factor in addition to other allometric adjustments to account for species differences due to kinetics.
                  Comparing the NOAELs and LOAELs from all of the reported animal studies support EPA's choice of endpoints.  Nearer toxicity endpoints were consistently the most sensitive followed by hepatic reproductive developmental renal.  There was a 105-week study that showed LOAEL of 62.5 parts per million for respiratory toxicity.  
                  Human case reports, several indicate neurological effects at around a hundred parts per million, maybe not be able to use these for dose response, but, as EPA describes, certainly add weight of evidence support for neurological findings as a critical endpoint for human health risk assessment.  
                  Regarding other publications, there was a publication in 2018 by Miao et al. regarding mechanism of action in humans.  It's in Human Experimental Toxicology, Volume 37, pages 3 to 12, where it talks about mostly proteomics.  And there's some editorial issues which, if you want to see them, I can point them out afterward.  
                  Recent data should be considered for carcinogenicity.  Specifically, Stelljes (2019) which supports a nonmutagenic threshold mechanism of action for carcinogenicity.  And other data that EPA could have considered for hazard identification and dose response associated with chronic inhalation exposures were identified.  However, the data can be further supported with the limited epidemiological studies or case reports.  Thank you.  
                  DR. KENNETH PORTIER:  Thank you, Dr. Barton.  Next, I think, is Dr. Willett.
                  DR. CATHERINE WILLETT:  I think Dr. Barton captured all of my comments.  Thank you. 
                  DR. KENNETH PORTIER:  Dr. Johnson.
                  DR. MARK JOHNSON:  Initially, I had some issues with the post-implantation risk.  Just looking at the benchmark dose algorithm, it looks like the controls have some baseline reductions in post implantation.  It's hard to see because the y-axis isn't labeled.  Point zero two occurs at the controls in the metric, and I'm not sure exactly what .02 means.  
                  But I also noticed that the high dose was removed, and I assume that's being done to fit the curve better, to meet the data, because it looks like the highest dose actually goes down.  I don't know if that's statically accurate, but it looks -- in other words, it looks like the dose response in the high dose kind of goes down.  
                  And so I'm not a DART person, so I refer to my colleagues who are experts in developmental and reproductive toxicology, but I really question the use of looking at post-implantation risk.  I really don't know what it means for humans; rats have lots of pups, people don't.  However, reduction in litter size does look like a good endpoint, and it comes out to about the same number.  So it's not that big of an issue, but it's something that I would consider.  
                  Regarding the question -- we're just in Question 5.1?  Is that correct?
                  DR. KENNETH PORTIER:  Yeah, we're still on 5.1.
                  DR. MARK JOHNSON:  Okay.  So I'll go beyond that.  I also have some issues with -- or questions -- regarding the use of the BMR one versus five versus ten.  It looks like a BMR of one was used for brain weight, but that's a continuous variable.  And I thought in the text somewhere it said it used plus or minus one standard deviation for continuous variables.  
                  And that's another endpoint I'm not sure about.  What's reduction brain weight mean?  Is that an adverse effect?  It doesn't justify having a one percent BMR, which is really what you're saying.  If you're doing a BMDL, you're saying that your 95 percent confident that one percent or fewer will exhibit this effect exposed at this level in that organism.  So, I guess, I'm just a little confused about that.  And that's all I had.
                  DR. KENNETH PORTIER:  Mr. Kaufman.
                  MR. ALAN KAUFMAN:  Thank you.  I was going to say most of my comments have already been covered, but one issue was in looking at dermal exposure.  In one case, there's fractional used.  I think it was for occupational.  I can't remember which way it went.  And, in the other case, it's consumer; I think it was permeability, and so, I guess, the question is why the difference?  And I think it leaves you open to criticism kind of splitting the difference that way.  In other words, sort of pick a lane and defend it rather than trying to say we're going to go one way for one set of exposures and another way for another set of exposures.  It didn't seem to make a lot of sense, and I think it does open you up to criticism.  I think everything else has pretty much been covered.  Thanks.
                  DR. KENNETH PORTIER:  Thank you.  Dr. Pessah.  Dr. Pessah, if you're speaking, you're on mute.  We'll move on to Dr. Vorhees, and we'll come back to Dr. Pessah.
                  DR. DEBORAH CORY-SLECHTA:  Dr. Portier?
                  DR. KENNETH PORTIER:  Yes.  
                  DR. DEBORAH CORY-SLECHTA:  This is Dr. Cory.
                  DR. KENNETH PORTIER:  Yeah.
                  DR. DEBORAH CORY-SLECHTA:  I can hear my -- I think you skipped over me.  I wanted to -- I'm hearing myself in the background, so it's a little difficult.  But I wanted to respond to the comment.  I'm on this question, but I also want to respond to this comment about brain weight.  
                  A one percent decrement in brain weight -- any decrement in brain weight -- means that we are losing cells that we shouldn't be losing.  And we have a long history of understanding what a poor prognostic indicator it is to have reduced brain weight.  So I personally believe it's a good metric, a very serious kind of metric, and I don't have any problems with a one percent.  I certainly wouldn't want to find out that my child had a low brain weight ever.  Thank you. 
                  DR. KENNETH PORTIER:  Is Dr. Pessah on?  Dr. Vorhees.
                  DR. CHARLES VORHEES:  Okay.  One thing I would like to comment on is under the reproductive section, the Agency uses this post-implantation loss, and the way it's calculated is the number of implantation sites minus live births.  Then under developmental reproduction, we use litter size as an independent outcome except those two both rely on litter size.  So they're basically using the same data in two different ways, and it seems to me that's one outcome, not two outcomes.  So I didn't see why it fell under reproductive and developmental; it's really developmental and it's litter size.  
                  So, when it comes to the brain weight concern, I agree with Dr. Cory-Slechta that a reduction in brain weight is serious, but, in this case, a reduction in brain weight was proportional to body weight; therefore, I'm skeptical that it has significant independent meaning in this case.  In fact, most treatments that reduce body weight to a significant degree will reduce brain weight at the same proportion or less because of brain sparing.  So, if there was a brain weight reduction that was greater -- not proportional or out of proportion to -- body weight, I would think that would be highly significant, but that's not what is seen here.
                  For the neurotoxicity, they cite a study by Honma and another study by Ichihara.  The Honma study effect, which is reported as key, is a reduction in hang time at postnatal Day 21, and, although it results in a very low calculated benchmark dose, I'm not sure that it's a very reliable measure.  At 21 days of age, hang time is not a very reliable index of neurotoxicity.  And what's of greater concern to me is this is in a study where there are only five animals per group.  So this study is not a very strong study, and it's using a fairly unreliable measure of neurotoxicity using hang time.
                  I think the better study is the Ichihara study which uses hind limb grip strength and finds effects after testing at several different exposure and post-exposure intervals.  And there they use nine animals per group, and they find a consistent pattern of effects that persists both at the end of the exposure period and several weeks after exposure.   And moreover, the effect is basically dose dependent in that, at the first exposure point at the end of exposure, they get an effect at 200 parts per million.  At 400 parts per million, they get an effect at the end of exposure and then four weeks later.  Then, in the high dose, 800, they get an effect at the end of exposure four weeks later and 14 weeks later.  So it really looks dose dependent; the greater the dose, the longer the effect persists.  So I think those are the stronger data and the data that would be best used for the benchmark MBCL.  So those are my comments.
                  DR. KENNETH PORTIER:  Try Dr. Pessah again.  I'll open it up to the panel.  Does anybody else want to comment on?
                  DR. DEBORAH CORY-SLECHTA:  This is Dr. Cory-Slechta. 
                  DR. KENNETH PORTIER:  Yes.
                  DR. DEBORAH CORY-SLECHTA:  I used both of these measures -- the hang and the grip strength -- and in young and developing animals.  I disagree with Dr. Vorhees about grip strength.  I think it's actually easier to get standardized data than it is with grip strength.  Grip strength requires a lot of training, if you will, of the investigator, whereas, grip strength is more of a do or die kind of phenomenon and animals simply do it.  So I disagree with that.  I also felt that the grip strength was an interesting parallel to see effects that are seen in humans.  Thank you.
                  DR. KENNETH PORTIER:  Just to be clear, you're disagreeing with Dr. Vorhees.  Is that right?  Dr. Vorhees.
                  DR. CHARLES VORHEES:  Yeah, let me ask you, Debbie, if I can say that, are you not concerned about the five animals per group in that study?
                  DR. DEBORAH CORY-SLECHTA:  I am concerned about that.  I would love to see nine or ten animals in that group.  On the other hand, I feel like grip strength has a lot more variability built into it than does the hanging time.  So, yeah, I am concerned about that.  I would like to see that, but I'm less concerned about hanging time is a measure than grip strength.
                  DR. CHARLES VORHEES:  Well, the grip strength, I will say, that I mentioned the hind limb, they also found forelimb grip strength differences.  They just found more effects on hind limb than on forelimb, but they measured both and they found effects on both.
                  DR. DEBORAH CORY-SLECHTA:  Yeah.  Right.  I understand that.  I'm just responding in my experience to different individuals doing the measurement and sort of the learning curve of being able to do that assay.
                  DR. KENNETH PORTIER:  Any additional comments?
                  DR. CHARLES VORHEES:  I can't resolve that.  I mean, we see that.  We both use those tests, and we both have our concerns about one versus the other.  But I've had some concerns about hang time tests and other early developmental tests in the assessment of neurotoxicity because, not only do I have questions about the consistency of measuring them, but also because a lot of them don't predict adult effects very well.  So, for those reasons, I have concerns about them.
                  DR. KENNETH PORTIER:  Dr. Willett.
                  DR. CATHERINE WILLETT:  I think this conversation also points to some other comments that will come up in the later section about being more transparent about the weight of evidence decisions that went into selecting which endpoints for the dose-response point of departure calculations because, I mean, if you go on a straight weight of evidence, you would do all of the things that you consider.  But I think they chose the other endpoint also because it's protective.  I mean, it results in a more protective estimate but those things -- the factors that went into how they chose them should be more transparently described, I think -- I mean, specifically described maybe in an additional table in the document.
                  DR. KENNETH PORTIER:  Ms. Rudel.
                  MS. RUTHANN RUDEL:  I just have a question for the neurotox folks or, actually, just maybe more broadly, but has there been a developmental neurotox study for this compound?  Or, based on what you've seen, what's the study that you think would show -- would be the design and the endpoint that you would look at to find the sensitive measure and reliable measure of the effect?  And has that study been done or not done?
                  DR. CHARLES VORHEES:  Is this study the best that could be designed to determine developmental neurotoxicity?  No, I don't think so.  I think any of the neuro people could design a better study, but the better study doesn't exist in this area.
                  MS. RUTHANN RUDEL:  So what would that be?  Like, what would you be looking for?
                  DR. CHARLES VORHEES:  Well, it would be interesting to hear what Debbie would say.  From my point of view, I would say that to put less emphasis on these early developmental parameters -- the emergence of various reflexes and things like that -- and put more emphasis on adult assessment to see whether there are long-term consequences of a developmental exposure.  Actually, the study by Honma does that.  There are a number of other later behaviors that were assessed, and they didn't really show much.  So, to me, that indicates that the hang time may not really be telling you anything about the ultimate neurotoxicity of this compound.  They did learning tests and activity tests, and really they only found effects at the higher concentrations.  In some of the tests, they didn't find much of anything at any concentration.
                  There's also some issues around -- there's statistical analysis.  I tried to make sense of the degrees of freedom that they show for their ANOVAs, and they don't appear to match the design.  And so I'm not really sure, in some cases, how they even analyzed their data.
                  DR. KENNETH PORTIER:  That was Dr. Vorhees.  Dr. Pessah, have you joined us?  I'm still trying.  Yeah.  And just reminding the panel to use your name and speak into the microphone because the people on the phone are having a little hard time hearing us.  Dr. Barton, did you capture all of that including this disagreement on the --
                  DR. CHARLES BARTON:  I will have to follow up with the speakers.
                  DR. KENNETH PORTIER:  Good.  Dr. Davies.
                  DR. HOLLY DAVIES:  I have a question.  Since you're talking about Honma, I would like to go back to the data quality evaluation within the systematic review because they hadn't looked at the paper in depth because I wasn't assigned to Question 5, so I was looking forward to this.
                  So I want to go back to the fact that it was unacceptable for the chronic when it was ranked in the systematic review and, if any of you, who clearly delved into this paper and know this to the extent you can disagree about it, looked at the data quality evaluation.  So, for instance, I'm looking at it now, and it got high for statistical methods that you just said you had trouble understanding what they did.  And also it was evaluated both for short-term and for chronic.  Had you looked at that?  I don't know if anyone's looked at that.
                  DR. CHARLES VORHEES:  No, I just looked at the study, and I had questions about it because I thought that the sample size was small and some of the statistical analyses are unclear and I expressed my concern about the hang time test.  And so I, personally, didn't think that the study was a particularly strong one for use in determining the benchmark dose.
                  DR. HOLLY DAVIES:  Okay.  Because I did have that question about the unacceptable and then using it for the point of departure.  Anyone wants to follow up, please tell me.
                  DR. DEBORAH CORY-SLECHTA:  Could you repeat what your question was?  It's a little hard to hear on the phone.
                  DR. HOLLY DAVIES:  Sorry.  The question was in the Data Quality Evaluation, the Honma paper had two ratings, and it was rated high for the acute and unacceptable for the chronic.  And I was hoping that when we got to this point and with people who had dug into the paper would be able to shed some more light on that if you'd looked at the Data Quality Evaluation.
                  DR. DEBORAH CORY-SLECHTA:  Yeah, I hadn't looked at that though.
                  DR. HOLLY DAVIES:  Because that's really the question is how good the papers are as we're using them for the risk assessment.
                  DR. KENNETH PORTIER:  But didn't I hear Dr. Vorhees say that he thought it was better for acute and not acceptable for chronic?  Was that right?
                  DR. CHARLES VORHEES:  No, I wasn't saying that.  No, no, no.
                  DR. KENNETH PORTIER:  Just checking.
                  DR. CHARLES VORHEES:  No, I think the study has serious concerns whether it's for acute or chronic.
                  DR. HOLLY DAVIES:  So I'm just muttering, and I shouldn't mutter to my partner -- not partner, my neighbor -- that sometimes we define chronic and acute differently, and that's why I wasn't sure how people were defining chronic and acute.
                  DR. KENNETH PORTIER:  I think I'm going to look to the EPA and say, did this raise more issues than it answered the question?
                  DR. STAN BARONE, Jr.:  I think it raises some important issues.  I do have some concerns in, at least, some of this dialogue.  There seems to be some misunderstandings with regard to the neurotoxicity risk assessment guidelines, and I want to call out a couple of things and, again, we had provided guidelines to you in our orientation session.  
                  BMRs, the issues of BMRs, and I think, we need to be more clear about why we chose the BMR of one percent for brain weight.  Again, it's because of several factors: One is the adversity of the endpoint.  It is an adverse endpoint called out specifically in our neurotox guidelines.  We do not correct for body weight, and it's because of a very large literature on brain sparing.  It is a pathology.  It is adverse.  Second, with regard to the BMR, brain weight is very well conserved.  The variation in brain weight is very tight, so, from both a biological and statistical sense, the BMR is low.  And I think I've already made my point about adversity.  
                  The comments about having a DNT guideline-like test for 1-BP, we don't have that.  We have the data that we have, and we tried to evaluate it in the most comprehensive way.  Your input on the weaknesses and the strengths of the findings is important.  I would also ask you to consider having a lower number of animals, would that lead us to a false positive or a false negative?  Which direction would you expect?  Again, those considerations will also be helpful to us in revising our evaluation.  Thank you.  
                  MS. RUTHANN RUDEL:  Can I just --
                  DR. KENNETH PORTIER:  Ms. Rudel.
                  MS. RUTHANN RUDEL:  This is Ruthann Rudel.  Thank you.  I thought that the write up of that section was very good, actually.  So I didn't have some of the issues that were brought up here with it.  And, on the developmental neurotox, I'm maybe only suggesting that in that section, there would be a sentence that says, there is not a modern developmental neurotox for this compound.  Just being clear about data gaps throughout, I think it's helpful.
                  DR. KENNETH PORTIER:  Dr. Vorhees, something you said about fitting the curve and dropping the highest dose, was that dropping justified?
                  DR. CHARLES VORHEES:  It wasn't me.
                  DR. KENNETH PORTIER:  Oh, Mark.
                  DR. MARK JOHNSON:  That was me, yeah.  Well, I mean, you do that to get a good fit for your model.  And sometimes, if you have a bell-shaped curve --
                  DR. KENNETH PORTIER:  Yeah, but the guidelines usually say that's not the excuse for dropping the highest dose.  The excuse is usually that's it's above minimum tolerated dose, and, if that wasn't the discussion, then I think we need to point that out.  Just because dropping that highest improves your fit is not an allowed excuse as far as at least from the cancer guidelines.  I don't know about the non-cancer kind of fits.  
                  But, if you read the cancer guidelines, it says usually the argument is that it exceeds some minimum tolerated dose and/or is producing an adverse effect that is clearly toxic, kind of an overwhelming toxicity, not just because it doesn't follow the nice curve.  I know because I just looked at this recently.
                  DR. MARK JOHNSON:  My primary concern has to do with the variability and just three data points.
                  DR. KENNETH PORTIER:  But three data points is --
                  DR. MARK JOHNSON:  You can draw a line.
                  DR. KENNETH PORTIER:  That's usually what you have.  You can draw a line and then assess on nonlinearity.  Did you want to add something?
                  DR. STAN BARONE, JR.:  So part of the issue with the highest dose is the animals weren't getting pregnant, and so, again, they're different.  And so we need to be clearer about the rationale for dropping the highest dose.  
                  And the second aspect in the dose-response analysis, which I believe we captured, was we're interested in what's happening in the low-dose region and the shape of the curve in the low-dose region versus what's happening at the high-dose region, which may be a totally different mode of action.  So, again, we need to be more clear about the rational definitely from your read and your input.
                  DR. MARK JOHNSON:  I just also want to add, what gives me comfort is that reduction in litter size corresponds very well with those responses done with post implantation.  But they are sort of the same thing, aren't they?  They're related.  
                  DR. STAN BARONE, JR.:  The measures what we're looking at, they're related, but the data that are going into it aren't exactly the same thing so biological relationship?  Yes.  But the data aren't exactly the same, and what's reported we're actually looking at different reporting measures to do those calculations.
                  DR. KENNETH PORTIER:  Dr. Davies, your flag is still up?  No.
                  DR. KENNETH PORTIER:  Any additional comments from the panel?  Dr. Vorhees.
                  DR. CHARLES VORHEES:  Can I ask Dr. Barone a question.  So you say that the data aren't identical, but they're not orthogonal either?
                  DR. STAN BARONE, JR.:  True, and that is why we are reporting them separately.  They're not exactly different biologically.  They are related.
                  DR. KENNETH PORTIER:  But, in a weight of evidence, that just means it doesn't add one plus one.  It's like 1 plus 1.25 or something.  They go together.  I understand that.  I wasn't challenging -- I remember reading that section and coming to the same conclusion and saying, oh, that's an acceptable justification for dropping that dose.  I just wanted to make sure we were reading the same thing.
                  Any additional comments on this question?  Good.  Let's move on to 5.2.
                  DR. KATHERINE ANITOLE:  Please comment on the weight of evidence analysis for the choices of non-cancer endpoints for the acute and chronic risk scenarios.  Please provide additional data, data interpretation, or information that would have informed the weight of evidence analysis and selection of critical studies for PODs.
                  DR. KENNETH PORTIER:  Dr. Cory-Slechta, you have the lead.
                  DR. DEBORAH CORY-SLECHTA:  Okay.  I wasn't aware.  I thought we were -- these were together.  I think we already actually discussed this topic under 5.1.  We went through a lot of these considerations already.  I am not personally aware of any other data that would add to or modify this weight of evidence at this point.
                  DR. KENNETH PORTIER:  I tend to agree that Dr. Barton had a lot of weight of evidence discussion in his comments, but let's move on through the group.  Dr. Sheela, I think you're next.
                  DR. SHEELA SATHYANARAYANA:  Well, I thought that these sections were really well written, and, given the data that was there and for the studies that I reviewed, I thought it was appropriate.
                  DR. KENNETH PORTIER:  Dr. Willett.
                  DR. CATHERINE WILLETT:  I agree mostly, entirely.  I just wanted to reiterate what I said a little bit earlier which is it might be helpful to follow the weight of evidence decision-making process at EPA if there was -- this might be a lot of work but -- if there was an additional table that just had the studies and endpoints that were chosen and went through the argumentation for choosing them, or the high-quality studies and why one was chosen over the over, just with the argumentation listed out because it would just make it easier to understand.
                  DR. KENNETH PORTIER:  So that's kind of a presentation recommendation because I think those arguments are in the document.  You're just saying kind of consolidate them into some kind of table that facilitates that, and that's a good recommendation.  Dr. Johnson.
                  DR. MARK JOHNSON:  Yeah, I would like to echo Catherine's comment for the same reason, possibly in Table 3.7 where we could just track it a little bit better.  Personally, I'd like to see the weight of evidence scores, and I think the scores are there as well too that help us understand the relative difference but actually why some were selected, and some were not.  
                  And this is my soapbox and I apologize to you if you've heard this before, but, statistically, 1 out of 20 times, we're going to be wrong.  And it just helps me to see those relative differences in the weight of evidence to be able to determine which is more likely to be a positive response than not.
                  DR. KENNETH PORTIER:  Mr. Kaufman.
                  MR. ALAN KAUFMAN:  Sorry.  Al Kaufman.  Nothing to add.  Thanks.
                  DR. KENNETH PORTIER:  Dr. Pessah?  Dr. Vorhees.
                  DR. CHARLES VORHEES:  Under weight of evidence, I would just say that I think more weight should be put on the Ichihara study because it's a longer exposure.  It's a 12-week exposure, and then they also measure the grip strength at various intervals at the end and various weeks afterward.  Whereas, the Honma study is only a three-week exposure with more limited sort of follow-up in the very small sample size.  So, in terms of weight of evidence considerations, I think the Ichihara study should be given greater weight than the Honma study.
                  DR. KENNETH PORTIER:  Anyone else on the panel want to weigh in on weight of evidence for non-cancer endpoints?  No?  Dr. Gilbert.
                  DR. KATHLEEN GILBERT:  Thank you.  Well, when we're talking about non-cancer endpoints, I have to ask why the immunotox was not considered.  They mention it several times that there was data showing immunotox and they talk a little bit about the Lee and Anderson papers, but then it sort of gets dropped.  So I was just curious why the EPA decided not to use that as an endpoint.
                  DR. KENNETH PORTIER:  So you're saying there's basically an incomplete conclusion on immuno in the discussion.
                  DR. KATHLEEN GILBERT:  Well, there were several of the other endpoints that were chosen that didn't have really more papers backing them up, so I was just curious as to why it was kind of dropped.
                  DR. KENNETH PORTIER:  Yeah, kind of goes to Dr. Willett's suggestion that being able to kind of see a bigger picture.  It might have been that the scores on the immuno papers were much weaker.  
                  I think Dr. Barone is searching through to see an answer for that.  While he's doing that, any additional comments?
                  DR. SHEELA SATHYANARAYANA:  This is Sheela Sathyanarayana.  I was just going to say that I think a lot of these comments go back to the systematic review and relates to some of the things that Holly said yesterday that, if you really trust your systematic review, then you wouldn't have these kinds of comments and issues coming up now.  And so not all of us could read all of the studies, you know.  I focused on the reproductive developmental tox and, of course, Kathleen's going to focus on immunotox.  
                  At some level for the readers, they have to trust that Data Quality Evaluation.  And I went through that document for my studies and I tended to agree, but I think that there's been comments for other studies where people did not agree.  And so, if you're going to go through that exercise of systematic review, it really needs to be done thoroughly.  And I know that that keeps being brought up again and again.  It's a very difficult process and I appreciate that, but I think that there's still some holes here.
                  DR. KENNETH PORTIER:  Dr. Barone.
                  DR. STAN BARONE, JR.:  So, if you look at the systematic review table and the point of departures, at least the NOAELs and LOAELs that are listed for the immune section, there are on the data quality evaluation quite a number of high-quality immune studies, at least according to our evaluation, in several mediums.  I think the issue that -- and this might be an issue of completeness in our description of what we choose for the point of departure -- in general, the point of departure for the immune effects is in the higher dose range.  There are a couple of studies that show points of departure in the lower dose range, and I'm talking about a hundred ppm range.  And the issue is those studies, whether they were suitable for dose response?  Were they modelable?  Those kinds of issues.  We need to be a little clearer about that.  
                  We are calling immune effects out as an important domain, and I think on the hazard ID, if you look at the table on page 361 of the assessment, you can see the whole sequence of immune studies.  So your comments are valid.  We'll look at that more closely and provide additional justification and look back at the modeling work because there was a lot of things that we tried to model, not everything was modelable and not everything could be carried forward for dose response, and we may have lost some explanation there.
                  DR. KENNETH PORTIER:  Dr. Davies.
                  DR. HOLLY DAVIES:  Just to go back to the Data Quality Evaluation, I did like the explanation.  I did think it was well written since we're talking about the weight of evidence.  But a lot of the studies in the weight of evidence didn't go through the data quality evaluation, so, if you look through -- if you say as I did -- go through and look at all of them and write down the references and which ones went for data quality evaluations and which ones didn't, you'll see that a lot of them weren't.  And so it wasn't -- so that's why as Sheela said, why we're all evaluating them ourselves, it's because we couldn't just rely on that document.  But it was a well-written document.
                  DR. KENNETH PORTIER:  Dr. Willett.
                  DR. CATHERINE WILLETT:  I just wanted to add a thought about the immunotox.  So I looked at the immunotox too and decided that they didn't use it because it was a higher.  But also there was a lot of high-quality studies in there.  And the other thing I wanted to point is those funny studies I found about the proteomic studies in humans.  Immunotox related proteins came up a lot, so that's mechanistic evidence that immune might be an important thing to consider, and those references are in the comments.  So that's just supporting evidence that immuno might be a good thing to reconsider.
                  DR. KENNETH PORTIER:  Dr. Anderson or Dr. Cory-Slechta, do you want to add anything?
                  DR. HENRY ANDERSON:  I don't have any.
                  DR. DEBORAH CORY-SLECHTA:  I don't have any.  Nothing to add.
                  DR. KENNETH PORTIER:  Anyone else?  Good?  Let's see.  Dr. Slechta, did you capture all of this for the minutes.
                  DR. DEBORAH CORY-SLECHTA:  I hope that I'll be back in touch with everyone.
                  DR. KENNETH PORTIER:  Good.  Let's go onto 5.3.
                  DR. KATHERINE ANITOLE:  While the majority of exposures are occurring via inhalation, and inhalation exposure is the most important, dermal exposures might be an important contributor to aggregate exposure.  Given the limited toxicological data available by the oral route, human equivalent dose or HEDs for dermal exposures were derived by extrapolating from the inhalation PODs.
                  Question 5.3.  Please comment on the assumptions, strengths, and weaknesses of this approach including using an inhalation study instead of a limited oral study for route-to-route extrapolation for determining dermal PODs in the non-cancer assessment.
                  DR. KENNETH PORTIER:  Dr. Gilbert.
                  DR. KATHLEEN GILBERT:  Okay.  I'm going to start with a couple of caveats: (A) I'm not a modeler, so I'm going to defer to my other discussants on a lot of these points; and (B) I just got this question two days ago, so I'm going to be mostly talking about what Dr. Bruckner sent in his notes because he was supposed to be the lead the discussant on this.
                  As we know, the reviewers of the 2016 Draft Risk Assessment recommended the inclusion of dermal exposure.  The EPA assessed the dermal exposure to workers using the dermal exposure to volatile liquids model.  This model estimates .16 dermal exposure for nonoccluded exposers based on a 2011 in vitro dermal penetration study.  
                  The report presents several occupation dermal exposure scenarios that take into account the potential for evaporation and glove use.  The potentially exposed workers were primarily confined to two companies that manufactured 1-BP in the U.S.  Dr. Bruckner agrees that dermal exposures may be an important contributor to systemic exposures to BOCs in the workplace and that an estimate of dermal exposure to 1-BP should be included.  The PBPK model of Garner et al. cannot be used for route-to-route extrapolation as it has not been validated with actual TK data and its model structure does not provide for skin exposure.  Unfortunately, there are apparently no published TK data from animal or human studies other than an IV experiment.  The approach described in Section 2.3.1.22 for calculating dermal exposures seems quite reasonable as does its assumption of .16 for the fraction absorbed.
                  Dermal absorption of 1-BP depends on the type and duration to the exposure.  Significant dermal exposure may occur in cases of occluded exposure, repeated contact, or dermal immersion.  Work activities with a high degree of splash potential may result in 1-BP being trapped inside the gloves.  It should also be noted that 1-BP easily travels through most glove materials with the exception of certain polyvinyl alcohol or multiple-layer laminates.  So one thing that was not clear to me is why the underlying EPA dermal model assumes only one exposure event per day which seems likely to underestimate exposure as workers come into contact with 1-BP several times during the day.  Based on these uncertainties, the EPA has expressed only a medium level of confidence in the access baseline exposure.  
                  My major concerns were why they only used one exposure per day and the information about the glove use being important.  That's all I have.
                  MS. TAMUE GIBSON:  Okay.  Thank you.  I'm the chair for a couple of minutes.  Okay.  Dr. Sheela, do have a response to that?
                  DR. DEBORAH CORY-SLECHTA:  I can't really comment on the calculations for the dermal exposures, and I'll leave that to my dermal colleagues to do.  But, in terms of the oral studies, for the two oral studies that are listed in the Data Quality Evaluation, I feel that they were evaluated appropriately.  One of them was listed as unacceptable because it didn't meet an OECD guideline, and I thought that was an appropriate determination for that specific study.  And I did feel that they were limited, so for that piece of the question, I agreed.  For the dermal piece, I'm just not the right person to address that.
                  MS. TAMUE GIBSON:  Okay.  Thank you.  Dr. Johnson.
                  DR. MARK JOHNSON:  Briefly, I agree.  I'm not a modeler either, but I think using inhalation data may likely be more accurate than using oral data considering first-pass hepatic transformation implications.  But you don't have a refined PBPK model, I agree.  
                  But some strength comes from comparing a human experience and human equivalent concentrations with the human experience data.  This is qualitatively done but not as a data integration step in the primary document.  So I would encourage the EPA to consider providing a data integration step within the hazard assessment section; integrating controlled animal data, human experience information, relevant mechanistic and in vitro, and potentially a read across for even in silico information helps to support the refined human equivalent concentrations.
                  DR. KENNETH PORTIER:  Dr. Vorhees.
                  DR. CHARLES VORHEES:  Yeah, I have nothing to add to this section.
                  DR. KENNETH PORTIER:  Dr. Pessah.
                  DR. ISAAC PESSAH:  I have nothing to add.
                  DR. KENNETH PORTIER:  Thank you.  Good to hear that you've joined us this morning.
                  DR. ISAAC PESSAH:  Yeah.  It's a little early here.
                  DR. KENNETH PORTIER:  Ms. Rudel.
                  MS. RUTHANN RUDEL:  I don't know if this is useful at all, but one of the intermediates, I guess, there's the bromopropanol like in the metabolite chains and the compound might be too different but NTP does have a dermal long-term study for 2,3-Dibromo-1-propenal, so it has an extra bromine.  Otherwise, it's pretty much the same.  So I don't know if anything that could be useful from a read across from that compound, but it is there.
                  DR. KATHLEEN GILBERT:  Can you send me that so I can put it in the notes?  Thanks.
                  DR. KENNETH PORTIER:  That was Dr. Gilbert.  Dr. Willett, did you want to comment?
                  DR. CATHERINE WILLETT:  No, I didn't realize I was on this part of this question.
                  DR. KENNETH PORTIER:  Put your flag down.
                  DR. CATHERINE WILLETT:  Oh.  Oh.
                  DR. KENNETH PORTIER:  Does anyone else want to chime in on this?  Dr. Gilbert, did you get what you needed from this?
                  DR. KATHLEEN GILBERT:  Well, I hope the other discussants will send me comments that they have.  Thanks.
                  DR. KENNETH PORTIER:  I'll turn to EPA; do you have any questions?  Nope.  Sorry, I missed part of that discussion.  Okay.  From my agenda, it seems that we have a 15-minute break scheduled at this point, and it's probably needed.  I have 10:17 so let's reconvene at 10 -- what will it be?  35, 10:35.
                  MS. TAMUE GIBSON:  Yes.
                  DR. KENNETH PORTIER:  Okay.  Take a break.
                  
                                    [BREAK]
                  
                  DR. KENNETH PORTIER:  Okay.  The panel is caffeinated and hydrated, and we're ready to keep going.  Are there any additional comments on 5.3 before we move on to Question 5.4?  Let's go ahead and move onto 5.4.
                  DR. KATHERINE ANITOLE:  Decreased live litter size, that is reduced number of live pups per litter, was the endpoint selected as most relevant for calculating risks associated with developmental toxicity following chronic exposures.  EPA selected the BMDS nested dichotomous model, or NCTR, for this developmental endpoint to account for intra-litter correlations and litter-specific covariates.  However, this model can only be applied to increases in effects, therefore, increases in post-implantation loss was the endpoint selected as the most relevant for calculating risks associated with the developmental toxicity following chronic exposures using nested modeling.  
                  Question 5.4. Please comment on the nested modeling approach and the selection of endpoints and whether the Draft Risk Evaluation has adequately described the use of this model.
                  DR. KENNETH PORTIER:  We've asked Dr. Gilbert to start.
                  DR. KATHLEEN GILBERT:  Okay.  Once again as a non-modeler, part of this was confusing to me.  They talk about the BMCL and it's my understanding that in terms of the 2019 Draft Risk Assessment, the POD derived from the WIL Research was described as a BMCL not a BMDL as exposure was inhalation, not ingestion.  And they talk about the POD being from -- increased post-implantation loss was a BMCL of 24 ppm that assuming that is just an editorial error.  
                  We need to talk about the nested modeling approach and its endpoints.  For dose-response analysis of developmental toxicity studies, data from individual litters are generally considered as an N of 1.  And that was certainly what I was taught when I started doing developmental toxicity.  This is because fetuses from the same dam are not statically independent.  Pups of the same litter tend to respond more alike than do pups from different litter due to similarity in genetics and environment.  However, in a nested study, the endpoint examination is conducted in a subset of subjects from each treated individual, the pups of the treated mother for example.  Because each mother may produce 15 to 20 pups.  If you have 20 treated mothers, you would have 20 observations from each mother.  Obviously, this generates an observational sample size that is larger than that of the originally treated animals.  Since all the observations are made in pups, but not in the mothers, these data should be considered nested data.
                  It says by using litter-specific data, intra-litter correlations can be accounted for through modeling methods as nested dichotomous method models that rely on a binomial model of variability.  Now, I still think that the N of one is the method that is most commonly used, so I would be interested in hearing from the other discussants what they think about using this nested approach.  And the rationale for using it not clear to me anyway.  So, in the Draft Risk Assessment, the supplemental information on human health benchmark dose modeling, it states the application of nested dichotomous models to these data was possible because the incidence data for post-implantation loss was available for every litter and preferable because they can account for intra-litter correlation and litter specific covariates.  At least as far as I'm concerned that needs to be clarified.  I didn't really understand what that meant.  
                  However, the endpoints of decreased live litter size and increased post-implantation loss to develop a POD for chronic exposures based on the developmental effects seems appropriate.  A BMR of five percent, which represents a combination of reproductive effects, when a BMR of ten percent would be used and developmental effects for post-implantation loss, which is considered a severe effect like mortality, when a BMR of one percent would seem appropriate.  So the PODs calculated for post-implantation loss using the nested dichotomous model were similar to those calculated for the related decreased live litter size when using BMD modeling.  So that seemed appropriate to me.  I would just like to have seen a better explanation for the rationale for using the nested modeling versus the more traditional N equals one for litters.  And that's all I have.
                  DR. KENNETH PORTIER:  Dr. Sheela.
                  DR. SHEELA SATHYANARAYANA:  So I think that, for my understanding, this was the nested model was done in response to reviews of the 2016 Draft Risk Assessment.  From my reading -- and I admit that I don't do a lot of nested models but -- in terms of doing nested models, it seemed appropriate and also the results did not seem to vary significantly from the non-nested models.  So I didn't find it to be of significance regardless.  So overall, I was okay with it.
                  DR. KENNETH PORTIER:  Dr. Johnson.
                  DR. MARK JOHNSON:  I don't have anything further to add.  Thank you. 
                  DR. KENNETH PORTIER:  Mr. Kaufman.
                  MR. ALAN KAUFMAN:  Hi.  Al Kaufman.  Yeah, I think it was adequately explained and described.  Beyond that, I really don't have any comment because it's sort of out of my area of expertise.  
                  DR. KENNETH PORTIER:  Dr. Pessah.
                  DR. ISAAC PESSAH:  I have nothing to add.
                  DR. KENNETH PORTIER:  Dr. Vorhees.
                  DR. CHARLES VORHEES:  Yeah, it seemed appropriate to me.
                  DR. KENNETH PORTIER:  So this is Ken Portier.  I probably made that recommendation back in 2016.  Yeah, he's shaking his head, yeah, you did.  I remember that.  And I can explain to you why it's slightly better.  They're not going to be that different: the nested incidence model versus the N equal one traditional proportional model.  It's going to be a slightly different estimate for underlying uncertainty, so the tests are going to be slightly different p-values.  But I think in this case, because of the way the data are, it's not going to change things that much.  
                  Statistically, what they did is the appropriate model, okay, but not the traditional model.  Anyone else want to add?  Gosh we're flying through these which is not what I -- I mean, this is what I expected.  Like you said, this is the second round, so this has been polished from the 2016 report.  I'll look to EPA; do you have any clarifying questions or comments?  No.  Then let's move on to Question 5.4 [sic].
                  DR. KATHERINE ANITOLE:  For the cancer risk assessment, EPA derived the inhalation unit risk, or IUR, based on lung tumors in female mice.  While the precise mechanisms/modes of action of 1-BP carcinogenesis are not clearly understood, overall, reasonably available information and the weight of evidence analysis for the cancer endpoint was sufficient to support a mutagenic mode of action for 1-BP carcinogenesis.
                  Question 5.5.  EPA concluded in the human health risk assessment that 1-BP carcinogenesis occurs through a mutagenic mode of action based on reasonably available information and the weight of evidence.  Please comment whether the cancer hazard assessment has adequately described the weight of evidence regarding the mutagenic mode of action.
                  DR. KENNETH PORTIER:  Dr. Eastmond.
                  DR. DAVID EASTMOND:  Okay.  I'll slow you down a bit.  So let me address this in a couple of ways.  First of all, I'd like to get in some technical details that are important because it weighs into sort of the weight of evidence, and then I'll get into my sort of evaluation of this.
                  So the first thing -- and I mentioned this briefly but -- there's a little bit of concern I had about the literature search.  And, essentially, the way that I do this -- and I assume EPA did this as well -- it's described as, when you start this off, you go to sort of standard lists of genetic toxicity testing or other reports and look at them.  And so one of the first things I did is I went to the ECHA documentation database, and there were significantly more studies in the ECHA database than there were in the EPA document.  So then I'm saying, how come they're missing these?  What's going on?  So then I start digging more into sort of what's happening?  For example, there are six different mutation studies in salmonella listed in the ECHA study, but only two really discussed in the EPA document and two briefly mentioned.  So I said, what about these other?  Why were they missed and et cetera?  And part of this is the EPA focused very, almost exclusively, on this difference between open system and closed systems and put their energy into that.  And I'll come back to that in just a minute.  
                  But I also did a couple of reviews looking at literature search, are new studies that have been published?  And there are three studies that, I think, are particularly relevant: Two of them which one describes the formation of DNA adducts in vitro and (inaudible) DNA.  And the other one actually reports the formation of DNA adducts, 1-bromopropane in vivo, these N7 guanine adducts.  We'll talk about that briefly in a minute.  
                  And then one recent one which is essentially a publication of one of the unpublished studies which is referred to in the report.  I'm pretty sure it's the Weinburg 2016 study, which has just recently been published and was mentioned, I believe, in the ACC comments; they call attention to that paper.  
                  In addition, in looking at sort of reviewing some of the data, there's a document PDF called BioReliance 2015 in the references.  The PDF has two studies in it: the first part of the study is the salmonella study, which is referred to; and then there's a second study on the end of it which is by Roy et al.  So the first one's a Wagner et al. as the study director of the second one is by Roy who just happens to have been a former postdoc of mine.  But anyway, it's a study which shows positive effects in human peripheral blood lymphocytes for 1-bromopropane for causing structural chromosomal aberrations.  And so that probably ought to be carried out because it refers to sort of a positive in vitro study.  So that's the first thing.
                  The second thing is that, in my opinion, the sort of mode of action decision is so important that the key studies involved in mode of action determination really need to go through this sort of data quality review and critique.  And so, when I started looking up one of them, one of the studies referred to that 1-bromopropane caused DNA adducts.  And so I said, I want to see this paper.  So I went and clicked and followed the link.  HERO doesn't work for me.  I'll just tell you.  I mean I get there, but I don't find what I want, but I was able to find this particular study.  And I'm looking at it as, is this in the study of DNA adducts?  And then I go through it, where'd they come from?  And buried in the discussion is a sentence saying it causes DNA adducts.  
                  And the reference it refers to is an abstract which was presented at the SOT three or four years earlier.  Now whether that would be considered sufficient evidence to base kind of key decision in this document, I'm not certain.  And it ends up being it's not critically important because these are new papers and they'll replace that, provided information for it.  But it triggered in me that maybe some of these studies weren't as evaluated as thorough as they should be, at least the papers themselves.  
                  And then one of the other ones, one by Kim et al., the link is a bad link.  It doesn't refer to any -- the study it links to doesn't describe anything about genotoxicity.  So there's just a few of those sorts of things that flag up for me.  
                  Now there's a discussion that plays a heavy influence in this determination as a study that took place in 1981 where 1-bromopropane was, along with a number of other halogenoalkanes, were studied in the AMES test in a closed system and an open system.  And they were non-mutagenic in the open system, but when you carefully close the system to prevent evaporation, then they saw a significant increase in mutation.  And this is with and without metabolic activation which indicates it's probably a direct-acting agent.  
                  So the shift has been basically to sort of disregard in and the open system studies and focus only on those with closed systems.  I'm not sure that's totally warranted because, for example, the studies that were done by the NTP did as part of their cancer bioassay, they tested 1-bromopropane all the way till toxicity, so at ten thousand micrograms per plate, and it was totally toxic; it killed all the bacteria.  And so those studies should have been valid at the lower concentrations because you're seeing biological effects of the compound, and those studies were particularly negative.  
                  So the pattern is a little more messy, and you kind of have to look at each of these studies and look at them individually.  So, in addition, there's a comet assay which was positive in an open system.  There's this structural chromosomal aberration assay which is positive in an open system, and I believe the mouse lymphoma assay was positive in an open system, which I can get the original study too.  But it appears that was the case.  
                  In essence, once you break down this, there's quite a bit.  There's mixed results seen in the in vitro test, but there are quite a few positives in in vitro.  And so that's one of the reasons for this focus on the actual mutagenic mode of action.  There's other arguments as well.
                  The contrast on this -- and I'll jump to that -- is that it's negative in virtually every in vivo study.  So you've got -- originally there are two dominant lethal mutation assays.  It was negative in both of those.  There are three micronucleus assays: two in, basically, in mouse and one in the rat, negative in those.  And then there are two fairly recent studies which are looking for mutations in transgenes in transgenic mice and basically it was negative.  So these were looking at change in mutations in the lung, which is one of the target organs; in the colon, which is another target organ; and the liver.  Actually, two of them did the first two tissues, and one of them did one tissue.  
                  So you've got kind of a very unusual situation where your in vitros giving you lots of positives.  The class of compounds is one that you think would be mutagenic.  The sort of the profile of tumors that are seen would suggest maybe it's muted; it's a mutagenic carcinogen.  It tends to be that mutagenic carcinogens tend to be active in multiple species and multiple target tissues, and so that's just kind of crude estimate on this.  But, in fact, the data here isn't very supportive of that, so it becomes a difficult sort of call.  
                  One of the things they did was they did review in detail the Data Quality Evaluation of these two transgenic mouse studies for mutation and sort of a mix -- they came through this sort of formal evaluation as high quality.  They're both high quality, and one of them was moved to moderate quality for someone's professional judgment, which I can't quite understand why.  But, anyway, then they went through a very detailed -- and what I would consider being an extremely critical evaluation of these studies with expectations that are outside of a normal realm of an evaluation.  
                  So, for example, they criticized them because they had not measured respiratory rates, or they had not measured body temperature on these animals during the 28-day studies.  I have never seen a genotox study that has ever done that.  There was a series of things, and they criticized one of them because it wasn't done -- they treated five days a week rather than seven days a week, which the guidelines recommendation seven days a week for 28 days.  Well, the study was designed to mimic the MTB bioassay, which was given five days a week.  So I would not dismiss this study because of these sorts of reasons.  The bottom line is I'm sure you can find some criticism with any study, but they looked to be reasonably sort of very acceptable quality studies.  
                  So you're kind of looking at a very strange situation in here where you're -- usually you think of your in vitro study as sort of screening studies, then you follow up with in vivo studies for confirmation.  But, in fact, the follow-up studies in confirmation aren't confirming directly what you're seeing in in vitro studies, but knowing how this chemical works, knowing this sort of pattern of tumors that are caused and, as I've said, I might mention that all the tumors appear to be site of action.  They're direct site of exposure or site of absorption, so you're looking at tumors in the lungs of the female mice.  You're looking at tumors in the large intestine of the rats, and the other tumor was on the skin, so you're looking at skin tumors.  So, again, that was maybe consistent with sort of a direct-acting agent where you're having effects at the site.
                  So, for me, it comes down to the -- this is really a judgment call.  I mean, there's one or two other errors I'll pass on to some things, little mistakes here and there.  I think if I've got anything else, I need to -- oh, well, while I've got the stand, I might as well comment -- a little bit of a thing.
                  For me, there's some sort of consistent ways of citing studies and particularly in published studies where you use the study director's name, the organization that published it, and then the sponsor is given there.  In some cases, the study is attributed to the sponsor.  In other cases, it's to the organization that did the study, and sometimes it's the study director, study author.  So it becomes really confusing when you're comparing across this document with like the ECHA document versus other ones because they all have different study names.  You have to try and say, okay, which study is this?  Fortunately, there aren't so many of them you can kind of piece it together.  But it would be useful, for me, just their sort of a study.  For me, it's always valuable to go to the primary study and look at it and verify.  Now there may be some rare occasions where you can't do that, but then you should have the reference and then say, "as cited by" where you got the information.  Because, well, I think what happened with the DNA adduct thing, I don't think anyone went to the original study.  I think they were using "as cited by" sort of thing.
                  So there's some other things about metabolites.  1-Bromopropane is metabolized to a whole range of metabolites.  Some of those are mutagenic, but some of them are sort of putative metabolites and never have recovered.  They're sort of thought to be in the metabolites.  And so I'm not sure they play a role.  Some will be because they are very minor metabolite, presumably.  
                  The other thing I would recommend is if you're going to talk about these metabolites as being important, they should be in the metabolic pathways because they're glycidol and the propylene oxide aren't in your metabolic pathways.  They're in the IARC pathway, which might be useful to adapt, but because these show up numerous times in the document, it'll be helpful to have them there where you can see them.
                  So, as I indicated, I was involved with the World Health Organization, while they're an actual program for chemical safety, and we did an update in their mutagenic testing scheme.  And, if you plug this into the mutagenic testing screen where you screen in vitro and you follow up in vivo, this would come out as being non-mutagenic.  But, given all this other information we know, that's independent of animal cancer data.  Once you now know what causes cancer and you know it's in vocal tissues in this sort of pattern, it makes this really a judgment call.  What do you call it?  And my sort of inclination on this is rather than -- and I've been vacillating and agonizing on this, and I think it's something the EPA will probably discuss some more.  The one in vivo study I mentioned, you now have in vivo DNA adduct evidence, which wasn't there before, but the adducts are N7-guanine adducts which aren't closely related to mutations.  They tend to -- they're clearly associated with exposure.  They're indirectly associated with mutations, so it helps but it also complicates things a little bit.  
                  Anyway, so that my sort of inclination is, even if you say, there's not a viable alternative mode of action in my opinion for the other ones, there's some other ones that are possibles.  There's a possibility of cell proliferation, although the evidence for that I found to be very, very limited.  There's some inflammation, so potentially immunomodulatory effects, but I don't know it's that convincing.  There's also reactive oxygen species was bringing oxidative stress.  But, for me, lots of toxicants cause oxidative stress, and they're not necessarily carcinogenic.  So it contributes, but I'm not sure it's a driving mechanism in many cases.
                  So maybe you've been through these discussions, Stan, but, I mean, my might take on this is probably better.  I would lean toward saying the mechanism is not known which I've already stated.  And in the EPA, I know part of it defaults to a linear extrapolation, so essentially you end up with the same outcome, but it doesn't put as much -- you're putting a lot of weight on in vitro studies and all the in vivo studies and genotox really don't support that.  So, for me, I think, the unknown is an easier response than a clear mutagenic mode of action.  But I will say, if you go back and look at IARC, basically, the evidence for genotoxicity is moderate.  If you go to ECHA, they think it's non-mutagenic.  If you go to NTP and the report on carcinogens, they think it is a reasonably anticipated carcinogen, and I think it's through a mutagenic mode of action.  So, obviously, informed scientists can disagree on this opinion and you may disagree with me, but that's sort of my assessment of it.  Any questions?  I'm open to --
                  DR. KENNETH PORTIER:  Let's see if Dr. Willett agrees or disagrees.
                  DR. CATHERINE WILLETT:  Well, I don't have much to add to all of that detailed analysis.  I was thinking more of whether EPA had substantiated their decision-making process, and I think they've clearly articulated why they went one way versus the other.  And I think in the interest in being protective that if there's significant doubt about the mode of action and it could possibly include a mutagenic element to it, that it's best to err on the side of including the mutagenic possibility.  So that was sort of where I came down on it.  
                  But I agree there's a lot of uncertainty and a lot of conflicting data.  It doesn't bother me as much because there's a lot of conflicting data everywhere where you have to make really complicated choices.  So I think erring on the side of protectivity, that was a good choice.  But given the fact that that was this sort of weight of evidence decision, I do also agree with the comments from EDF and Earthjustice that maybe that indicates a need to consider the consequence of acute exposure in terms of carcinogenicity potential.
                  DR. KENNETH PORTIER:  Dr. Johnson.  Mark, did you want to add anything.
                  DR. MARK JOHNSON:  I have nothing to add.  Thanks.
                  DR. KENNETH PORTIER:  Mr. Kaufman.
                  MR. ALAN KAUFMAN:  Yeah, I similarly looked at the data and, when you look at the structure, you would sort of expect that you might see a mutagenic mode of action.  On the other hand, the data's sort of conflicting.  It doesn't point to one direction or another very strongly.  However, I think you probably made the best call that you could which is to err on the side of a mutagenic mode of action and just going in that direction, so I'm comfortable.  I think it would just -- you may want to beef up the section where you talk about some of the uncertainties.
                  DR. KENNETH PORTIER:  Dr. Pessah.
                  DR. ISAAC PESSAH:  One of the things that hasn't been said which -- and, again, I'm not a cancer biologist or toxicologist -- is that along with this sort of discussion about DNA adducts, nobody has really mentioned the data that seems to be consistent across studies about a change in redox balance in various tissues.  Both immune tissues, hemopoietic tissues, and in this most recent study that shows glutathione adducts, again, exposure is intraperitoneal, but the doses, I thought, were reasonable certainly in line with what's been done and considered by EPA.  
                  So carcinogenesis isn't just one hit, one mechanism, and so why aren't we integrating it or why isn't EPA integrating these two what I -- are sort of consistent observations across studies which is that you're challenging the antioxidant and defense mechanisms.  You're reducing glutathione.  You're producing increased ROS.  And, in addition, there is some evidence that you're producing DNA adducts.  It may not be as strong to identify a mode of action, but it seems like you have to look at these things in an integrated way, but that's just something that came to mind as I was reading the literature.  
                  DR. KENNETH PORTIER:  Thank you.  Dr. Vorhees.
                  DR. CHARLES VORHEES:  No, I have nothing to add.
                  DR. KENNETH PORTIER:  Does anyone else want to comment?
                  DR. CHARLES BARTON:  I do.
                  MS. TAMUE GIBSON:  It's Dr. Barton.
                  DR. KENNETH PORTIER:  Who is it?  Oh, Dr. Barton.
                  DR. CHARLES BARTON:  Thank you.  The cancer hazard assessment has not adequately described the weight of evidence regarding the mutagenic mechanism of action.  Also, not all available studies were viewed in the draft document.  There's been seven in vitro studies conducted to examine mutagenicity.  One Huang's assay was positive, and five Huang's assays were negative including two from the NIH.  An in vitro mouse lymphoma study was positive.  
                  When in vitro data are equivocal, in vivo studies are usually relied upon.  Six in vivo studies have been conducted.  Three in vivo mouse micronuclei studies, including one from the NIH, were negative.  Two in vivo dominant lethal assays were also negative and one in vivo transgenic rodent somatic cell mutation assay, aka Big Blue, was negative.  Thus, two in vitro studies were positive, but five in vitro and six in vivo studies were negative.  All studies need to be discussed.  While some studies were not used for evaluation need to be discussed for transparency and avoid the perception of cherry-picking data.  The weight of evidence from mutagenicity appears to be negative.  As written, it appears the mutagenicity mechanism of action was concluded, and it could be interpreted that it was included and then studies picked to support that conclusion and that needs to be avoided.  
                  For example, chromosomal aberrations in exposed Chinese workers were discussed to support a mutagenic mechanism of action while the study found the non-statistically significant correlation and exposure had a small number of subjects and lack of a control group, not just a nonsignificant trend from high exposures to medium exposures.  
                  As previously mentioned, even secondary references to abstracts were used to support mutagenicity while well-conducted studies of high quality demonstrating that it was not mutagenic were ignored.  Several papers have concluded that 1-BP has a non-mutagenic mechanism of action for carcinogenicity such as Stelljes (2019) and Rossman and Duel (2002).  This would mean that there's a threshold for carcinogenicity.  The weight of evidence suggests that the mechanism of action for carcinogenicity is a nonlinear threshold model.  This should be thoroughly explored and discussed.  
                  Furthermore, a linear no-threshold one molecule one-hit model of carcinogenicity would warrant a cancer risk assessment for acute exposures.  This would suggest that most fruits and vegetables, as well as grain products, cause unreasonable cancer risk since these contain known carcinogens.  These natural things we eat contain known carcinogens as explored by the carcinogenic potency database at UC Berkeley where 55 percent of the carcinogens are naturally occurring.  This includes constituent natural chemicals in coffee and pastry which many here today had for breakfast.  Thank you. 
                  DR. KENNETH PORTIER:  I'm looking at Dr. Eastmond to see if he wanted to comment on this.
                  DR. DAVID EASTMOND:  The only real substantive comment -- there are actually two transgenic mutation studies, both negative.  So the in vivo studies are two negative in vivo studies.  Basically, it's a mixed pattern as I indicated.  The in vitro gives -- there's multiple positive in vitro studies, and there's negative in vitro studies, and they have their limitations.  The in vivo studies -- a number of them -- so that the dominant lethal studies usually aren't very sensitive for picking up mutational effects; point mutation is what we're talking about.  So there's something useful there.  
                  The micronucleus studies for the most part, if you're looking at point mutations caused by AMES test, they may or may not be picked up in a micronucleus study, so you wouldn't expect it.  And that's why the transgenic mutation studies become important, but they're not perfect.  They said the in vivo stuff, if you look at it, you come across, it looks pretty negative.  But you do have all this other supportive evidence that makes it difficult to want to dismiss it totally, and that's why I kind of went with the unknown mode of action that still triggers your sort of linear approach.  But it's a little more conservative in how you interpret all these negatives because you do have multiple negative studies.  
                  DR. KENNETH PORTIER:  Dr. Blystone.
                  DR. SHERI BLYSTONE:  As the non-toxicologist in the room, I like that approach.  Given everything that I've seen and heard from all of you guys that do know this stuff, to unequivocally say it's a mutagenic mode of action is not appropriate but being protective and using a linear model seems appropriate based on the equivalent, so I like Dr. Eastmond's approach.
                  DR. KENNETH PORTIER:  Dr. Johnson.
                  DR. MARK JOHNSON:  Yeah, I have a quick question for Dr. Eastmond.  Does not assays like AMES and mouse lymphoma have a relatively high false-positive rate?
                  DR. DAVID EASTMOND:  Actually, the reason AMES is so well liked is it has a very low false-positive rate.  So its sensitivity in your sort of classic is about 55 percent; its specificity is like 85 percent which means it has very few false negatives.  So that's one of the reasons, like, it doesn't flag too often for false positives.  So, when you have an AMES test that's positive, usually your -- it's more likely than not that this will be carcinogenic.  
                  A lot of the other -- the AMES test is not as prone to the false positives which happen with a lot of the other in vitro assays.  So, like the comet assay, the chromosomal aberration assay, the micronucleus assay, they tend to have a higher false-positive rate, and that's heavily influenced by toxicity.  Because, as you go to high concentrations, DNA strand breakage is a consequence of cell death, and so that can be picked up in these assays.  And so you're always -- cytotoxicity causes DNA breaks but DNA breaks cause cytotoxicity, so in this circle, where are you at in the circle?  And so it's sort of judgment call.  The AMES test usually isn't affected by that.  
                  The trick on it is, from my take, is the two tester strains which were positive -- and this is brought out by the NTP, and I think it was mentioned in the EPA document -- is that they have, I guess, they have endogenous glutathione transferase activity, so they think it may be this glutathione mechanism, which is involved, that may have contributed to the positives in those but who knows.  
                  The other thing I was going to mention is in the scheme of things, this is mentioned in the document is that these alcoholates, I guess, are thought to be more of the sort of the soft electrophiles rather than hard electrophiles.  And so they tend to be more targeting the soft nucleophiles like the sulfhydryl groups, and that's why you get a whole lot of these reactions and the adducts and the biomarkers.  
                  As evidenced in this most recent paper -- and it really ought to be evaluated for quality control because it's one of these adduct papers I don't like because they don't do their adducts in the controls -- they mention the controls are negative, and they show an example.  But they do have adducts in their low dose and their high dose, and their high dose is so much higher than their low dose that you can say, even if the low dose is controlled, you still have an increase.  But the EPA will need to look through that for quality assurance.  But that's in evidence of formation of DNA adducts in vivo which is an important piece of evidence at least.
                  DR. KENNETH PORTIER:  Dr. Blando.
                  DR. JAMES BLANDO:  This is Jim Blando.  I just wanted to support what Dr. Blystone said.  It sounds like, to me, that there's a fair bit of uncertainty about some of this, and I think in light of that, it makes sense to use a linear extrapolation.
                  DR. KENNETH PORTIER:  Dr. Barton.
                  DR. CHARLES BARTON:  Yes, using a linear no-threshold model, just because it is more conservative instead of relying upon the best available science is not appropriate for EPA in my opinion.  Thank you. 
                  DR. KENNETH PORTIER:  Ms. Rudel.
                  MS. RUTHANN RUDEL:  This is Ruthann Rudel.  I'm just curious, Dr. Eastmond, what you think about the Toraason paper so the comet assay damage in leucocytes in exposed workers, and how does that sort of relate?
                  DR. DAVID EASTMOND:  Actually, it was pretty negative.  The difference was pre-exposure/ post-exposure in the low dose, and nothing seen in the high dose of the workers.  The Toraason has the in vitro comet assay data and that looks to be real.  That was the DNA strand breakage.  But my take is the Toraason one has, at most, sort of equivocal suggested but this was a pre and post and the high dose -- I think the high dose, the post-exposure was lower than the pre-exposure, so it's certainly wasn't consistent with the clastogenic effect as I thought about that study.  Somebody's who more familiar with it can weight in but that was -- there's some suggestive stuff but we have effects at the low dose and the comet assay and not at the high dose and it's pre and post.  The comet assay's fairly prone to artifacts that you have to be cautious in interpretation of comet assay data.
                  DR. KENNETH PORTIER:  I see a number of flags up.  I just wanted to make sure they're still up for questions.  Dr. Blystone.
                  DR. SHERI BLYSTONE:  Did you want to go first?
                  DR. JAMES BLANDO:  Sure, I was just going to say that Dr. Barton mentioned the best available science, and it sounds like, to me, the best available science produced uncertainty.  So, from a public health perspective since our role is to be concerned about public health is the best available science has produced uncertainty and, therefore, we should be conservative linear in that threshold.
                  DR. SHERI BLYSTONE:  And I was echoing that same thing, best available science.  There's good quality studies that have conflicting data, so you can't say one is better than the other.
                  DR. CHARLES BARTON:  Are we mandated?  I have a question.  Are we being requested to determine what's best for public health?  Or are we being requested on, is there a mutagenic mechanism of action?  I don't think we're being asked.
                  DR. KENNETH PORTIER:  I think the question is on the mutagenic mechanism of action.
                  DR. CHARLES BARTON:  Right and not policy on what is best on public health.
                  DR. KENNETH PORTIER:  You're right.
                  DR. CHARLES BARTON:  Okay.  Thank you.  
                  DR. KENNETH PORTIER:  The question of threshold model versus linear model is one that panels -- I've lost count of how many panels we've had that discussion on and some of it becomes an EPA policy.  What's the default when you say unknown mechanism of action?  And Charles, you're aware of this as well as the rest of you are.  Sheri, did you want to say something else?
                  DR. SHERI BLYSTONE:  Just a fair point, Dr. Barton, but I think again, from everything that I'm hearing, it's not really appropriate to conclude a mutagenic mode of action based on the best available science.
                  DR. KENNETH PORTIER:  Dr. Davies.
                  DR. HOLLY DAVIES:  Since EPA's current practice is to default to the linear when it's unclear, I was -- it's a question for EPA -- about the update on what they're doing for that because I thought they were updating their guidance on cancer and how they're dealing with that.  And I was wondering if those updates were coming any time soon if there's going to be a change to that?
                  DR. KENNETH PORTIER:  I know that the Science Advisory Board recently provided a whole bunch of comments on the cancer guidance, and I, myself, put in some comments on mode of action and then modeling.  Dr. Barone.
                  DR. STAN BARONE, JR.:  There's no answer yet.  That particular issue, the guidance, there are discussions about the guidance.  That's a process and, again, we'll take public comment, and there will be peer review by the SAB.  So we're operating under the assumptions of the current 2005 guidance for cancer evaluation including the mode of action analysis.  
                  So, while I've got the mic open, I do want to -- and this is sort of a clarification -- to the degree possible that the committee can provide additional data or evidence or emphasis to the interpretation for the mode of action analysis, we would find that very valuable for a revision or our mode of action analysis.  
                  In our mode of action analysis, we did, as Dr. Eastmond also pointed out, followed the mode of action framework.  We did consider the evidence for urinary metabolites actual reactive species and epoxide in mouse and in rodent urine as another point of evidence for our mode of action analysis for a mutagenic mode of action.  So, again, by inference and read-across, that was a significant part of our determination.  But, for alternative modes of action, again, we are trying to describe alternative modes of action to the degree possible that we can.  If there's any additional evidence that you've found in your reading and in your review that we should incorporate, please let us know.
                  DR. KENNETH PORTIER:  I'm just letting it sink in here and see if something floats to the top.  I'm watching Dr. Eastmond.  I can see the neurons flying, but I'm not sure anything's going to come out.
                  DR. DAVID EASTMOND:  That's okay.  I was just curious.  The challenge for me, if I recall, and I'm not so familiar with glycidol, but the propylene oxide, generally, I put that in my brain as being a pretty high dose sort of carcinogens site of exposure.  So you go through the metabolic pathways, propylene oxide's got to be a pretty minor metabolite within 1-bromopropane.  So that's why I thought I can't see how this would be -- I mean, I haven't compared the potency, and that's the critical thing.  And maybe you've done that, but for me, I thought, well, this is relevant that there are mutagenic metabolites formed.  It seems to me that they aren't formed at such a level that we would expect to see much of an effect there, but that's just the kind of a seat of the pants guess.
                  DR. STAN BARONE, JR.:  So, in our metabolism section, we did look at that.  It's in the appendix page 338.  We did not include any kind of potency evaluation, and, if you think that might be helpful, we can -- if that's a recommendation, we can, again, look at that.  We didn't do that because we don't have a PBPK model, and that's, again, one of the drawbacks in the tools that we had for this evaluation.
                  DR. DAVID EASTMOND:  I mean, for me, I don't think it's critical as you clearly have tumors, and even if you can work it out, you've got multiple metabolites.  You don't know which one's going to be contributing specifically, so I wouldn't spend a lot of time on it.  That was just sort of my impression.
                  DR. KENNETH PORTIER:  Dr. Pessah, you wanted to comment?
                  DR. ISAAC PESSAH:  Well, again, I just want to ask whether if the data that shows the lowering of glutathione due to formation of glutathione adducts and changes in some of the antioxidant defense enzymes is at all considered as part of an indirect sort of susceptibility or factor in a cancer modeling?  Again, I'm not an expert in this, but it seems like no one's really talked about that.
                  DR. KENNETH PORTIER:  Dr. Barone.
                  DR. STAN BARONE, JR.:  We can add additional text, but the glutathione depletion is a mechanism mode of action that is talked about for both the cancer and the non-cancer.  And that may be more prominent for the peripheral neuropathies and the non-cancer effects for the neurological deficits, and we speak to that in the assessment.  So we can definitely bolster that hypothesis, and there is evidence to support that hypothesis as a mode.  Whether it's definitive for the cancers, we are very clear that it's not definitive for the cancer.
                  DR. KENNETH PORTIER:  Dr. Pessah, was that an adequate -- do you want to follow up?
                  DR. ISAAC PESSAH:  Yes, just that I'm not sure how you would determine that it's not an important factor in the cancer risk evaluation in the sense that there's quite a bit of literature now, not on 1-BP, but in several other areas of toxicology as well as just cancer biology that the depletion of antioxidant defenses is really a factor.  It's been considered a peripheral factor but, nevertheless, as someone mentioned, quite a number of fruits and vegetables we eat are, in fact, contributing to the depletion of antioxidant defenses.  And here's now a solvent as antibiotic which is contributing to that, and I don't see that, again, from a lay perspective being incorporated into an overall risk model or integrated into another risk model.  The answer that Dr. Barone said satisfies me for now.
                  DR. KENNETH PORTIER:  Dr. Johnson.
                  DR. MARK JOHNSON:  Yeah, I have a question for my colleagues too, and maybe Dr. Barone as well.  What kind of bothers me about some of the in vitro mutagenic assays is they didn't really consider, I guess, the spatial differences within the liver.  And, I guess, it's my understanding that the glutathione processes occur first.  Then, they're depleted then you move onto the CYP2E processes where you get the oxidative metabolites where you produced more CO2.  And, I guess, then you have liberation of bromine through the inhalation pathway.  But the Geiger et al. study suggests just the opposite, doesn't it?  It suggests that you get the oxidative metabolites first and then you get the glutathiones.  So it contradicts it in my mind and maybe I just don't know enough about it.  Maybe Dr. Pessah could add a little bit more to that.  So I'm curious in why, I guess, glutathione depletion isn't a more important thing or at least the mid-primary metabolite for low exposures.
                  DR. KENNETH PORTIER:  Dr. Pessah.
                  DR. ISAAC PESSAH:  Well, that's a question that came to my mind.  That it seems that, at lower doses, glutathione is impacted, and I can't tell you why that would be.  But it just seems like a pattern that I picked up from reading.  And I didn't get a real sense that that was a major input into the modeling that they were kind of done separately.  Well, you had the classic mutagenesis assays in the model, but the glutathione and, in general, the antioxidant defense wasn't really -- was that an input into the modeling?  Or was it just considered and then put aside and not a factor?
                  DR. KENNETH PORTIER:  Dr. Eastmond.
                  DR. DAVID EASTMOND:  Let me try to address a couple of things.  1-Bromopropane will react directly with glutathione nonenzymatically, so you're going to see a reduction in glutathione levels, so that would support an oxidative stress mechanism.  As far as zonation in the liver, that is generally the case for things via oral route of exposure where you're going through the hepatic portal vein and then you go in.  But, in this case, we're looking at inhalation or effects at site of exposure.  
                  Because of the direct reaction with the compound with glutathione to form a conjugate, it doesn't surprise me at all you'd see glutathione levels drop, and it's very likely that oxidative stress or additional sorts of things because this taxes the pentose phosphate shunt and other things.  There's going to be other physiological changes that will take place, and I would agree it very likely could be involved in the neurotoxicity that Dr. Barone mentioned.  
                  I mean, I think it is happening, but there are a lot of chemicals.  In fact, the very reason we have glutathione is to protect us against many, many different types of chemicals: Some are carcinogenic.  Some are noncarcinogenic.  So Tylenol, acetaminophen is sort of classic examples.  When you get to higher doses, it starts conjugating with glutathione.  That doesn't necessarily mean that that compound is carcinogenic.  It's activated as something that reacts with glutathione.
                  DR. KENNETH PORTIER:  Dr. Barone.
                  DR. STAN BARONE, JR.:  Just one clarification for the committee, just to make this hopefully clearer, we're doing our best to qualitatively describe the mode of action and the evidence for a mode of action.  The evidence for mode of action, or lack thereof, informs our choice of dose-response models.  We are not doing dose response on glutathione conjugation.  We're doing dose response on the tumor incidence data.  So there's a distinction there.  Until we have a full quantitative AOP, it's very difficult for us and, even as much as we know about cancer biology, we are not to a point where we have a dose-response model or a quantitative AOP that incorporates multiple mechanisms of action which many of these chemicals have into a dose-response tumor model.  
                  That's where we are so I just want to be clear what we have and what we don't have, and the more evidence or more that you can provide critical input to the mode of action qualitative analysis, the better our assessment will be at informing the uncertainty for our cancer evaluation.
                  DR. KENNETH PORTIER:  Dr. Schlenk.
                  DR. DANIEL SCHLENK:  This is kind of just to throw out.  Glutathiones in really, really high concentrations in most tissues, particularly the liver, maybe in some of the other tissues, it's not in such high concentrations like the central nervous system which is probably qualitatively why you actually see those effects.  But, from a carcinogenicity perspective, to get to oxidative stress, the depletion that you need for glutathione's got to be fairly extreme because you also have other oxidative stress mechanisms that protect.  That glutathione tends to be the primary one, but there are other mechanisms that protect.  So the dose has to be extremely high, I think, to get to an oxidative stress capacity to where that makes a difference at least in the liver anyway.  But maybe some of the other tissues, it's a little -- it's possible, but I don't see it as being a primary mode of action at least for the carcinogenicity.
                  DR. KENNETH PORTIER:  So Dr. Eastmond, as we write this up -- I mean, I think this has been an excellent discussion of weight of evidence and how you're kind of weighing all these things, and Dr. Barton's thrown in a little cite issue that we probably should acknowledge as well.  But you think you've got enough to -- once you Dr. Barton's notes and maybe Dr. Johnson's comment.
                  DR. DAVID EASTMOND:  Certainly, I mean, I've written a bunch of stuff, and I can add to it and try to reflect sort of the discussion and the considerations and certainly put Martin's comments as well there.  Certainly, we would reflect the valid point of view.  
                  One of the things is the challenge, as you know, is that scientists can differ on the interpretation of some of these things.  This is one of these areas where there's mixed data, and it's reasonable to interpret it differently in how you do it, and decisions you make in the face of uncertainty, that's usually not a scientific thing.  It's sort of comes back to sort of core values.  Anyway, we'll see what we can do on that.
                  DR. KENNETH PORTIER:   So I like the way you kind of laid out this signal for a mutagenic mode of action in your discussion, and, as I was listening to Dr. Barton, I was thinking, what's the signal for a threshold kind of model, and I guess I don't know that one as well.
                  DR. DAVID EASTMOND:  I mean, certainly, you can lay out the signal.  There's evidence pro and con, and my sort of bottom line was evidence in favor tends to be all in vitro, and the evidence opposed is in vivo and that usually trumps the in vitro.  One's a screening and one's the sort of definitive test.  But given there's all this peripheral evidence around that makes you want to be cautious because structural analogs cause cancer.  You've got no reactivity of the chemical.  Now we know it causes adducts in vivo.  So it's a mixed bag.  
                  So, for me, I go back to saying, if I'm going to do this in a systematic sort of approach, I might end up with saying, we just don't know the mode of action which then would trigger the same sort of outcome of linear dose response, so they don't have to change everything basically.  But, for me, it's sort of a fair thing.  You can go either way.  There's arguments pro and con on it.  And I'm sure they'll have additional discussions on this as well at least with some of the new information.  But we'll give them our point of view.
                  DR. KENNETH PORTIER:  EPA, does that -- any follow-up?
                  DR. STAN BARONE, JR.:  We appreciate this robust discussion and the divergence of views.  Hopefully, that will be captured and the evidence to support the divergence of views on this particular topic.  Again, as Dr. Eastmond's pointed out, and we did too, there are similar views to our assessment by other assessment bodies and there are differences.
                  DR. KENNETH PORTIER:  Okay.  On my agenda, I had that we were going to break for lunch around 11:45.  I have 11:40.  We're way ahead of our agenda, so I'm suggesting that we meet back from lunch at 1:00.  It gives you an hour and twenty minutes for lunch, a little bit more relaxed lunch than we normally have.  I'm informed by members of the panel that we're not so unprepared, but we'd like the time this evening to fully prepare for Question 7.  So we're going to complete Question 5 and Question 6 this afternoon, and then, when we're done, we're done for the day.  And tomorrow we'll take Question 7 and 8.  So that's kind of thinking about you prepare your time and effort.  I'll see you guys at 1:00.  Thank you.  
                  
                                    (BREAK)

                  DR. TODD PETERSON:  Okay.  The public is showing up, so that's good.  Dr. Gilbert has asked that 5.6 and 5.7 be read in, and they're going to address it as a combined question.  We'll go ahead and read both of those.
                  DR. KATHERINE ANITOLE:  Question 5.6; typically, EPA uses the benchmark dose modeling software, BMDS, with a BMR of 10 percent and the models are restricted to multistage models or the broader suite of dichotomous models in BMDS and a single best model is chosen for the POD.  EPA used an alternative approach to calculate the cancer POD versus the standard approach of choosing the best fit model and to assess the impact of model uncertainty.  
                  Briefly, EPA used two model averaging approaches -- frequentist and Bayesian -- considering multiple benchmark dose models to calculate the POD at benchmark response levels of 0.1 percent and 10 percent for added and extra risk.  Please comment on the assumptions, strengths, and weaknesses of the model averaging approach for determining the POD in the cancer assessment.
                  DR. KENNETH PORTIER:  Dr. Gilbert.
                  DR. KATHERINE ANITOLE:  Oh, I'm sorry.  Did you want me to read 5.7 as well?
                  DR. KENNETH PORTIER:  Yes, please.  Read 5.7 in as well.
                  DR. KATHERINE ANITOLE:  Question 5.7; in agreement with EPA's long-standing approach, all three tumor types from the NTP study -- NTP, 2011 -- were dose-response modeled with multistage models using the typical constrained model coefficients greater than or equal to zero from U.S. EPA, 2012.  Under the U.S. EPA's 2005 cancer guidelines, quantitative risk assessments from cancer bioassay data were calculated by modeling the data in the observed range to estimate a BMCL for a BMR of 10 percent extra risk, which is generally near the low end of the observable range for standard cancer bioassay data.  
                  The BMCs and BMCLs are shown for each of the three cancer datasets.  The results for a BMR of 0.1 percent added risk are presented for comparison.  Please comment on the assumptions, strengths, and weaknesses of the multistage modeling approach for determining the POD in the cancer assessment.
                  DR. KATHLEEN GILBERT:  Okay.  Cancer modeling is not my forte, so I'm really hoping I can get some help from other folks here.  EPA did benchmark dose response modeling of the NTP's 2011 study for all three significantly increased tumor types -- skin tumor in male rats, intestinal tumors in female rates, and lung tumors in female rats.  
                  The dose response modeling section on page 175 of the draft risk assessment describes the rationale for the cancer modeling, the details of which were available in the supplemental file entitled "Information on Human Health Benchmark Dose Modeling," which I could not access in HERO.  But I'm not sure I would have the expertise to evaluate it anyway.
                  The usual multistage model was used with both a 0.1 percent added risk and a 10 percent extra risk to calculate the BMC and the 95 percent lower confidence limits of the MCL ppm.  In addition, to assess the impact of model uncertainty, BMCs and BMCLs were also calculated using two model averaging models, namely the frequentist and the Bayesian.  The results calculated by the three models were similar; less than one log difference.  For example, BMCs with a BMR 0.1 percent AR for male rats; squamous cell carcinoma with 2.96, 3.73, and 9.81, using those three different methods.  And the BMCL was even closer; 1.78, 2.25, and 1.47.  
                  This kind of consistency using the three models, lent support to the accuracy of the BMC calculated for the cancer risk.  Consistent with the case law and 2017 NIOSH Guidance, the EPA uses one times ten the mice as a benchmark for the purposes of this risk determination for individuals, and industrial and commercial work environment.  I will have to defer to the other discussants whether they think this benchmark is stringent enough.
                  They also calculated the inhalation unit risk for humans using the three different methods based on respiratory carcinomas in mice.  The IUR determined by the multistage modeling, 8 times 10 to the minus 7th, was pretty similar to that determined by the Bayesian model, which is 1 times 10 to the minus 6th.  Once again, this provides support for the multistage modeling approach.  However, it didn't seem to me that the details assessing the rationale for the multistage approach were sufficient.
                  Limitations include, as we've heard before, the lack of an estimate of cancer risk for short term 1-BP exposure.  However, as EPA mentions, the uncertainties involved in extrapolating from continuous studies in animals to a case of a once in a lifetime human exposure with their potentially divergent biological mechanisms make such an assessment problematic.  
                  As a result, there are no U.S. federal agency regulatory standards applicable to less than a lifetime exposure to carcinogens.  So, I would just ask some of the other discussants how they felt about that rationale too.  That's all I have.
                  DR. KENNETH PORTIER:  Dr. Johnson.
                  DR. MARK JOHNSON:  I personally prefer the use of Bayesian methods because it implies, as I understand it, the use of prior information distributions for developing your BMR.  So, I applaud the EPA for using that.  I would go further and suggest that maybe you should think about looking at the Bayesian BMR approaches for the non-cancer data and to think about using Bayesian methods for extrapolation development of uncertainty factors.  And I provide a reference, Simon et al. (phonetic), if you care to follow that.  That's all I have.
                  DR. KENNETH PORTIER:  Mr. Kaufman?
                  MR. ALAN KAUFMAN:  Yeah, this is outside my area of expertise.  I'm going to pass.  Thank you.
                  DR. KENNETH PORTIER:  Dr. Pessah?
                  DR. ISAAC PESSAH:  No comment.
                  DR. KENNETH PORTIER:  Dr. Vorhees.
                  DR. CHARLES VORHEES:  I have nothing to add.
                  DR. KENNETH PORTIER:  I am going to help Dr. Gilbert out at this point.  I did look through the section and I looked through the supplement.  I scanned through all of it.  The only thing -- and it looks pretty standard.  It's kind of what was recommended in 2016 that EPA do with these data because, at that time, model averaging was kind of an emerging technology that a lot of people were using.  And I think the panel, at that point, wanted to see that become a bigger part of the risk assessment protocol.  
                  What surprised me is you used it only on the cancer, not with the non-cancer ones.  With the non-cancer ones, you fit the models.  Part of that, I can understand, is because you used the multi -- what is it called again?  The model that was used for the non-cancer effects, the NCTR, nested dichotomous model, doesn't really lend itself to the model averaging approach.  I don't think that the science is there yet for handling multistage models and a weighted averaging approach.  
                  On the cancer point though, I think, if I read correctly, the document recommends the point of departure from the best-fitting model even after you fit the frequentist and the Bayesian.  And I guess my choice would have been, kind of like Dr. Johnson, to use the Bayesian weighted averaging value.  Now, that said, in almost every case, it doesn't change the point of departure but by maybe a point one way or the other.  So, it's probably not enough to argue from my point of view.  
                  The models fit in the weighted averaging are very close to each other surprisingly, until you realize, as Dr. Johnson pointed out to me right before lunch, that in most cases, you're only dealing with three -- at most, four -- dose levels.  So, these models are pretty much fitting a very linear line through.  
                  And they're all fitting the best linear line that they can, which means that the BMD and the BMDL estimates are going to be very close to each other as well.  So, while it was a nice exercise, it probably doesn't produce a point of departure that would be that different than fitting the best-fitting model.  Good exercise, nice write-up; I enjoyed it.  But I didn't see anything that I felt needed to be changed.  
                  One of the other things I looked through is I like to see BMDs and BMDLs that are close to the dose.  Right?  The minute they start extrapolating way below the lowest dose level, you start to get nervous.  One of the things that happened is the uncertainty around the BMD gets big, which means the BMDL starts going towards zero.  And I worry that there's an extrapolation effect going on.  
                  In this case, most of the BMDLs, the BMDs are close to the lowest dose.  So, even if they threw all this out the window and used the NOAEL or the LOAEL, they'd get almost the same result.  So, in this case, I think this is the best that they could do.  I didn't see anything that I needed to see changed here.  
                  Anybody else want to join in on that?  Weighted through that supplement?  No?  We all have our preferred supplements.  EPA has provided enough supplemental files for each of us to have our own, right?  That took care of 5.6.  5.7, did you have some additional comments on the tumor types from the NTP study?
                  DR. KATHLEEN GILBERT:  Not really.
                  DR. KENNETH PORTIER:  Does anybody else want to comment on that?  All right.  I felt that EPA did a pretty good justification of why they chose a BMR at 10 percent, why they threw in the 0.1 percent for comparison purposes.  All of that was pretty well explained. 
                  MS. RUTHANN RUDEL:  I have one clarifying question.
                  DR. KENNETH PORTIER:  What was that, Ms. Rudel?
                  MS. RUTHANN RUDEL:  Oh, it's Ruthann Rudel.  I have a clarifying question.  If a question's about the age-dependent adjustment factors decision, is that part of this question or a different question?
                  DR. KENNETH PORTIER:  That's a good question.  Maybe I'll look to EPA and say, is this where you want to see us discuss that?
                  MS. RUTHANN RUDEL:  Page 178. 
                  DR. KENNETH PORTIER:  I don't see another place where that would come up, unfortunately, unless you wait until Question 8 then bring it in and say, I think you should have asked us the question.  You might wait for that one on Question 6.1, although the 6.1 is MOE.  
                  Let's go ahead and take that question.  Maybe that's a good question.  Because it does relate to the calculation of the point of departure, and I think this is the one question where we're talking about rats.  So, what was the comment, Ms. Rudel?
                  MS. RUTHANN RUDEL:  My comment is just that an age-dependent adjustment factor is included in the cancer risk assessment if children under the age of 15 are going to be exposed.  And I thought that's not that realistic to think that there aren't going to be children under the age of 15 exposed to this compound either through consumer uses or through indoor air.  Because it's daycare or a home that's sharing a building with a dry cleaner, or some other user of this chemical which is used a lot.  
                  I know that there's one scenario that has children in a dry cleaner.  But here it says ADAFs will not be applied for younger-aged children spending time in the workplace, like a family-owned business, because it's unlikely their exposures are chronic in nature.  I didn't find that persuasive.   
                  DR. KENNETH PORTIER:  Does anyone else want to comment on that, on the adjustment factors?  The risk assessors?  I'm looking at Dr. Blystone.  Did you guys feel that what EPA did was -- I don't know.  When I read it, I felt this was kind of the standard thing they would do.  
                  But the question about the younger than 15-year-old children in dry cleaners did come up in the public comment, and that's what you're reacting to.  So, I guess my question, looking at the panel, does EPA need to go back and address that better?  Dr. Gilbert?
                  DR. KATHLEEN GILBERT:  Well, it sort of comes back to the question we were talking about before in terms of internal exposure levels.  We don't seem to really have a handle on that, so it's hard to know how relevant that age adjustment is.
                  DR. KENNETH PORTIER:  Dr. Cobb.
                  DR. GEORGE COBB:  Just to layer on top of that comment I made briefly yesterday, I didn't see anywhere any indoor air considerations in schools that weren't related to direct use in machine shops and those kinds of things, and any off-gassing that you might be experiencing from the insulation materials or computer cleaning could very easily increase these exposures to people of this age class.  
                  DR. KENNETH PORTIER:  Dr. Sheela.
                  DR. SHEELA SATHYANARAYANA:  Well, in general, I would agree with Ruthann.  If there's no exposure data, that doesn't negate the fact that there could still be exposures.  There needs to be exposure data that's confirmatory that children are not exposed.  
                  I think that, over the course of the past couple of days, several people have brought up scenarios and we also heard it from public comments.  So, I think you need to be really transparent that you are not accounting for that population in these calculations.  
                  DR. KENNETH PORTIER:  I was trying to think whether this was discussed in the uncertainties.  Was this discussed in the uncertainties around the point of departure?  I can look to Dr. Barone.  He might be able to answer it quicker than whether I can find it in the document.
                  DR. STAN BARONE, JR.:  So this does get back to a critical conversation we were having before lunch.  Just to remind folks, the 2005 Cancer Guidelines and the Supplemental Guidance specifically call out application of age-dependent adjustment factors for children, again, 0 to 16, for a mutagenic mode of action.  It is not called out in a default in your extrapolation to do age-dependent adjustment factors.  It's only called out in the cases where we observe and declare that a mutagenic mode of action is partially or entirely part of the mode of action for that chemical carcinogenesis.  
                  So I'd have to put that in context with our earlier discussion.  We'd also discussed earlier the conditions of use and the scope of the assessment.  The question is -- and the committee can comment on -- whether children in dry cleaners -- after-school care, during the workday -- that would be in the back of the shop or in the shop, whether those would be a vulnerable population.  We, again, because of the chronicity of the exposure, didn't call those out.  
                  So, if you think we should do something more about that or supply additional information or analysis, that is where the children and the exposure come together.  Because there are exposures and we've outlined what those exposures are for near-field and far-field.  Again, occupational non-users would be included in that occupational non-user group.  That's the only population where children really exist where there's some degree of repeated exposures.  
                  In the consumer products arena, we are looking at the consumer products and the consumer product uses as acute exposures only.  In our Cancer Guidelines, in general, seven years, 10 percent of the lifetime or greater, is what we apply our cancer risks to.  Okay?  Seven years, one-tenth of the lifespan is what generally is recognized as being what we apply the cancer risk models to.  Not one day, not one hit, not one exposure, but repeated exposures in some form or fashion for one-tenth of the lifespan.  
                  DR. KENNETH PORTIER:  Dr. Blando?
                  DR. JAMES BLANDO:  As I'm processing, I think you just answered my comment.  But I'm still processing, so I'm not sure.  But I would just confirm that, in fact, when we did our studies in dry cleaners, we did have dry cleaners that had their children doing their homework at the shop with mom and dad.  And in fact, we had the children who, when mom and dad needed an extra hand, would say, hey, come over here and give me a hand with whatever.  
                  Not to add a funny anecdote; we even had one dry cleaner ask us if we could do our sampling on their dog because they brought their dog to the shop.  And we were going to try and do it but, of course, we didn't.  But I would just confirm that, in fact, we did observe on our one little study that children would be present.  Obviously, those children are not there for seven years.
                  DR. STAN BARONE, JR.:  But there are children -- 
                  DR. JAMES BLANDO:  Well, we didn't ask.  I mean, the children we observed were young, elementary school-aged children all the way up to kids that looked like they were freshmen in high school.  So, I think that's reasonable.  
                  The other thing to note is -- and I wasn't sure if we were going to discuss this somewhere else -- most of the dry-cleaning shops that we would go to were at all co-located buildings.  For example, we had one dry cleaner who their giant axial exhaust fan in their ceiling above their dry-cleaning machine went out their one-story roof, but it was connected to a building that had an apartment with a window.  And I remember commenting, thank God it's cold today, because the window's closed for the apartment complex.  
                  So, I guess there are a numerable number of configurations that you could bring up, but I would just say that I think it doesn't seem uncommon to me based on our anecdotal field experience that these are going to be in co-located buildings, potentially having apartments on the second floor or having kids help mom and dad out in the shop after they get done with school.  Because mom and dad are working 12-hour shifts, so what are they going to do with their kids when they get out of school?  They're going to come to shop and do their homework and everything else.  That was our experience.
                  DR. KENNETH PORTIER:  I'm sitting here having déjà vu, because I heard this conversation in 2016.  And I think, if you read the minutes, you'd hear the same kind of discussion.  Maybe my question to EPA is, in the scoping document, was this discussed?  And the decision made not to include it in scope.  Or was it just not discussed?  That's a good point.  Because I know it was discussed in 2016.  Stan?
                  DR. STAN BARONE, JR.:  Déjà vu.  Yes, it was discussed in 2016.  Yes, it was in scope.  The issue of dry cleaners or the dry-cleaning use was brought in post the 2016 assessment.  That was part of the scope.  The issue of co-located facilities and adjacent facilities and daycares and so on were part of the consideration if we were doing general population exposures.  It was part of the scope that we put out in the previous scoping problem formulation document.  
                  Since the declaration that the agency is moving forward with the listing of 1-Bromopropane as a hazardous air pollutant, those will come under the regulatory authority of the air program.  And further work and analysis and restrictions would come from that effort, not from this effort.  So, we're not looking at risk estimation for co-located facilities and the adjacent buildings.
                  DR. BILL DOUCETTE:  Just a quick follow-up -- Bill Doucette -- this is indoor air we're talking about.  Doesn't that regulation deal only with outdoor air?
                  DR. STAN BARONE, JR.:  We'll be looking at the risks in the indoor environment in the workplace.  That's our job.  Not for the adjacent or co-located facilities.  That's under the Office of Airs -- 
                   DR. BILL DOUCETTE:  Okay.  And a follow-up to that, if I may, other consumer products other than dry cleaning we mentioned that could potentially be a source of indoor air, would that be under your purview?  It doesn't seem like it should fall under the outdoor air.
                  DR. STAN BARONE, JR.:  Again, for the consumer products that are generating or using 1-BP, that comes under our scope.  And that is part of our scope, so that's the reason why we did the market analysis we did.  We've googled and we've searched our own proprietary databases and we've contacted the industry associations to a pretty extensive degree.  
                  It doesn't mean we wouldn't appreciate any new information.  But on consumer products, we are looking at that.  There are consumer products included in this assessment, and that is indoor air.  And that's why we're modeling the insulation board as well as the other uses of consumer products that we think are significant sources of 1-BP exposure.
                  DR. HOLLY DAVIES:  Could I just follow up quickly?  Just because, when I think of HAPs, I think of big facilities.  I know you're not the air program -- and this is Holly Davies -- but if you could just give a little information on how the Clean Air Act would regulate the dry cleaners, I think that would just help alleviate some of our concern.
                  DR. STAN BARONE, JR.:  I'm not in a position to give the detailed background on the HAP analysis.  But it is an extensive program.  I think there were comments the other day that the agency is still in the process of deciding about accepting the petition.  That's not true.  The agency has accepted the petition to list 1-BP as a HAP.  Listing 1-BP is where we are now, the actual listing.  And that is an administrator's activity.  The administrator actually has to sign off on the listening.  
                  As was mentioned, frankly, the agency hasn't listed a new chemical on the HAP list.  This is the first, so this is a big deal.  It's eminent, we have been told.  But that's a process.  So, when that will happen, I can't say.
                  DR. HOLLY DAVIES:  But would dry cleaners have to have permits?  Like I said, I just think of big facilities, not little facilities.  I guess Sheri knows.  Maybe I'll just ask Sheri later.
                  DR. SHERI BLYSTONE:  Well, just a couple of comments.  I'm not a Clean Air Act person either, but it's all about the sources, right?  So, there's stationary sources, mobile sources, indoor sources; I'm just looking at the EPA's website right now.  They work on all of those and decide, based on the reporting and everything, what needs to be controlled.  So, it would include indoor air as well.  If dry cleaners are a source, then they would be covered under the program, right?
                  DR. HOLLY DAVIES:  And they're not too small a source.  Keeping in mind the decisions have been made, I get that.
                  DR. STAN BARONE, JR.:  Dr. Blystone's correct.  The Office of Air will be taking information that we've gathered and using that in their source characterization.  There's a rulemaking process.  The rulemaking process, the max standard develops a rule and a standard for testing and monitoring for those different source categories.  There's monitoring that follows that rulemaking, that data collection and the actual measurements -- real measurements, not modeling -- and modeling.  But there's actual monitoring measurements that are part of a process.  That's a multi-year effort.  
                  Then, following the monitoring, there's an assessment and then there's a residual risk assessment for the different source categories.  So, again, risk mitigation follows that.  Previous efforts go back in time.  You can look at the perc rule and the max standard for perchloroethylene in dry cleaners.  Dry cleaners were assessed for perchloroethylene.  There is a phase out.  There are efforts underway to get perc out of dry cleaners.  
                  Again, those kinds of multi-year efforts were significant for the perchloroethylene effort.  Whether that would be applied for 1-BP is somewhat debatable.  The facts on the ground are we're talking about six or seven dry cleaners in the whole country using 1-BP.  Facts on the ground.  This is not 20 thousand dry cleaners.  This is not a broad-spectrum exposure source category today.  So, again, let's put our focus on risks where the risks occur.  Just wanted to make that clear.
                  DR. KENNETH PORTIER:  Mr. Kaufman?
                  MR. ALAN KAUFMAN:  Yeah, hi.  Al Kaufman.  I think you may have just answered my question when you reminded me that there are only 20 dry cleaners using the 1-BP.  But I wanted to go back for a second to children's exposure.  
                  Considering that dry cleaning is very often an entry occupation for a lot of immigrant populations where everybody in the family kind of pitches in and works after school, I suspect that for those situations where kids are present, they might not be just in the back doing their homework.  They might actually be working.  And they might have a chronic exposure.  But if we're only talking about 20, we're not talking about a huge risk.
                  DR. KENNETH PORTIER:  Ms. Rudel?
                  MS. RUTHANN RUDEL:  I agree with your point about focusing on risks where risks occur.  I think I commented earlier that it's hard to feel a lot of confidence in knowing how and where the chemical is used.  So, we don't have a lot of great information about that.  I'm concerned about overlooking some exposure pathways and scenarios that could be important.  
                  One of them is this family business situation.  And another one I'm interested in for any kind of product that's a consumer use product, if you're a child in a household and there's a hobby room or somebody collects coins or they're doing some arts and crafts and they have these materials in their home, then that seems like a chronic indoor air exposure from consumer product uses.  
                  We could consider including that as a susceptible population or something like that.  I'm not exactly sure why it would be not considered chronic if you have a consumer product, because people have them sitting in their houses and might use them regularly and have them in their indoor air as a result.
                  DR. STAN BARONE, JR.:  So, if the panel has information about the periodicity of any of these consumer uses, again, that would apply to a chronic exposure, information we have, at least from consumer products or obvious kinds of uses of these products, they are not used every day.  They're not used repeatedly every day.  So, that's very different than a chronic exposure.  
                  Some of the information that was brought forward about the off-gassing could inform some of our approaches and our modeling, but that isn't necessarily a chronic exposure.  That could be a subchronic exposure.  We'll have to look at that.  But again, if you have additional information that would help inform those activities or patterns, that would be really helpful to us.  
                  Something that we struggle with, again, there's been some discussion prior about surveys and publications on this particular topic.  For consumers, there is a paucity of data.  There's very few comprehensive studies of consumer products.  And we have, of course, research efforts in our Office of Research and Development on this as well.
                  DR. KENNETH PORTIER:  Ms. Rudel and Dr. Gilbert, it sounds to me like you might want to make a recommendation or a suggestion that this be better discussed in, it looks like, Section 4.4.1 on potentially exposed or susceptible subpopulation, other risk-related considerations.  I don't see discussion of what you just talked about, and you might make a recommendation that there's obviously a data gap.  
                  Although we acknowledge the fact that there's not a lot of locations today that are using 1-Bromopropane in a dry cleaner, that's not to say, next year, that shoots up 100-fold.  So, I think it's valid to at least have a suggestion that that uncertainty be discussed in that section.  Maybe the two of you could put that in.  Dr. Gilbert, had you wanted to make another comment?
                  DR. KATHLEEN GILBERT:  I just have a question.  I know this isn't really what we're discussing, but the whole Clean Air Act petition for 1-BP, we said 20 dry cleaners this might be relevant for.  They will actually put a chemical on the Clean Air Act for such a small number of facilities that it might be relevant to?  Or is there some other rationale that I don't understand?
                  DR. STAN BARONE, JR.:  The petition was for 1-BP manufacturing use processing.  Dry cleaners are just one source.  Again, you had heard about the estimates of 24 million pounds production.  Production, processing, other facilities could be sources, and actually, more significant sources of general population exposure.  So, the Clean Air Act would be addressing those kinds of activities/sources to the general population.  
                  This is the dissection between TSCA and the Clean Air Act.  We're actually looking at the internal, inside the house.  We're looking at the manufacturing, the facility, the processing facility and the workers there, and the occupational nonusers, whereas the Clean Air Act covers the general population.
                  DR. KENNETH PORTIER:  Hence my recommendation that we put it in there.  Because it doesn't take it off the table; it leaves it there as something that EPA, at some point, will have to consider.  
                  Any additional comments on points of BMD modeling and calculation of the points of departure for non-cancer and cancer endpoints?  I don't see any, so I think we'll move on.  Where are we?  Question 6?  Is that where we are?  Yes?
                  MS. RUTHANN RUDEL:  I have a request for that, because I'm lead on 6.1 and 6.3 and they're acute versus chronic.  Could we combine those two?
                  DR. KENNETH PORTIER:  6.1 and 6.3?  I'll tell you what.  Why don't we read 6.1, you make your comments for 6.1 and 6.3, and then when we get to 6.3, say, "I've already made those comments."  And that gives the panel the chance to address both of those separately.  So, let's go ahead and -- yes, Ms. Rudel?
                  MS. RUTHANN RUDEL:  It's Ruthann. Have we passed the question where we talk about uncertainty factors?  Or is that coming?  That's coming?  Okay. 
                  DR. HOLLY DAVIES:  Could we just take a quick break between questions?
                  DR. KENNETH PORTIER:  Take a break?
                  DR. HOLLY DAVIES:  Or does no one else want a break?
                  DR. KENNETH PORTIER:  Okay.  Well, let's take a 10-minute break.
                  DR. HOLLY DAVIES:  I feel like a small break between questions.
                  DR. KENNETH PORTIER:  Take a 10-minute break.  What time do I have here?  I've got 1:42, so we'll come back together five minutes before 2:00.
                  (BREAK)
                  DR. KENNETH PORTIER:  Before we move onto Question 6, Dr. Blystone said she had one more comment she wanted to get on the record.
                  DR. SHERI BLYSTONE:  Yeah.  I mean, because at one point, you had pointed to me and asked for my opinion, but I never gave my opinion.  But I wanted to follow up with Ruthann exactly what her concern was.  So, in the break, I did do that.  
                  Her concern was that children in dry cleaning operations should be considered for a chronic hazard.  And if you're talking about every day probably, that they're there every day throughout their childhood, seven years is going to be part of that.  So, it seemed reasonable to me as well that that is a pathway that should be considered for chronic hazards.
                  DR. KENNETH PORTIER:  All right.  A chocolate fix is going around the room.  It stopped.  Now, Dr. Anitole, please read in Question 6.1.
                  DR. KATHERINE ANITOLE:  For non-cancer risk estimates, EPA used the margin of exposure, or MOE, approach.  MOEs for acute and chronic exposures were calculated separately based on the appropriate non-cancer POD and the estimated exposure concentrations adjusted for durations.  The endpoint-specific MOEs were compared to the endpoint-specific benchmark MOEs.  If the MOE estimate was less than the benchmark MOE, the exposure scenario for non-cancer endpoints was interpreted as a human health risk.  
                  For cancer risk estimates, EPA multiplied the occupational scenario-specific exposure estimates by the cancer IUR to estimate the extra cancer risks.  Extra lifetime cancer risk estimates from 1-BP exposure were compared to the benchmark cancer risk levels.

                               CHARGE QUESTION 6

                  I think I got a little mixed up here.  Question 6.1; please comment on the assumptions, strengths, and weaknesses of the MOE approaches used to estimate the non-cancer risks to workers and occupational non-users -- for example, adults of reproductive age -- following acute inhalation exposures to 1-BP, including the MOEs presented in the document.  Specifically, please suggest alternative data that could be used.  Please comment on the selection of uncertainty factor values in deriving the benchmark MOE for acute inhalation exposures.
                  Did you want me to read 6.3 in combination?
                  DR. KENNETH PORTIER:  I think we'll get to it.  Let's just tackle 6.1, and we'll just fold Dr. Gilbert's comments into 6.3 at the time.  Dr. Gilbert?
                  DR. KATHLEEN GILBERT:  Well, I think I'm supposed to be talking about -- okay, the non-cancer endpoints.  There's an amazing table, 3.1, that describes all of the different organ systems and talks about the uncertainty factors which, as far as I can tell, are perfectly appropriate.  And they also describe obviously the BMCLs which, for a particular target organ system, seem to be fairly consistent across the different studies.  
                  So, the main question, I think, has to do with the study that they used which, for many of the endpoints, they used the WIL paper to talk about acute inhalation exposures.  This is somewhat problematic, as they discussed, but I have a rationale for.  They say that support for using this endpoint for acute short-term exposures is the fact that the male and female reproductive effects, that is in the F-zero males and females collectively, contributing to the decreases in live litter size, which is the endpoint they're looking at, all occurred within a short window of exposure between ovulation and implantation.  
                  They suggest that other studies have shown, which is absolutely true, that short-term exposure to certain chemicals during this critical window can induce long-term effects.  That's very well documented.  However, showing that chronic exposure, that happened to overlap this crucial short development window, caused long-term effects in the pups is not the same.  
                  Because, in the WIL study, yes, they are exposed during that short window of time, the parents.  But they're also exposed for 70 days premating.  And they're making the argument that the effects that they see on the live litter are only due to the short window of time during gestation as opposed to anything that's happening during the 70-day preexposure.  Described by the EPA as a very health protective effect, not necessarily accurate, and that could overestimate the risk to workers and consumers following acute exposure.  
                  However, they think it's justified enough because of the studies that have shown that short-term exposure during that time period can cause the long-term effects that they saw in the pups.  So, I am basically okay with that since they're calling it a health protective, even though I don't see the rationale.  I don't think you can actually say that you can guarantee that the effects that they got in the pups had to do with just that narrow window of exposure and, therefore, was a result of the acute exposure.  I'd be interested in what other folks have to say about that.  
                  Let's see.  Table 3.1, as I said, describes the endpoints and the ATCs selected for the inhalation non-cancer dose response.  Applied a composite uncertainty factor of 100 for acute and chronic inhalation benchmark MOE based on default intraspecies uncertainty factor of 10, an additional 10 to account for additional toxicokinetic differences, including variability in expression and function of CYP2E1, which is involved in 1-BP metabolism.  
                  Since BMD modeling was used to derive at HEC, a NOAEL to LOAEL uncertainty factor of one seemed appropriate.  And the sub-chronic to chronic was not needed because a key study used a chronic exposure protocol.  
                  One thing I had a question about.  They talked about the different HECs provided for acute versus chronic exposure; when describing the decreased live litter size and post implantation loss in F0 females, there were different HECs for the acute versus chronic.  Explained by the fact that HECs are calculated by duration adjustment and a human equivalent DAF.  
                  The adjusted POD is the POD times duration adjustment.  And the baseline used for the duration adjustment was 8 hours per day for exposure for occupational and 24-hour exposure for consumer.  For acute exposure, the duration adjustment was hours per day exposed divided by 8.  And for chronic, it was hours per day divided by 8 times days per week to reflect a 40-hour work week.  
                  To me, this needed a little more explanation.  I think that's basically it.  My main concern was the use of the WIL study, which is a chronic exposure to look at acute endpoints.
                  DR. KENNETH PORTIER:  Thank you.  Dr. Anderson, do you want to add to this?
                  DR. HENRY ANDERSON:  Yeah.  I don't have much to add.  The only thing that I saw is that there's really a lot of studies that are categorized as being high quality.  The one thing here that I thought is, if you look at the human data, what's striking is that at least the short-term or a higher exposure impacts appear to be all neurological.  And there's quite a number rated high quality neurologic studies.  
                  The data quality in the neurologic short-term studies, which gets you some fairly low NOAELs,  some of those haven't been evaluated.  There's two rated high that have fairly low levels.  It would seem to me it would be helpful, when you have a plethora of studies and different endpoints, to provide the analysis of what would be the benchmark or the selected study if you were to use different endpoints.  But I just think that neurologic endpoints are more consistent with what actually is on the ground experience in the human category.  
                  So, it would seem that it would be helpful to have a better explanation for why they chose to only use a single study when there are multiple choices.  And why they then didn't choose the neurologic endpoints, or at least do the calculation with them.  With the issue of uncertainty factors, when there's pretty good consistency between the animal studies and the human experience, one could reduce the uncertainty factor in the animal to human translation there.  
                  So, I think it would be helpful to perhaps add some reliance on what is seen in the human studies, even if those data are not used.  That's one of the issues.  And then, again, the identification of the subpopulations.  Clearly, individuals with neurologic disease, those with diabetes in the workforce, all of those are diseases that have an impact on the neurologic system.  That places them at greater risk to other neurotoxins.  
                  So, that would be some recommendations I have.  Perhaps, some greater discussion of alternatives to the cancer risk.  That's diligent in this case to utilize for the acute exposures.  That's it.
                   DR. KENNETH PORTIER:  Thank you, Dr. Anderson.  Does anyone else want to comment on this question?  Maybe somebody else on the line.  Dr. Cory-Slechta?  Dr. Pessah?
                  DR. CORY-SLECHTA:  Yeah, no comment.
                  DR. ISAAC PESSAH:  No additional input.
                  DR. KENNETH PORTIER:  Dr. Gilbert, I have a question.  Using a chronic developmental study to set an acute point of departure, is that conservative or anti-conservative?  Is that going to produce a lower point of departure or a higher one?  I was just trying to think of the implications of doing that on the risk assessment.
                  DR. KATHLEEN GILBERT:  That's a good question.  I'm not sure.  I mean, you could make the argument that something occurred during that chronic 70-day exposure that made the uterine wall less habitable for the pups.  You can't necessarily say that that was the one short window during gestation that caused it.  But whether it's going to make it more or less sensitive, I am not sure.  
                  It could be argued that they're using it as more conservative, and I think they're probably right.  So, they're using it as, let's be on the safe side and use this as their study to use to devise these numbers from.  And I think that's probably okay.  I just want to acknowledge the fact that it's not clear-cut science.  
                  DR. KENNETH PORTIER:  Any other comments?  You guys are happy with that?  Just checking.  Dr. Willett?  You know, your body language like this tells me, I don't want to say anything.
                  DR. CATHERINE WILLETT:  Actually, the thing I wanted to comment on was, why not use the neuro?  The neuro for the non-cancer, it was included as a consideration.  But I also think that -- and maybe for short-term -- in the human occupational studies, they wouldn't necessarily have seen the reproductive effects because they weren't looking for it.  So, it doesn't mean they're not there.
                  DR. HENRY ANDERSON:  To me, that issue would be, is the neuro result in a less conservative or less protective number than using the reproductive?  And I would say, okay, the reproductive gives some added security and it's the same as -- you use a chronic study when you have a lot of more acute studies, and then make the assumption that there could be smaller windows in there, and things mixed into the chronic studies.  
                  So, I would just like to be assured that, if the decision was you'd use the reproductive one, because it's a more cautious, protective number, that's hard to argue that, well, let's be cautious.  But if it ends up with a less protective number than if go through using the neuro data, then I'm not sure which of those is more protective if that's the decision, as we want to pick the study that results in the most protective approach for acute exposures as opposed to which actually is the strongest or most relevant.
                  DR. KENNETH PORTIER:  I'm not sure I heard a recommendation come on it.  I heard comment on it.  Dr. Gilbert, is that correct?
                  DR. HENRY ANDERSON:  Oh, my recommendation would be to do the analysis on multiple endpoints or at least one other endpoint to see what the result would be.  And then, if the decision is you want to go with the one that's most protective, then do that, rather than just say, well, this one using a chronic are anticipated to be more protective.  But I would like to actually see that to be the case.  
                  DR. KENNETH PORTIER:  Yes, Dr. Gilbert?
                  DR. KATHLEEN GILBERT:  I don't really have any recommendations.  And they did a good a job explaining that they knew this wasn't completely accurate, but they thought it was more health protective, so they were going with this study.  So, I don't really have a recommendation.
                  DR. KENNETH PORTIER:  Dr. Anderson, on page 171, Table 3.1, there are a lot of developmental studies that they analyzed and presented points of departure from other than just the WILs post-implantation result.  That just happened to be the one that produced the lowest POD.  But your point is that there is no -- well, there are four nervous system studies that are analyzed.
                  DR. HENRY ANDERSON:  I was just looking at the summary tables.
                  DR. KENNETH PORTIER:  But none of that analysis was done on epidemiology data.  It's all done on animal data, obviously.
                  DR. HENRY ANDERSON:  No, no.  I agree.  And it's true.  They say there are no studies of reproductive exposure in humans, but there's quite a bit of neurologic studies, which, again, some of the exposures appeared to be relatively low.  Now, again, some of the criticism of those studies as well are used as time-weighted average and there could be peak exposures in there.  Well, that's true of all of the studies where you use a time-weighted average.  
                  So, that's kind of why I take a greater interest on neurologic animal studies being consistent with the human experience, simply because we don't have reproductive studies.  And in most instances, reproductive outcomes are, in fact, more protective.
                  DR. KENNETH PORTIER:  Dr. Vorhees?
                  DR. CHARLES VORHEES:  Well, for this, they do cite human data for neurological effects down to very low doses, down to 0.07 ppm.  And normally, you would use human data in preference to animal data.  But you then say that there are methodological limitations to these studies, and then I presume that that's the reason that they were not used.  
                  Now, I didn't go in to look in detail of what the deficiencies in those studies were.  But I presume they were pretty serious, or you would have deferred to the human data over the animal data.  Is that correct?
                  DR. STAN BARONE, JR.:  This is Stan Barone.  That is correct.  Again, the Chinese worker studies had some methodological concerns that were called out and explained in the systematic review data quality evaluations.  We also received significant public comments about the use of those studies in our prior assessment as well.  There's a lack of blinding of the physicians who are doing the testing.  There's, again, from the description of the studies, the methodologies submitted inappropriate aspects to the testing.  
                  There's probably something significant there, but to say that they're ready for dose response, we did not choose to use them for dose response.  They're a part of the hazard ID weight of evidence.  And again, we fell back on the animal studies where we could actually do dose response and felt we had more confidence.  If you look within, for the developmental effects and the neuro effects in the animal studies, we're basically talking about within the order of magnitude for those effects.  We're not talking about several orders of magnitude.  
                  So, they are close to one another, but we appreciate your comments.  We recognize that using this developmental outcome following a repeated exposure has some controversy around it.  
                  DR. KENNETH PORTIER:  Dr. Davies?
                  DR. HOLLY DAVIES:  This is Dr. Davies.  I wanted to speak to the question then the suggestion to use other endpoints for the acute.  On page 145, EPA refers to their Guidelines of Developmental Toxicity Risk Assessment and explains that the reason they were using the developmental endpoint is because of the critical windows of fetal development and that that was appropriate to use the developmental endpoint for the acute because the one exposure, and that wouldn't be appropriate for the other endpoints.
                  DR. KENNETH PORTIER:  Dr. Anderson, did you want to comment?
                  DR. HENRY ANDERSON:  Oh.  Again, yes, they're close.  But that doesn't say which is more than the other.  There's a lot of comparative language used without -- and the quantifier we're seeing, it could be, if that's the case when you analyze it, then say that and that just further supports the choice of the reproductive outcome.  And I'm not against using that; I'm simply saying there's a lot of other data there that I wasn't overly convinced of the reasoning for not at least using that.  
                  Now, I can understand the human epidemiology data that typically, in studies like this at least, one of the criticisms was they don't describe the selection process.  Well, you can't particularly randomize workers.  But there were a number of concerns.  There always are with these types of studies that do an observational type of study.  But whether those are fatal flaws or not, I would tend to argue with.  But I'm not against using the reproductive study.  
                  Just, my sense would be, I would like to have seen a little more review of alternative.  Maybe it's all there and, given all the pages that are there, I misinterpreted, or I didn't read it as carefully as I should have.  But I just felt the neuro disease -- again, that's what's been seen in the people.  
                  DR. KENNETH PORTIER:  Okay.  Before we leave this question, I just wanted to make sure everyone was happy with the selection of the uncertainty factors in all cases and a total uncertainty factor of 100 was used.  I didn't really hear anybody commenting one way or the other.  It seemed pretty standard to me, but I wondered.  There didn't seem to be any reason to go higher or lower.
                  Okay. Question 6.2.
                  DR. KATHERINE ANITOLE:  6.2; please comment on the assumptions, strengths, and weaknesses of the MOE approaches used to estimate the non-cancer risks to consumers following acute inhalation exposures to 1-BP, including the MOEs presented in the document.  Specifically, please suggest alternative data that could be used.  Please comment on the selection of uncertainty factor values in deriving the benchmark MOE for acute inhalation exposures.
                  DR. KENNETH PORTIER:  Dr. Davies.
                  DR. HOLLY DAVIES:  I thought we were doing 6.3.  
                  DR. KENNETH PORTIER:  We'll get to 6.3.
                  DR. HOLLY DAVIES:  Kathleen, did you want to do them together?  
                  DR. KATHLEEN GILBERT:  Sure.
                  DR. HOLLY DAVIES:  Or it's fine with me, whichever we'd prefer.  
                  DR. KATHERINE ANITOLE:  If that's okay, we might as well do 6.3 now.
                  DR. KENNETH PORTIER:  Well, please read 6.3 into the docket, and then you guys can discuss it how you want to.
                  DR. KATHERINE ANITOLE:  Question 6.3; please comment on the assumptions, strengths, and weaknesses of the MOE approaches used to estimate the non-cancer risks to workers and occupational non-users following chronic inhalation exposures to 1-BP, including the MOEs presented in the document.  Please comment on the selection of uncertainty factor values in deriving the benchmark MOE for chronic inhalation exposures.
                  DR. KENNETH PORTIER:  Dr. Gilbert.
                  DR. KATHLEEN GILBERT:  Well, there's not really much to add here.  Once again, they're using the WIL study for many of these.  And as far as chronic exposure goes, it would be hard to argue that they would not be appropriate.  They're using obviously the same uncertainty factors and all the other parameters.  So, I don't really have any kind of problems with this question.  I don't have any suggestions either.
                  DR. KENNETH PORTIER:  Dr. Anderson, did you want to add anything?  6.3?
                  DR. HENRY ANDERSON:  No, I don't think so.
                  DR. KENNETH PORTIER:  Does anyone else want to comment on 6.3?  It seemed pretty straightforward.  Yes, Dr. Barton?
                  DR. CHARLES BARTON:  I just wanted to bring up again that the exposure model for consumers, the way that it was used in the consumer exposure model, evaporation wasn't taken into consideration.  Whereas, for the workers, the fraction that's absorbed does.  So, two different models are being used, one for workers, one for consumers.  
                  It just appears a little strange why one was chosen for each.  May be best if it's explained why one was chosen.  For example, why was this one chosen for consumers?  And giving a rationale behind different inputs on why these are best would go long ways in allowing the reader to see what is occurring and allow more transparency.  Because now we're having to assume the assumptions that are being made.  Thank you.
                  DR. KATHLEEN GILBERT:  I just think that's a good point.
                  DR. KENNETH PORTIER:  Thank you, Dr. Gilbert.  Any additional comments on 6.3?  Then let's go to 6.2.  Dr. Davies.
                  DR. HOLLY DAVIES:  So, since it's already been read in, I want to make the point just that the MOE, of course, is dependent on all the uncertainties in estimates of consumer exposure that are going to be in Question 6.5 and in the human health hazards.  So, all the uncertainties and the assumptions and strengths and weaknesses that we've talked about in different parts.
                  I thought there was really a nice explanation in the risk characterization section on page 251 on the uncertainty factors.  It laid out what all the uncertainty factors were, although it kind of directed back to Table 3.1 that Kathleen mentioned before.  So, for each specific ones, I was surprised to see that they were all the same.  I was expecting to see different ones picked for different reasons, but they were all 100.  
                  I didn't have any comments as to why they shouldn't be 100.  And it sounds like nobody else has a comment on that either, but maybe they do for consumers.  I was taking Chuck's comment on the consumers modeling for this one also.  And then Table 26, it was very nice to have all of the MOEs for the consumers.  I didn't have that much for this question, if anyone else has anything.
                  DR. HENRY ANDERSON:  I don't.
                  DR. HOLLY DAVIES:  One more thing.  One thing I did really like in comparison to 2016, is how there's so many endpoints.  Because in the draft risk assessment, there was just the one endpoint for non-cancer, just developmental.  So, here, it's nice to see all the endpoints of the points of departure that were done for multiple endpoints and to see them from line-up.  It gives that nice weight of evidence that it's not just reliant on one point of departure from one study, and to see that weight there is really nice.
                  DR. KENNETH PORTIER:  Any additional comments?  Dr. Anderson?  Dr. Cory-Slechta?
                  DR. HENRY ANDERSON:  No.  I like that table that had them all as well.
                  DR. ISAAC PESSAH:  This is Isaac.  It was very helpful.
                  DR. KENNETH PORTIER:  I think they like this section.  That was good.  It's wonderful when you have data.  You know?
                  DR. STAN BARONE, JR.:  Yes.
                  DR. KENNETH PORTIER:  Any additional comments?  Let's move onto 6.4.
                  DR. KATHERINE ANITOLE:  Question 6.4; please comment on the assumptions, strengths, and weaknesses of the approach used to estimate extra lifetime cancer risks to workers, which EPA derived from an inhalation unit risk based on lung tumors in female mice for estimating incremental or extra individual lifetime cancer risks.
                  DR. KENNETH PORTIER:  I'm looking at Dr. Gilbert.
                  DR. KATHLEEN GILBERT:  Okay.  Once again, I hope some other people can chime in on this.  I don't really have that many suggestions or issues with this.  The calculations and the parameters they used seemed appropriate.  The cancer risk estimates for acute inhalation were not derived as we've talked about before, but the chronic exposure ones seemed fine.  The uncertainty factors, again, seem appropriate.  
                  The data for the lung tumors was based on a combined incidence of alveolar adenoma or carcinoma in female mice and was selected for derivation of the inhalation unit risk and the dermal slope factor.  They used multi-stage modeling as well as the averaging, the frequentist.  I know some people like the Bayesian especially, so they used that model as well to determine the BMCL ATC values for both the 40-hour work week and 24 hours per day.  
                  The BMCL ATC values represented 95 percent lower confidence limit estimate of the occupational exposure concentration expected to produce a one in ten or one in a thousand lifetime extra or added risk of lung cancer was used.  Although not directly comparable due to different BMR values, they produced IUR for both the 40-hour week and 24-hour exposure models that were similar.   
                  The values selected as a POD for the inhalation unit risk and dermal slope factors were thought to be protective of more sensitive individuals, which is always good.  Of course, the MOA of 1-BP is not known.  Therefore, the inhalation unit risk and dermal slope factors were calculated using a default linear approach, which seems appropriate.  
                  Since the IURs and dermal slope factors calculated by all three modeling approaches were similar, the model averaging can be used to determine that model uncertainty was not significantly impacting the IUR or the slope factor.  So, that means that the slope factor determined using the multi-stage modeling was consistent with EPA guidelines for cancer risk assessment.  Additional high confidence in the IUR or dermal slope because they were based on high quality animal data.  
                  So, basically, I think it sounded really good and I didn't have any issues or recommendations for this section.
                  DR. KENNETH PORTIER:  Dr. Anderson, do you want to comment?
                  DR. HENRY ANDERSON:  Why do we always start at the top of the alphabet?  Yeah, I just have a question here.  Let's use this cancer risk only to workers who are -- in all the others, it includes occupational non-users.  I'm assuming they're not omitting the question about exposure and cancer risk to occupational non-users.  It's in the document, but I was curious that it wasn't part of this particular question.  
                  The only other thing, I don't know where it fits, but it's come up a couple of times before.  That's the integration of the inhalational exposure and the skin exposure to a data integration into a whole list, like risk to the individual, rather than just parse the risks by skin or by inhalation.  So, somewhere, I think that combined exposure and then what are the risks of the combined exposure should be in this document somewhere.  So, that would be a recommendation.
                  DR. KENNETH PORTIER:  Dr. Anderson, I'm looking at Figure 4-1 on page 230 where it just displays the cancer risk estimates for the occupational users.  That kind of combined the dermal inhalation risk.  Would you do that for all of these or only for the ones that have the lowest risk, like the import worker or the disposal worker?  
                  I'm just trying to figure out whether this recommendation is a global one.  I mean, most of these already show significant risk just from inhalation.  So, to my way of thinking, this would be maybe an added analysis that might be done for those that are a marginal low risk and that that dermal might put them over a 1 times 10 to the 5 or 1 times 10 to the -6 threshold.  Does that make sense?
                  DR. HENRY ANDERSON:  Yeah.  I think part of the discussion earlier was if the inhalational exposure is not terribly high, then the contribution of skin was up.  It becomes more significant where they have a high exposure and the skin gives the marginal addition to it.  I just would like to have it at least discussed.  
                  It doesn't necessarily have to be calculated for all of them, but it could be, is it important?  And when is it important to consider the two together?  Because again, somewhere, I would again suggest having a discussion of mitigation impacts rather than, in many of these tables, having the respirator use and the glove use right in the table with the others.  It seems to me the issue here for TSCA is, is the environment the worker is in hazardous or not?  Then, if it is hazardous, what can be done?  
                  Certainly, recognizing the impact of if respirators are used, being positive, I think that's very important.  But it almost here implies that the environment is somebody wearing a respirator when, in fact, what's in the air, not necessarily what the person is inhaling, is the target here.  
                  So, I don't know whether the combined -- how much difference that would make.  I mean, if it's already excessive, then that's fine.  But it does help if you're going to say, well, how are we going to control this exposure in the workplace?  Is it more important to spend your money on a respirator program or buying expensive impervious gloves for everybody?  Or if you'd like to do them both.  That's really my issue about having somewhere in here how these two interact.
                  DR. KENNETH PORTIER:  Dr. Blystone?
                  DR. SHERI BLYSTONE:  I'm on a different topic.  Sorry.
                  DR. KENNETH PORTIER:  Okay.  I think Dr. Barone wanted to comment on that.
                  DR. STAN BARONE, JR.:  I wanted to get some clarification, because I heard two things and I'm not sure I understood the two issues that Dr. Anderson was raising.  One, I think, he was raising issues about occupational non-users.  Again, our assumption is occupational non-users are not using PPE, whether it's inhalation or dermal.  And there are risks for occupational non-users.  
                  We're showing the impact on the risk estimates for inhalation APF reduction for the workers for each of the scenarios.  And then, the protection factor, if it's appropriate for dermal.  The second part of the question, I think you were talking about aggregate of the dermal exposure and the inhalation exposure.
                  DR. HENRY ANDERSON:  Yeah.
                  DR. STAN BARONE, JR.:  And if you have some suggestions on how you think we should do that, we would much appreciate what approach you're proposing.
                  DR. HENRY ANDERSON:  I'm not sure I have a method.  What I would do is talk to NIOSH experts or others that are doing that.  For instance, you do biomarker measurements.  That really is an integrated exposure of the two.  How you can calculate the exposures, you're already calculating the skin exposure based on what's in the air measurements.  You have a relationship there and that is somewhat proportionate.  So, I think you can sum up what the exposure would be.  But that's probably an approach that could be used.
                  DR. STAN BARONE, JR.:  We're still not clear on how to make that actionable.  We struggle with this, Dr. Anderson.  Without a PBPK model to do the aggregation, we're really struggling.  So, if you have some other approach that we don't know about, we'd appreciate it.
                  DR. HENRY ANDERSON:  You've got to remember, I'm a physician and my job is diagnosing and treating ailments and looking at the exposures of the individuals.  Therefore, I look at the aggregate exposure of the individual and then try to parse out where those exposures are coming from.
                  DR. STAN BARONE, JR.:  I totally get it.
                  DR. HENRY ANDERSON:  I'm just saying that I don't see that when you break out the various routes of exposure and treat them independently.  You know?  You may have missed a totality.  Or maybe you can ignore the skin, and with a high inhalational exposure, the skin is not going to really contribute much to the aggregate exposure.  In my practice with clinicians, I don't use models to do those or the PBPK kind of things.  
                  So, you'd have to ask somebody with that expertise how you can do that versus a simply additive kind of approach.  I'm just recommending that you'd take a look at that.  If it can't be done, then simply put that in the document.  We explored doing this and we can't come up with a way to do it.
                  DR. KENNETH PORTIER:  I think what we've identified is another methodological gap.  We've identified data gaps.  This is more of a methodological gap and one that we can possibly at least point out.  It's something that might need to be discussed in Section 4.3.1 on uncertainties of the occupational exposure assessment, something that, in a sense, hasn't been discussed that's integrating this inhalation and dermal.  
                  I understand what Dr. Barone is saying about without having a PBPK model, it's hard to integrate those two sources of exposure into some kind of endpoint that utilizes both of those.  I hadn't thought that far.  I'm glad you mentioned that.  I was thinking, well, just add them up.  It doesn't work that way, does it?  It's not that simple.  So, Dr. Gilbert, maybe make a point to get that in there, okay?  Now, Dr. Blystone, another topic related.
                  DR. SHERI BLYSTONE:  Yeah.  It is a different topic, but a question to the rest of the panel members, really.  In thinking back to some of the conversations from yesterday, I understood that the mouse model may not be the most relevant to humans for carcinogenicity.  If we're using that here, the question was, does that add any uncertainty in this part of the report?  If so, was that adequately captured?
                  DR. KENNETH PORTIER:  So you're thinking in terms of the uncertainty factors and whether 100 was adequate?  Is that part of what you're talking about?
                  DR. SHERI BLYSTONE:  I mean, if they're using that model to derive their additional analysis of risk, there's some question about whether that model is appropriate to humans.  If that uncertainty mattered here, I'm not sure.  I'm asking the question.  And if that answer is yes, then was that adequately addressed in the report?  Because I don't remember anything particularly around that here.
                  DR. KATHLEEN GILBERT:  Well, I guess the issue of whether mouse studies are relevant for human or how relevant they are to human comes up all the time.  I'm pretty sure they address that uncertainty factor as part of their calculations.
                  DR. KENNETH PORTIER:  Any additional comments related to estimation of extra lifetime cancer risk to workers?  Okay, Dr. Gilbert, I guess you've got something to write here.  
                  DR. HENRY ANDERSON:  I'm not saying -- 
                  DR. KENNETH PORTIER:  Hello?
                  DR. HENRY ANDERSON:  I said I'm sorry I couldn't be more specific, and clear.
                  DR. KENNETH PORTIER:  No, but I think you raised a good point that needed to be discussed.  It ties into our discussion yesterday and it kind of identifies a method gap that we at least can point out.  It's not actionable on EPA's part today, but part of our task, as a committee, is to identify these things, get them documented.  
                  So, who knows?  Somebody may look at that, some smart academic may read this report and say, I know how to do this without a PBPK model.  Dr. Schlenk's shaking his head, no, that isn't going to happen.  Okay.
                  DR. HENRY ANDERSON:  Well, kind of the issue is if you're making a determine of unreasonable or reasonable risk, and you're saying in manufacturing there's no reasonable risks for those workers, doing a sum may make a difference.
                  DR. KENNETH PORTIER:  Okay.  Question 6.5.
                  DR. KATHERINE ANITOLE:  Question 6.5; please comment on whether the risk characterization has adequately described the assumptions, uncertainties, and data limitations in the methodology used to assess risks from 1-BP. Please comment on whether this information and the risk conclusions are presented in a logical, transparent manner and provide suggestions that could increase clarity in the risk characterization.
                  DR. KENNETH PORTIER:  Dr. Davies.
                  DR. HOLLY DAVIES:  This is a really big question that spans all of the risk assessment.  So, as I was going over the 14 pages of assumptions and uncertainties for all of them, it just seemed like -- this was in my preamble -- we've gone through a lot of this with the other risk assessments, pieces.  This question, it seems hard to answer for me having not dug deeply into these.  Not everyone who was assigned this question had dug into all of those, and I didn't copy down everyone's comments, even though I was trying to.  
                  So, we're going to need help from the rest of the committee on this.  I'm trying to put all of my clarity questions in Question 8 on my list for Michael for that, so some of these might overlap a little bit.  So I'm really not sure how to write this up.  Because I think some of these are in the other questions, and I'm not sure how to present this.  But I'm going to start.  Also, I got some comments from other committee members, but I didn't get them early enough to integrate.  So, this isn't as smooth as I would like.
                  I'm going to start with some other general questions.  Uncertainties and occupational exposure modeling and confidence -- it's nice that they explained the confidence.  That went through, and it was very qualitative.  More than one person had this comment that the qualitative confidence was good, whether it was high or medium, but it wasn't really clear because we weren't quite sure yet for what was high and what was medium.  
                  So, it was nice to have the qualitative discussion, but there was no quantification or even criteria around when it was high and when it was medium.  When I went through, I didn't write down which of these were actually in here and which of them I added because they weren't in here.  And then I'll add other ones that I don't have in here.
                  One of the uncertainties that I had seen was in the amounts manufactured, imported, and released that we've discussed before.  Also, the list of other occupational uses on page 100 that are not quantified.  One of the key uncertainties that they identified was whether they actually correctly identified the potentially exposed and susceptible subpopulations, which will be the next question, 6.6.  
                  I think there's a lot of uncertainties on number of workers in all of these.  I know I haven't had time to really delve into that.  Especially the ONUs, the exposures variable, and the studies don't stay where they are.  And the HHE is not typical exposure.  These are all things that the report mentions that are really good.
                  Issues that we've brought up about the post engineering controls when they look at that.  We've brought up issues on whether that's described, whether there's data for that, personal protective equipment.  All the models that they used, two issues -- that they were not scored for quality in the systematic review, and the other issue is the current criteria are so weak it wouldn't have helped because they would have all just scored high.  
                  An issue that I have not brought up before but that I've been thinking about is, at least in King County -- because I know the age of all the dry-cleaning equipment in King County -- many of the dry-cleaning machines are old, past their expected lifetime and they're more likely to be leaking.  If that's representative of other areas in the country, there could also be another exposure because there'd be more leaking.  
                  Of course, balanced on that is, as you know, 1-BP is not used very much.  In fact, it might not be used at all in King County.  But the age of the machines would be more leaking, balanced with this isn't very commonly used in dry cleaning.  
                  I do know that we have not seen 1-BP used as a spot cleaner in King County.  So that assumption in the draft risk assessment that it's not used much is supported by what we see in King County while we're inventorying all the spot cleaners used in dry cleaners.
                  Questions that have come up on the Westat Survey that been brought up about the age, the survey we have.  We've brought up uncertainties around consumer products.  I haven't been able to find any evidence of consumer uses.  This has been brought up by some other people also.  Issues around whether the consumer uses are really industrial uses, and then, of course, the fact that consumers sometimes use products that aren't consumer products.  They go out and buy industrial products and bring them home and use them.
                  One thing that's in the document is EPA refers to the SDSs as high degree of confidence, which I would definitely disagree with.  I don't have any confidence in information that I've gotten off of SDSs.  So, I wouldn't put that as a high confidence for that information.  I usually found information on SDS is not so reliable.  The phrase I really liked about the modeling inputs for the consumer exposure was that they were well established and commonly used.  They're what we normally use for risk assessment.  There's other things in all these pages about modeling and risk assessment has all the normal caveats about risk assessment.
                  One question I was actually wondering though, while I've just dissed SDS information, I was wondering -- SDSs do often have data about acute aquatic toxicity, and if some of the industrial products are high levels of 1-BP -- if there was data there about acute aquatic toxicity.  Something I thought earlier, this is on the cancer MOA that has been brought up again multiple times on that issue.  Dr. Eastmond brought up that ECA had determined the non-mutagenic MOA; and looking at the ECA that that was actually the registrants that had brought that up.  I just wanted to make that point, minor point.
                  DR. DAVID EASTMOND:  But they agreed to it.
                  DR. HOLLY DAVIES:  Some things, I'm not as familiar with.  The factors that were used in the environmental risk assessment, it always seems odd to me that you have to just divide by 5 here and divide by 10 here and call it good.  I think that would be, again, the well-established and commonly used.  Another question, the physiochemical properties and environmental fate characteristics that hasn't been brought up, they seem fairly good.  We don't have a lot of monitoring, but the characteristics don't seem like we would find it.  
                  So, is that enough information to go forward?  That's kind of a question for the committee.  What kind of precedent are we setting?  Let me see if I can find others.  And Mark sent me some stuff that I find hard to understand.  Other things that came up with just more explanation that came up in the previous questions around weight of evidence decision makings and choosing the studies that were used for the point of departure.  And then Mark said he liked the Bayesian models.  
                  Mark Johnson had to leave early, and he sent me something and it's kind of mixed in with some of the other things because, as I said, these are also mixed in with the other questions.  So, I think he had also brought some of these up earlier.  He liked the Bayesian model approaches.  I think that's the comment he most wanted me to make.  So, with that, I will leave it open to the rest.
                  DR. KENNETH PORTIER:  Dr. Anderson?
                  DR. HENRY ANDERSON:  Okay.  I would say, assumptions that were made are clearly stated.  Where I would say there is a problem is the uncertainties and the reliability of those assumptions are not clearly laid out.  Where again, not to beat a dead horse too much, but the full issue of respirator use, the assumption of 100 percent use, I have no confidence that that's the case, or similarly, that the expensive impervious gloves would also be used.  
                  So, while it's stated that you have the risk as respiratory protection is not used, it would be very helpful either to have it as a separate section to talk about the benefits of use of mitigation opportunities.  Again, the data limitations, EPA provides no data on respirator use.  There is some observational data that respirators weren't used, whether they would be used now or not.  There's really no data on that.  
                  Again, I would say the assumptions are clearly stated.  But why one would assume 100 percent use of protective equipment, the same as would go for engineering controls.  I think there's been plenty of discussion about all equipment that isn't quite so effective.  When they replace it with better equipment, the old equipment gets shipped to another user, so exposures (inaudible).  
                  Again, I would just raise the issue about assumption of 100 percent respirator use.  And not only that, but that they are in fact effective to the stated protection factor.  I think it was read out in one of the public written comments that with respirators, especially when you have cartridge respirators, they would be the ones you'd use in this instance that you could get breakthrough in the cartridge and not realize that.  
                  So, the effectiveness will actually deliver the level of protection that you're assuming.  You have a compound, and you're assuming it is 100 percent effective up to that level and that everybody is using them continuously.  I think those are unrealistic real-world expectations.  That probably needs to be stated somewhere.
                  One way to approach that, as I've said, recommendation would be to, I think, not get rid of that information in the report, but have it as a separate section on the impact of mitigation efforts.  Because I think it's very important.  So, have that information that, when this exposure environment is hazardous, some type of mitigation effort, such as respirators or gloves, needs to be implemented.
                  DR. KENNETH PORTIER:  Dr. Sheela, did you want to add comments?
                  DR. SHEELA SATHYANARAYANA:  No.  I agree with everything that's been said this far.  And I thought, overall, the descriptions were pretty transparent.  And then the few issues that are there have been brought up.
                  DR. KENNETH PORTIER:  This is Ken Portier.  When I looked at this question, I guess I thought this was more the 30 thousand-foot view rather than the detailed view.  So, if I could rephrase, it would say, please comment on whether the risk characterization in the document as presented on pages 188-253 adequately describes the assumptions, et cetera.  
                  I looked at this and I said, if I can go to a particular table, Table 4.6, and look at a number, can I trace it back?  Can I trace back the point of departure?  Can I trace back the exposure?  Can I come up with the same number?  Are there discussions about uncertainties and assumptions on each of those components in the document?  And I couldn't find any places where we didn't have that.  
                  Now, I agree that sometimes we don't agree with their assumptions.  We don't agree with their uncertainties.  But in terms of what they've written here and kind of linked together as a logical presentation on risk characterization, I had no problems with it.  To me, that's the answer to this question.  
                  We've already pointed out a lot of deficiencies in the pieces that go into this.  This is the question of, when EPA put it together in the document, is something missing?  Or can something be better clarified?  And I guess that was the way I was looking at the question.  Dr. Eastmond?
                  DR. DAVID EASTMOND:  Just to clarify, Holly is correct.  The information I found was on ECA website.  But the ECA information on the ECA website is largely put in by the applicants.  Because they tried to flip the burden of proof, so the registrants had to evaluate the studies and make their own assessment.  And then, periodically, it gets evaluated by ECA itself.  But there's no evidence that ECA actually reviewed these.  So, just to clarify that.  Sorry I misled people.
                  DR. HOLLY DAVIES:  I had thought earlier you had been -- because you talked about the WHO model, the World Health Organization model.  But I had thought you'd also just meant that -- we'll talk about it later.
                  DR. DAVID EASTMOND:  On that one -- the WHO model, I was chair of that committee.  So, I'm pretty familiar with it.  That's the guideline.
                  MR. ALAN KAUFMAN:  Yeah, this is Al Kaufman.  I think we had already decided that it was kind of a judgement call and you could kind of go either way, depending on which way you looked at the data.
                  DR. KENNETH PORTIER:  Dr. Blystone.
                  DR. SHERI BLYSTONE:  I just feel the need to comment on safety data sheets since it was brought up.  Certainly, under OSHA, any hazardous ingredient in a product must be disclosed.  Most often, that is in ranges.  And I think that's the data we were talking about here was whether or not it's in a product and how much.  
                  It's probably going to be reported in ranges.  It is hazardous, so it will be reported over 0.1 percent because it's a carcinogen.  Those ranges will be -- and if you take the high-end value, it's going to be overly conservative for that.  So, I think there's limited ability there to claim trade secrecy for ingredients.  That is possible from that percentage.  But I think the information on the safety data sheets and hazardous ingredients is pretty straightforward.  
                  Your question about the eco tox, the aquatic toxicity -- of course, under OSHA, they will defer to EPA for aquatic tox.  So, we are not, in the United States, required to classify for environmental hazards.  However, many companies do.  And they put the information in Section 12, like you said.  So, they would classify for GHS hazard aquatic.  And most likely, those people are going to go and look at publicly available data like the ECA database and use that information to write their classifications.  So, it's unlikely that there's no data somehow hiding in a safety data sheet.
                  DR. HOLLY DAVIES:  I did just want to say, whenever I've worked with companies on specific products, we have often gone to the SDSs and worked with the company and found they're not actually correct with what's in the product.  And it's been an out of date SDS or the SDS isn't totally correct, so they're not as good as the scientific paper.  If I was ranking things on high, medium, low, I wouldn't put them up there without a peer review journal from sources of information.
                  DR. SHERI BLYSTONE:  But a useful source, I wouldn't call it unacceptable.
                  DR. HOLLY DAVIES:  Well, I think if it were identifying that 1-Bromopropane was in the product on an SDS that that would be fairly reliable.
                  MR. ALAN KAUFMAN:  This is Al Kaufman.  Listening to the conversation between the two of you, it sort of triggered a thought.  I wonder if this rises to the level of requiring hazard labeling under the Federal Hazardous Substances Act for consumer products.  That might be something we want to think about as well.
                  DR. HOLLY DAVIES:  I also wanted to clarify, when I said earlier something about not having seen 1-BP in consumer products, that I actually went out and searched for it.  I have actually searched all of my usual places where I find data on consumer products and didn't find it.  So, it's not just that I didn't see it presented here.  It was that I went and looked myself in my usual places and didn't find any additional data.  After actively searching, I didn't find it either.
                  DR. KENNETH PORTIER:  I was looking for it in nonflammable hairsprays.  But I couldn't find it.
                  DR. HOLLY DAVIES:  I didn't look in my basement, but I was looking more like in the Nordic databases and the SPIN database and other places more like that.
                  DR. KENNETH PORTIER:  Go back to Question 6.5.  Again, I really do think this was a 20 thousand-foot question because I think it's key for this document.  If this section is not logical, not transparent, then the next section where EPA goes to its risk determination doesn't follow logically and transparently.  
                  So, I just really want to make sure you guys are comfortable with this risk characterization section, at least on the human health.  We had a conversation on the eco risk characterization yesterday, but at least on the human health.  Mr. Kaufman?
                  MR. ALAN KAUFMAN:  Is the question, are we comfortable with how it was explained?  Or are we comfortable with -- do we agree with it?  Because I think you nailed it when you said we may disagree with the characterization, but in general, it was very well explained as to how they got where they got.
                  DR. KENNETH PORTIER:  In previous questions, we have identified issues with uncertainties and assumptions and models and data gaps.  So, given our answer to the previous questions, the way they presented the risk characterization here.  Does that look okay?  Not necessarily every number we agree to.  That's not what I think they're asking us.  They're asking us whether this way we presented it.  Logically, I think of it as linking backwards from these particular tables to the hazard and exposure discussions, so that it comes together logically at this point.
                  DR. HOLLY DAVIES:  This is Holly Davies.  After this, there was all the pages of the assumptions and key sources of uncertainty.  So, I still think it's -- when you're saying the 20 thousand view, I didn't think it was just linking the tables in there.
                  DR. KENNETH PORTIER:  So, you're saying, I want to add something to that section, right?  For clarification.
                  DR. HOLLY DAVIES:  No, no, no.  I'm still saying that this section should be reflective of everything is what I was seeing the question as.
                  DR. KENNETH PORTIER:  To me, it's integrated.  You know?  They looked at a whole bunch of conditions of use and they did a whole bunch of scenarios, and then they kind of came down to one number which is going to be the number on which they're going to base their risk determination.  And the question is, have they made the case down to that number?  Is that case followed logically?  That's the point I'm asking to get a feel for the panel.  Dr. Barone, am I reading this question correctly?
                  DR. STAN BARONE, JR.:  You are.  That is the intent of the question.  That's what the intent of the risk characterization is.  Have we integrated the information, told you what the risk drivers are, made it clear how we got to those decisions?  And I think you have pointed out some areas where we could be more transparent, some areas where we might consider alternatives.  
                  But for the most part, other than our prior discussion about a summary table of all the risks, which we did not have time to implement into this particular effort, I hope you will see that in the very near future in some of our upcoming assessments which will also be addressing this particular question.  Go ahead.
                  DR. HOLLY DAVIES:  This is where I didn't know if this is this question or it goes into Question 8.  Because, for me, it was really weird to see this whole section with everything together rather than see them integrated into each part, if that makes sense, to have all the assumptions separated out into one section.  Cathy knows what I'm saying.
                  DR. CATHERINE WILLETT:  Yeah.  I was going to mention that, but I guess it goes in Section 8.
                  DR. HOLLY DAVIES:  Question 8.
                  DR. CATHERINE WILLETT:  But trying to review the section is really difficult, because you have to keep flipping back and forth.  You know?  So, I had all my piles all over the place because there's a table in one section, then you have to look at a table in another section, then you have to go to the appendices.  I mean, I know it's really hard to organize this complexity of information in a way that always makes sense.  
                  But I think maybe, by putting these discussions at the end, like in the presentation, that was really great because you had all the information about a certain kind of exposure in one place.  You went through it all at once.  Of course, this is all trial and error trying to get it to be great.  But maybe that organization might.  But I guess this is Question 8, right?  Can I bring up a couple other things while I'm talking?  Okay.
                  DR. KENNETH PORTIER:  I just want to clarify.  Dr. Davies, one of the things that you're pointing out is Section 4.3, Assumptions and Key Sources of Uncertainty.  They've kind of integrated that.  And that didn't make much sense to you.  You're kind of saying, well, why didn't we have that under each table?
                  DR. HOLLY DAVIES:  Well, reading through that, it was just part of everything.  And I think a lot of it is in here, and maybe it's duplicated.  It could totally be duplicated so that people who like it in see it in, and people who like it out have it out.  Because some people like it both ways.  
                  DR. KENNETH PORTIER:  Yeah.  You see, as I was thinking about your comment, I'm thinking, well, if I'm looking at one of these tables and there's a number, it would be nice to be able to click that number, go back to the previous exposure or hazard table, look at that number in that context, look at the uncertainties and discussion there.  Why do I have this global uncertainties and discussion when you do the risk characterization?  I guess I haven't read that chapter, that section, clear enough to see what its purpose is.
                  DR. HOLLY DAVIES:  I read it because I thought that's what this question was about.  Here we have all the uncertainties, assumptions and uncertainties for everything.  It did adequately describe it.  And it was, in general, really well done.  As you said, it has nice qualitative, but also integrated in here was a lot of this.  I, of course, don't need the hyperlinks as much because I print things out like a dinosaur.  And different people are going to want, but other people do what the hyperlinks and different styles.
                  DR. KENNETH PORTIER:  I think it does bring the question though.  What is the purpose of Section 4.2?  What are you really trying to accomplish there?  Because it's very repetitive.
                  DR. HOLLY DAVIES:  Section 4.2 or 4.3?
                  DR. KENNETH PORTIER:  4.3, the assumptions and uncertainties section.
                  DR. HOLLY DAVIES:  Which goes to question 8.
                  DR. KENNETH PORTIER:  I'm sorry.  4.3.
                  DR. HOLLY DAVIES:  It goes to Michael tomorrow.
                  DR. KENNETH PORTIER:  Okay.  But the whole point is, if that confused you, that's a point that you need to bring up.  Why was this here when we've had all this distilled discussion in the hazard section, and in the exposure section where you've had all this discussion about uncertainties and assumptions there?  What is the point of trying to integrate and squeeze that down?  Now I see where you were coming from.  I just wasn't sure where you were coming from with your comments.  Dr. Barone?
                  DR. STAN BARONE, JR.:  If I may, again, this is something we struggle with too.  As a risk assessor and a team lead looking across these assessments, one of the challenges you see when you're reading a lot of these documents is risk assessors will just create laundry lists of all the uncertainties.  And at the end of the day, it's like, oh, wow.  What really matters?  So, what we're trying to get to in this section and the risk characterization, what matters?  
                  From your sensitivity analysis, which of these uncertainties or which of these limitations, what of these strengths, what matters?  Versus back in the exposure section or in the hazard section where it seems like there's just a listing of uncertainties.  So, we're really trying to bring this together.  
                  Again, I think some of this conversation actually helped solidifying how we should ask this question more clearly.  But that's the goal.  That's what we're trying to get to, is have we really outlined the key uncertainties, the key limitations?  What is overestimating or underestimating, impacting the risk estimation process?  That's our goal.  
                  I think the latter questions where we actually moved them from the beginning really talk about the linkages between exposure, fate, the uncertainties, and all the prior sections, and bringing those forward.  Are they being captured?  Are they being captured in the supplemental information?  Are they being captured in the appendix and being brought into the body of the assessment?  
                  We've had some challenges with that, and you've called those out in previous assessments.  So, those are two different dimensions to the same problem.  Again, I think you guys have been struggling with that and giving us some good input.  I do have a few clarifying comments or questions.  I'll come to those.  
                  DR. KENNETH PORTIER:  Go.
                  DR. EDWARD PERKINS:  I figure I'll have to say something.  
                  DR. KENNETH PORTIER:  Yeah.
                  DR. EDWARD PERKINS:  I totally agree with what you're saying.  I think that's the way I viewed it, but I noticed there was often a lot of uncertainties associated with each calculation in the earlier sections.  But they weren't really described what the potential impact of tweaking that would be, overrepresenting and underrepresenting.  You noticed a lot of that when we were talking about different areas, how it might affect it.  
                  But I did notice in this section of the 4.3 it was really talking about potential impacts of those different things.  It was a little easier to see how the different elements would affect the overall risk assessments.  I think that was valuable right there.  It might do, to even clarify it farther, to highlight that a bit more.  Just bold out something simple like that or just highlight it.  Because there is a lot of text.  But I thought it was very useful and well explained.  
                  DR. KENNETH PORTIER:  Something you said came to mind.  Key uncertainties, key assumptions.  And to me, it's almost like each table needs two bullets: key uncertainty, key assumption.  What do we think is the most uncertain about the numbers that are in Table 4-53?  To me, that's probably more useful to somebody who's reading this than going to Section 4.3 and reading a global summary.  
                  I don't know if anybody's going to try to take these 53 tables and do a gestalt feeling for that, and then think in terms of global uncertainties.  They're going to look at numbers and say, why should I not trust this number?  Or what do I need to know as I assess that number?  That would be my feeling.
                  DR. HOLLY DAVIES:  This is Holly Davies.  I appreciate that in the consumer exposures when I was looking at those models.  It said, these parameters are what affect the outcome the most.  And it was good to have that in -- I can't remember if it was in the body or the appendix.  But as I was looking at those, it was good to have that and to see which would affect it the most, and that's why you were varying those.  
                  Because I actually found things in here that I hadn't seen, like on the potentially exposed and susceptible subpopulations, right?  I thought I had seen everything you had written about that, and then I found new stuff.  I was like, oh, look at that, I missed this.  Because it was kind of tucked.  But that's that style.  And again, some people are going to be like, did EPA look at the uncertainties?  I want a chapter just on uncertainties.  And they wouldn't see it if you integrate it.
                  DR. KENNETH PORTIER:  This is Ken Portier.  To me, the uncertainties and assumptions were part of the context for each of those tables.  And to separate them out and have a discussion in a whole separate chapter kind of disengages that content.  It's like taking the metadata from your data set and putting it over here.  Here's the data.  Oh, if you want the metadata, here it is.  But to me, this table means very little unless I know something about assumptions and uncertainties.  
                  But I was willing to link back and look at the numbers and see the previous discussion.  But it's fair for us to ask EPA to find the one key assumption, the one key uncertainty for each of these tables and highlight that for us as either a footnote or a part of each of those -- they're each a sub-sub-section of their own.  Right?  There's a lot of tables in this chapter.  By the way, I really do like figures 4-2 and 4-1.  Those are cool.  Whoever did those, good job.
                  UNIDENTIFIED FEMALE:  What figures is it?
                  DR. KENNETH PORTIER:  4-1 and 4-2, the central tendency inhalation cancer risk estimates with the nice, bright, red line where you could look down.  I mean, the only suggestion I would make is rank them from the most to the least, so people can say what's the most important one that I need to worry about?  But that's a minor thing.  
                  Gosh, what is this section number here?  4.2.2.2.  So, you might have a discussion in each of those ".2.2.2" sections.  But that Section 4.3 just didn't catch my attention when I read it.  But obviously, it caught Dr. Davies' attention.  So, it had some value.
                  DR. HENRY ANDERSON:  It was very helpful to have the color specifics in the risk section.
                  DR. KENNETH PORTIER:  Any additional comments before we go onto 6.6?  Oh, I'm sorry.  Dr Doucette.
                  DR. BILL DOUCETTE:  Bill Doucette again.  I'm not sure it belongs here.  I guess it does, because of the way the discussion went.  But I know there was a comment across the table about how much nicer it was for the introductory presentation, how well organized it was.  And I really appreciated at the end of each one of the sections, because you have to break this document into readable sections, where they went through and talked about essentially the strengths and limitations within that section.  
                  I missed the second get-together, but the first get-together with PV29, that was also brought up how much clearer that arrangement was in the slide presentation than it was in the format.  And I don't know if that's the appropriate model.  Again, I think that's open for discussion.  But I like the fact, at the end of each section or the beginning of each section to warn me what I'm looking for, that I have that table of strengths and limitations.  
                  I found that more useful to me than a long summary of trying to put everything all together.  And that's just my opinion.  I'm not sure everybody else feels the same way.  But I think the consensus is that the summary presentation that you folks give is really very helpful.  And maybe that's the format for the document.  
                  DR. KENNETH PORTIER:  Dr. Barone?
                  DR. STAN BARONE, JR.:  This may sound pedantic, but with the risk evaluation rule, we're actually asked to specifically call out in the risk characterization assumptions, uncertainties, and data limitations.  So, you can imagine why we have a template and an outline that actually has specific sections, just like we have an aggregate/PESS section in the risk characterization now.  There's a logic to this.  
                  I do appreciate and we definitely are in the process of adaption to your comments about the format.  We're also discussing, now, the template for the next 20.  So, stay tuned.  I think some of your comments and, again, lessons learned will be incorporated into that.  I have a couple of clarifying points.
                  DR. KENNETH PORTIER:  Let me address that.  One second.  We'll push back just a little bit.  So, while you have to have a Section 4.3, it doesn't have to be a reiteration of a lot of what we've covered before.  It could be much more of a global summary, a narrative about what we think is really the critical things that's going on here.  And the way I read this section right now, it's a lot more repetitive of what we've seen before.  So, maybe our recommendation is to think about distilling it a little bit more than it's currently distilled.
                  DR. STAN BARONE, JR.:  Thank you for that.  Yes, exactly.  And that is the goal.  Again, the instructions to our teams are to make it more of an integrative section.  Again, you'll see the more we're including sensitivity analysis and uncertainty analysis, it becomes a more integrative section.  Again, that's where we're trying to head.  I think you've already commented about the improvements.  Hopefully successive approximations.  We'll get there.
                  One of the couple of clarifying issues on comments that were made, SDSs -- going back to the safety data sheets -- are not a source for our eco tox evaluations.  We're not using those for eco tox evaluations.  In fact, trying to figure out what eco tox study was actually being cited or referred to in an SDS is more than a trivial challenge.  
                  On the models, there seems to be confusion about systematic review of models.  With Section 26-H, when we talk about data quality and the quality of the science, it's peer review and/or systematic review.  It doesn't have to be both.  In fact, many of our models are not going to be appropriate for systematic review.  The inputs and the science that goes into the models, but really the models are peer reviewed.  And we use peer review, like you all, in what we're doing here to get at the evaluation and validation of the models, not systematic review.  
                  So, I think interpreting that every model has to go through systematic review, it's just a different process.  It's for a different kind of product, so I want to be clear about that.  Then there was some discussion about the ECA database and the ECA information.  I want to underscore the issue that it is the sponsor's information.  ECA does a quality review.  Their cap is 5 percent.  Their data quality is capped at 5 percent of that information that's submitted.  They would like to do more.  But they're basically only able to check up to 5 percent.  They don't have the capacity to check more than that, but that is where they stop.  
                  So, they're not going into every file, every study, or every dossier to review that for quality, which is different than our requirements.  We're sort of in a different place statutorily as far as data quality review.  Just the facts.
                  DR. KENNETH PORTIER:  Dr. Doucette?
                  DR. BILL DOUCETTE:  Bill Doucette.  Sorry.  Just a follow-up clarification.  So, in terms of the models, for example, the EPI Suite values that you gave in the physical/chemical properties, for example.  You rank them as high, and that was exclusively because we had the peer review for that particular model.
                  DR. STAN BARONE, JR.:  That's right.
                  DR. BILL DOUCETTE:  Okay.
                  DR. STAN BARONE, JR.:  That's right, and those particular values are within the model applicability domain.  Again, there are certain data quality criteria, but essentially because of the peer review value of the model and the model's output.  There are cases where we're using models and we're stretching the models.  We may be at the limits of the applicability domain and we're going to tell you, in many cases, where we get to those cases.
                  DR. KENNETH PORTIER:  Any additional comment on 6.5?  We have a second mid-afternoon break scheduled, but we really only have one more question to address.  So, I'm going to look at the panel and see.  Do you want to take a break?  Or do you want to just soldier on?
                  UNIDENTIFIED FEMALE:  Soldier on.
                  DR. KENNETH PORTIER:  Soldier on.  Okay.  Please read question 6.6.  Bill, put your flag down.
                  DR. KATHERINE ANITOLE:  Question 6.6; please comment on whether the risk characterization has adequately identified and characterized the "potentially exposed or susceptible subpopulations" -- PESS -- based on a thorough review of the reasonably available 1-BP exposure and health effects information on both potentially exposed and biologically susceptible subpopulations.
                  DR. KENNETH PORTIER:  Dr. Davies?
                  DR. HOLLY DAVIES:  I just noted that this was new to TSCA and was updated in 2016.  This was put in in several places, including in Section 6, for these risk evaluations to include these populations and these determinations.  And it had a definition of what that means, including infants, children, pregnant women, workers, and the elderly, but not a great definition.  EPA's working, as with so many of these, exactly how to do this. 
                  I wanted to acknowledge there are several challenges in identifying and characterizing the potentially exposed or susceptible populations, including the lack of definitions in TSCA and the lack of data on these groups required under TSCA.  Also, clearly, if EPA included everyone possible from the definition in TSCA, then everyone would be included.  Because if you include children, reproductive-age adults, and the elderly, well, we've really just now included everyone.  So, that wouldn't work.  
                  I think the goal is to identify which subpopulations are likely to be more effected by this chemical and address that risk.  Ideally, one would take into account both intrinsic factors such as age, genetics, and disease, and external factors such as poverty, racism, and exposure to other chemicals.  Some communities have different exposures, not just higher exposures, such as tribal members.  And children are particularly vulnerable because they take in more air and water per kilogram of bodyweight.  They're still developing, and they interact with items differently than adults do.  In addition to protect potentially exposed or susceptible subpopulations, the aggregate exposure should be estimated.
                  In this document, again -- we've mentioned this a couple times -- in the presentation we had the first day, the slide had a nice summary of this.  But in the draft risk evaluation, I found kind of skipping through.  Like, a little bit here and a little bit here and a little bit there.  And it was harder to pull it together.  EPA listed what they did to assess exposure for workers and consumers by age group to break it down into reproductive age that we've talked about before, and children were included as bystanders, but not in the occupational.  We've had this discussion before.  
                  EPA noted the importance of acute exposure and the critical window of fetal development but didn't mention that as a potentially susceptible subpopulation.  On the example of the genetic polymorphisms in CYP2E1, expression that had been linked to alter disease susceptibility is a great example of an intrinsic factor that's a susceptible subpopulation.  Unfortunately, there's not a lot of data as to how many people there are and who they are.  But that is a really good example of a susceptible subpopulation that they noted.  
                  There was a section in that section that we just talked about on the last question that says EPA identifies the following as PESS.  But I had missed the first several times on the run through, because it was tucked in there, and the aggregate and sentinel exposures.  
                  So, my suggestions were that this would be improved by putting it all in one place with a more clear statement of, these were the groups that were identified.  Because TSCA does say it's identified by the administrator.  This is what was identified, and here it was addressed more like slide 78 in the presentation.  
                  And then, even if EPA does not intend to address the risk to TSCA, which is another conversation we've had before, the aggregate exposure should be estimated to fully characterize the risk so that we know, what is the full risk that this subpopulation is getting?  Even if it's going to be addressed through the Clean Air Act or RCRA, or some other law, what is the full aggregate exposure that the subpopulation is getting?  And if the data is missing for these subpopulations, appropriate uncertainty factors a protective defaults can be used to be protective.  
                  The other comment I had gotten for susceptible subpopulations was for consumer uses and actually on a cleaner for implantable prosthetic devices and that being a, perhaps, susceptible subpopulation, a population that would be more exposed to this although we don't have estimates on the size of that subpopulation.
                  DR. KENNETH PORTIER:  Thank you, Dr. Davies.  Dr. Anderson?
                  DR. HENRY ANDERSON:  Well, I think I've said most of my points.  I think it would be helpful if they summarize what are the groups upfront.  It seems, throughout the document, exposed or susceptible populations, at times, are not the same.  So, that's one of the issues that just makes the document longer if you have to list a long set of individuals.  
                  The other group that has not taken mention in these is individuals with chronic diseases, and I would just use the general term "chronic diseases" rather than go through people with neurologic disorders or diabetes or hypertension.  Some of these compounds, those individuals are particularly sensitive to them, especially in the workplace.  I would just add chronic diseases.
                  DR. KENNETH PORTIER:  I don't have anyone else listed as a discussant on this question.  But is there anyone -- oh, Dr. Blando.  Good.
                  DR. JAMES BLANDO:  It's Jim Blando.  I just want to quickly mention, I have the same thought that Holly did about this definition which I read for what a PESS population is.  But, especially being hired in the general population, it seems to me that our discussion of the HAP petition for people who, for example, live in co-located buildings with industrial operations, it would strike me that they would meet that definition of PESS.  
                  So, although I know they're not included in this risk assessment, it strikes me, as Holly had mentioned, that those folks would meet that definition.  That's just the comment I would make.
                  DR. HENRY ANDERSON:  Yeah.  That would hold for neighborhood people as well.
                  DR. KENNETH PORTIER:  Dr. Anderson, we didn't catch the beginning of that.  Could you say that a little --
                  DR. HENRY ANDERSON:  That would probably also pertain to people living in the neighborhood as a potentially exposed group.
                  DR. HOLLY DAVIES:  Yes.  And they also (inaudible) fence line.
                  DR. KENNETH PORTIER:  Dr. Cobb, and then Dr. Barton.
                  DR. GEORGE COBB:  I thought about introducing this earlier.  In poking around the Albemarle website the other day, looking for some of their data, it seems that 1-BP is marketed as a cleaner for implantable prosthetics.  And I don't know if that's been assessed at all, which will maybe throw a real curve ball at you. 
                  MR. ALAN KAUFMAN:  Can somebody explain what an implantable prosthetic is?
                  DR. GEORGE COBB:  I'm going with a hip or a knee replacement.  It may be something different.  But it says, on the website, implantable prosthetic devices.
                  MR. ALAN KAUFMAN:  But once you've implanted it, you wouldn't continue to clean it.
                  DR. SHERI BLYSTONE:  It also would be regulated by FDA in that case.  But it probably had the similar scenario to decreasing in general.
                  DR. GEORGE COBB:  Well, it's still potentially an exposure to a population that would be at risk.
                  DR. HENRY ANDERSON:  It would be impregnated into the hard plastics.
                  DR. GEORGE COBB:  Excuse me.  George Cobb again.  Not a population that would be at risk, a population that may be susceptible because they're undergoing, most likely, a surgical procedure.
                  DR. SHERI BLYSTONE:  I can't imagine it's still on the prosthetic when it gets implanted.
                  DR. KENNETH PORTIER:  Dr. Barton.
                  DR. CHARLES BARTON:  Thank you.  The lung cancer endpoint risk assessments were based upon a decrease in live litter sizes in the rats, and it was thought to occur from exposures possibly varying between ovulation to implantation.  It's not known whether it was exposure to the males or the females or if it really was.  In a way, it hasn't really been demonstrated if it wasn't really a developmental effect where that started initiating with the same number of implants.  But they did not stay.  They died away as simple multicellular fetuses.  
                  If that's the case, the developing fetus would be a susceptible subpopulation.  As it's written, if we go along with it's actually effects on the F-zero generation, then the susceptible subpopulation would really be males and females of reproductive age.  As it is, what's been written really hasn't made a strong enough argument to say that it's actually a reproductive effect and not effects directly on the fetus.  Thank you.
                  DR. KENNETH PORTIER:  This is Ken Portier.  Dr. Davies, something you said just kind of got me looking at the outline again.  You're right; in Chapter 2, susceptible population is a main heading, all caps.  Then you get to Section 3, and it's four levels down under dose response.  And then level 4, now it's a third level.  So, I think you need to kind of take a step back and look at your outline and decide if this is going to be a major 2.2, 3.2, 4.2 kind of heading and integrate all that stuff.  
                  I understand why, for example, you have that susceptible population might be a subsection on dose response.  But for emphasis and for the reasons that EPA indicated that it's called out specifically in the TSCA regulations, you may want to consider, in the general format, pulling that up to a main subheading and just do your integration that way.  Any other comments?  Dr. Gilbert.
                  DR. KATHLEEN GILBERT:  I just want to say, I understand this kind of characterization is really difficult.  But as far as the potentially exposed population, once again, it comes down to this unresolved issue as whether or not we have a baseline exposure and majority of people, including pregnant women, which, once again, as Chuck would say, is putting fetuses at risk.  
                  So, until we resolve that, do we have to consider this as a baseline exposure that would have to be added into other kinds of exposure for the risk characterization?
                  DR. KENNETH PORTIER:  This is Ken Portier.  You bring up a good point.  I just zipped back to Section 2.2 and there's no discussion of that, the biomonitoring implications for susceptible populations at the exposure section.  So, we probably should make a recommendation somewhere that they include that.  And we may have done that yesterday.  I have all of yesterday integrated.  
                  But those of you who had sections on that exposure might want to think about how we might want to incorporate a recommendation, put something like that.  Section 2.2 of the discussion, susceptible populations exposures and indications thereof.  Any additional comments on Question 6.6?  Dr. Barone, questions?  Issues?
                  DR. STAN BARONE, JR.:  Just some clarifications.  TSCA, again, focuses on conditions of use.  So, it's not the broad rush of everything in the population.  It's about the conditions of use.  And we have divided up industrial/commercial general population and consumers into the three human health receptor subpopulations.  Within those are additional susceptible populations and potentially exposed.  
                  So, to the degree you all can provide us comments about those linkages and where we might be lacking on those linkages, we've tried to make clear we do not expect children or infants to be affected by all these conditions of use.  We've tried to be clear about when we do think children -- I'm using children as an example -- will be affected by the conditions of use, and then what context, in an acute or in a chronic scenario.  
                  Your comments that you provide will provide us with impetus to look further or expand those context of acute versus chronic if chronic gets expanded.  I think we talked about at least one condition today where consideration for a chronic exposure or subchronic exposure might be expanded.
                  There's another context for susceptibility -- Dr. Davies and others have brought this up -- and that is preexisting disease.  Dr. Anderson brought this up too.  If you have additional information that's relevant to 1-BP exposure, that would increase the variability in response or increase the vulnerability of the subpopulation, whether it's workers, consumers, or others.  Because, a preexisting disease, we very much would like to include that.  
                  We did that in our prior assessments, HBCD and fatty liver disease and metabolic disorders.  If there's specific information here or that we're missing that -- Dr. Eastmond's waving his hand -- it would be really helpful to get that kind of additional information incorporated into the evaluation.
                  DR. KENNETH PORTIER:  Dr. Barton, and then Dr. Eastmond.
                  DR. CHARLES BARTON:  Thank you.  Since it appears that cytochrome P450 2E1 is embogged, and it could increase toxicity and you mentioned underlying diseases, alcoholics could be a susceptible subpopulation.  It's estimated that between 10 to 15 percent of the population have alcohol abuse disorder, according to DSM-IV or whatever they're out with now.  Thank you.
                  DR. STAN BARONE, JR.:  We did call that out to some extent.  If you have additional information please, be free in advising us.
                  DR. CHARLES BARTON: I did appreciate that it was called out, but I've never seen that category of the population that's considered a susceptible subpopulation.  But there would be.
                  DR. STAN BARONE, JR.:  If I can mention, you did bring it up in the document, although it's more indirect.  Because you said that those exposed to solvents -- alcohol, ethanol, or drugs -- would be at higher risk of cancer.  But it wasn't directly linked to the 1-Bromopropane.  
                  The other thing is, as mentioned, CYP2E1 is inducible.  And in alcoholics, it's believed to have higher levels that make them at higher risk.  It's also altered in diabetics and some other ones.  If you want, I can provide some information on that because it's under a variety of different controls within the body and those could influence susceptibility. 
                  DR. STAN BARONE, JR.:  The more that you can help us define the range of variability and response -- actually, you laugh.  But this becomes important information later on when we enter into cases where we do impact assessment, qualitatively and quantitatively.  This is the assessment phase.  If we move into another phase, this data serves as the foundation for that further assessment work.  
                  So, the more that we can build that scientific foundation, the references, the data, the understanding of those impacts, the easier that effort is later on, which is not a trivial issue if you've gone through rulemaking before.
                  DR. KENNETH PORTIER:  This is Ken Portier.  Before we go on, Dr. Bennett, on discussions on Question 3, I just realized that in Section 2.2, which is the potentially exposed or susceptible subpopulations, there's really no reference to the eight consumer uses at all.  It's all about workers and there's nothing about consumer uses, and I kind of missed that.  The question was not asked, and we didn't notice it.  
                  But coming out of this conversation, I think we really should add something, somewhere in 3.1, 3.3, where we mention the fact that susceptible populations for consumer conditions of use are not addressed at all in that section.
                  DR. STEVEN BENNETT:  I did mention that in passing yesterday, but I don't think you really got the full emphasis of where it was lacking or where it should be incorporated into it.  But yeah, I completely agree that there should be an expansion within that earlier chapter to incorporate those, if that is the case.
                  DR. STAN BARONE, JR.:  That will probably come up in Question 7 tomorrow.
                  DR. HOLLY DAVIES:  Ken, what were you suggesting for the consumer uses?  Were you suggesting that we put it in our answer?
                  DR. KENNETH PORTIER:  No.  I'm saying, in our discussion on Question 3, which was about consumer conditions of use which is Chapter 2 in the document, that Section 2.4 has no mention of consumer conditions of use and susceptible subpopulations.
                  DR. HOLLY DAVIES:  I guess I was specifically not mentioning it, because I knew this question was coming up.  And I like the idea of putting it in one place and not having it spread in multiple places.
                  DR. KENNETH PORTIER:  I'm an in-the-moment kind of person when I'm over here, so I wasn't looking at tomorrow.  I was looking at today.
                  DR. HOLLY DAVIES:  So, we can do it either way.  That was just my thought process.
                  DR. KENNETH PORTIER:  No, no.  We'll bring it up again tomorrow.
                  DR. KENNETH PORTIER:  It'll come up.  Yeah.
                  DR. STEVEN BENNETT:  This is Steve Bennett.  I think my recommendation is to incorporate, into the earlier section on the consumer use, the exposures in there and the incorporation of it.  In the PESS issues, that would be identified within the consumer uses that are in there.  And then in this section, there will be, whether that works into the risk consideration.  So, pair the two others.  It'll be two separate entities but two different considerations.
                  DR. HOLLY DAVIES:  I'll send you my notes.
                  DR. STEVEN BENNETT:  Thank you.
                  DR. KENNETH PORTIER:  Okay.  Dr. Blystone?
                  DR. SHERI BLYSTONE:  This is probably more on the Question 8 thing, but to satisfy both -- and I think, Dr. Portier, you mentioned maybe condensing things -- there's an opportunity there to hyperlink back to the other section do you don't have to repeat it.
                  DR. KENNETH PORTIER:  I think we've kind of worked all of Question 6 over pretty good.  I have us being right at 4:00.  And given that the panel was not quite ready to proceed with questions 7 and 8, I'm looking to the panel to end an hour early today and start again tomorrow morning at 9:00 and tackle the two questions.  I think it's a pretty good bet on our part that we will be done by noon tomorrow.  I think we'll be able to cover the two questions in three hours.  
                  That's my hope and my goal.  And I expect Question 7 people to work hard this evening because they've asked for this dispensation.  I'm going to expect a wonderful presentation in the morning.  Thank you.  With this, we'll end for today.
                   (MEETING ADJOURNED FOR THE DAY)
                        OPENING OF THE MEETING - DAY 3
                  
                  DR. KENNETH PORTIER:  Good morning.  Welcome to Day 3 of this meeting of the Science Advisory Committee under TSCA, and we're going to continue our discussion of the EPA questions.  
                  
                        FOLLOW UP FROM THE PREVIOUS DAY
                  
                  DR. KENNETH PORTIER:  I wanted to open this morning's discussion to see if there was anything left over from yesterday that someone wants to bring up.  We need final discussion, final comments on Questions 1 through 6.  Dr. Eastmond.
                  DR. DAVID EASTMOND:  I just want to make a correction, I guess.  The three target sites are -- two of them are clearly at the site of exposure and one in the large intestine.  Because this is an inhalation exposure, it's going to be more indirect.  You can have it through grooming or something but, it's not a direct site anyway unless you clarify that.  I realized that in the middle of the night.
                  DR. KENNETH PORTIER:  Anyone else?  So we're going to move onto the general risk characterization questions, Question 7.  And I'm going to ask that 7.1, 2, and 3 be read in together, and then we'll take 7.4 separately because 1, 2, and 3 kind of go together as a group.  And I'm told that the group responding to this have their act together, and they really want to really jump into it.
                  
                               CHARGE QUESTION 7
                  
                  DR. KATHERINE ANITOLE:  Question 7. General Risk Characterization.  This question pertains to Sections 4.1, 4.2, and 4.3 of the Draft Risk Evaluation.  EPA concluded that 1-BP does not present an unreasonable risk of injury to the environment under the conditions of use.  EPA concluded that 1-BP presents an unreasonable risk of injury to human health under the conditions of use.
                  Question 7.1 Please comment on the objectivity of the underlying data used to support the risk characterizations and the sensitivity of EPA's conclusions to analytic assumptions.
                  Question 7.2 Please comment on the characterization of uncertainties and assumptions including whether EPA has presented a clear explanation of underlying assumptions, accurate contextualization of uncertainties and, as appropriate, the probabilities associated with both optimistic and pessimistic projections, including best-case and worst-case scenarios.
                  Question 7.3 Please provide information on additional uncertainties and assumptions that EPA has not adequately presented.
                  DR. KENNETH PORTIER:  Thank you, Dr. Anitole.  The lead for the discussion is Dr. Cobb.
                  DR. GEORGE COBB:  Thank you.  First of all, I'd like to say that the things that I'm reading into the record and, when I indicate that there's more to come from other discussants, those have had broad consensus among the discussants, and we're trying to do this for efficiency so that you know who to direct questions to if there are any, hopefully.  And pardon me, I'm largely going to read this, so here we go.
                  The panel commends the Agency staff who are working directly on completing these complex risk assessments.  The 1-BP assessment has many attributes that present improvements from previous TSCA assessments that have been reviewed by this committee.  The panel appreciates these refinements.  The panel has specific comments, many of which were noted in-depth in responses to Questions 1 and 6.  We're not going to verbalize them here today, at least not all of them.
                  Related to the specific questions, the discussants found it quite difficult to follow the exclusion criteria that significantly winnowed the available literature to a small percentage of studies that were utilized for the assessment.  This limits the panel's ability to make an informed assessment of the objectivity used to retain or exclude the data.  And this exclusion criteria produced no or limited data -- for instance, an N of 2 for environmental exposures -- and cascaded into really large uncertainties in the environmental exposures.  No reproductive or developmental data were included in the environmental exposures.  
                  Techniques to include more data by lowering weighting for existing studies that might be of lower quality could be valuable for inclusion in the TSCA risk determinations.
                  We also observed that the process for winnowing the papers to a single or a small key number of studies may be appropriate for cancer endpoints in the human health risk assessment, but it is inappropriate for environmental assessments where a robust data set is needed to get species distribution analyses or species sensitivity analyses.  And that may have led to some of the tension in the discussions about some of these processes.  
                  Although it's understood that the Agency intended to be protective, the decision to use a single high-quality study for the dose-response non-cancer and cancer hazard evaluations may give us a false sense of precision.  Although we understand that is protective, and maybe that uncertainty can be clarified in the document.  
                  Also the diminution in scope between 2016 and 2019 doesn't instill confidence in objectivity.  
                  The 1-BP assessment includes a statement on page 246 that the model surface water concentrations represent a high-end estimate of exposure interval or a pulse exposure and there's no evidence to indicate that this is a high-end exposure.  1-BP's concentration in surface water were from a single reported release from a single site, and that release is three times ten to the minus fourth percent of all of the 1-BP used in the U.S. in a single year.  So it's hard to say that that's a high-end estimate.  And there are ways that you couldn't but do a high-end estimate.  They simply haven't been done.
                  There are gaps in the human health risk exposure assessment, and a quantified sensitivity analysis could help with these gaps.  There are also gaps in the consumer product exposure section that could be used to better describe uncertainties there.
                  The panel concurred that the data supporting the finding of unreasonable risk for consumers and in occupational users was justified.  
                  Also the Agency states that benchmark MOEs were exceeded by several orders of magnitude, and that's not true in all cases, so there's several places where we've identified.  Please go back and redo that again.
                  There are also areas where Dr. Perkins is going to address some of these topics.  I'm getting ready to just read off the topics.  He's going to address these later in more detail: sensitivity of human health conclusions to analytic assumptions, the assumption in age and length of exposure, and the possible underrepresentation of exposure.  
                  Assumption of nontarget exposures leading to uncertainties surrounding aggregate exposures and assumptions that little BP will enter the soil subsurface doesn't have any data to back that up.
                  So, going on, there are instances where the wording the agency uses implies more rigor or more confidence in the data then is justified.  We're not going to go through all of those but several of those are pointed out in our writeup.  Just for an example, the Agency states that, because of uncertainties inherent in deriving risk quotients, the values are protective so that the risk estimate can state with a high degree of confidence that risk quotients less than one are not an ecological risk and can be screened out for further analysis.  That statement is actually incorrect; the uncertainties make it less likely because the uncertainties are so high, and there's more to come on that particular topic in just a second.
                  The consumer uses associated with uncertainties need some refinement in Section 3.  I think we mentioned that just a second before. 
                  The consumer product data are old and the leaking household containers have not been considered, and we realize that you were just informed about the leaking containers in this meeting.  Probabilities of differing human health exposure scenarios were well characterized with the exception of the assumed emission rates for occupational exposures perhaps being inaccurate.
                  The draft evaluation focuses on the major exposed population and defines the limits of populations at greatest risk by estimating high-end exposures.  To these major populations, EPA does mention other populations such as pregnant women or people with other alleles for the CYP2E.  That may make these populations more susceptible to the effects of 1-BP, but the Agency used a blanket uncertainty factor of ten to protect these populations.  And references to support the biological significance to this should be provided or some discussion of uncertainty provided.
                  We talked about modes of action.  There was a section on uncertainties and confidence at the end of each section that qualitatively describe them, but it wasn't clear why some of these were high confidence and some were medium confidence, and that should be clarified.
                  Also the use of models, some of which were probabilistic and some of which were deterministic, is sometimes unclear and confusing, and that could be improved.
                  And there's also an uncertainty about the number of workers that were exposed or at least the precision with which those data are reported.
                  The aspects of adjustment factors in the environmental exposures that require attention -- and it's specifically related to Question 4 and perhaps this should have come up after Question 4 but -- I didn't get these references all pulled together in the information ferreted out until it was too late for that.  So, in Kienzler et al. in 2017 -- and we realize this after your risk assessment was done -- they pretty clearly demonstrated that acute-to-chronic ratios for single species of fish require an acute-to-chronic ratio of a hundred to protect 90 percent of the organisms.  So that acute-to-chronic ratio of ten is probably too low.  So the suggestion is to use that paper, and others that may differ somewhat, to use an acute-to-chronic ratio that approaches a hundred and then use an adjustment factor for inter-species sensitivity perhaps of 5X as was done in the acute study.  And that's the suggestion there.
                  Also, on 187, the report states that the lower bound of toxicity for ECHA studies was used; that's not correct.  The assessment uses the lower of two fish studies, but, if you've got an N of two, just choosing the lower mean is not the lower bound.  The lower bound of that study that was chosen is quite a bit lower in concentration than what is used in the assessment, and this all goes into the uncertainty.  And the panel recommends that the Agency evaluate uncertainties in data sets and ascribe these adjustment factors and uncertainty factors within the numerical assessments of hazard and then discuss the uncertainties in this section around those adjustment factors.
                  Generally, the panel agreed that uncertainty analysis would benefit from inclusion of more discussion on biomarkers and nonspecificity of the assessment endpoints and inclusion of rigorous treatment of uncertainty for the amount of BP that's moving into the environment.  And that concludes all of 1, 2, and 3.
                  DR. KENNETH PORTIER:  Dr. Cobb, this is Ken Portier.  I just want a clarification.  You mentioned diminution in scope 2016 to 2019.  Is that in the environmental risk or is that in general?  I think you needed to expand on that a little bit more.
                  DR. GEORGE COBB:  So, I believe we heard that the scope of the previous document included some air exposures and/or some general population exposures, and that's what that was referring to.  And I do not believe it was referenced to the environmental.
                  DR. KENNETH PORTIER:  Okay.  You said something about human health quantified sensitivity analysis.  Can you expand with that?  I was trying to figure out what that --
                  DR. GEORGE COBB:  Okay.  That was not my comment.
                  DR. KENNETH PORTIER:  I know.
                  DR. GEORGE COBB:  Can you tell me where it was?
                  DR. KENNETH PORTIER:  It was right after diminution in scope.
                  DR. GEORGE COBB:  All right.  Let's get to it.
                  DR. KENNETH PORTIER:  Probably in the next section.
                  DR. GEORGE COBB:  All right. 
                  DR. KENNETH PORTIER:  I was trying to figure out --
                  DR. GEORGE COBB:  Data gaps in human health exposures assessment, that was, I think, Dr. Davies, that was your comment on data gaps in human health exposures without quantified sensitivity analysis makes it difficult to understand the gaps in the effects in the results.
                  DR. KENNETH PORTIER:  Dr. Davies.
                  DR. HOLLY DAVIES:  This is, I think, partly from putting our notes and putting them together.  I think the data gaps I was referring to, I think, were general data gaps.  I don't know if -- I don't think I had human health exposures specifically.  I think Dr. Cobb might have added that.  I think I was pointing that we had data gaps that were important, but we don't know how important all the data gaps are.  So I don't think I was specifically talking about -- because some of them were -- one of them that comes to mind is, as we've talked about, with the lack of mass balance; we don't know what we're missing in some of the other data gaps.
                  DR. KENNETH PORTIER:  This is Ken Portier, so I think it would help if we had a little bit more specificity on that con- -- 
                  DR. HOLLY DAVIES:  On some of those.
                  DR. KENNETH PORTIER:  And then the last thing where I have a question -- assumed emission rates for population exposure.  I didn't quite understand the context of that and that was toward the end of your comments.  I just wanted to make sure I was hearing everything correctly.  There's a lot of stuff in there.  There's something right after the old consumer product data.  You were talking about --
                  DR. HOLLY DAVIES:  Sorry.  I'm trying to catch up to Ken who's trying to catch up to George.  So I mentioned the data gaps, but then the next comment was mine which was appreciating in the consumer exposures that sometimes we have to move forward with some of the data gaps that we can't just say we don't know enough.  We can't do anything.  And that was just one example of making estimates that are good if not perfect.  That was mine.  So what was your other question?  I don't know.
                  DR. GEORGE COBB:  Ken, this is George.  Is it where we're talking about the probabilities of different human exposures being well characterized with the exception that it's assumed emission rates for occupational exposure may be inaccurate?  Is that what you referred to?
                  DR. KENNETH PORTIER:  Yeah, that was it.  So say that again.  I just wanted to make sure because I think that was kind of an important statement.
                  DR. GEORGE COBB:  So it's probabilities of differing human health exposure scenarios were well characterized with the exception that assumed emission rates for occupational exposure may be inaccurate and, therefore, have produced release rates that underestimate actual releases.  I'll apologies; that was not my comment.  I believe that was either Dr. Cory-Slechta's or Dr. Willett's, but I'm not sure.
                  DR. KENNETH PORTIER:  Dr. Cory-Slechta, I think you're next on the list of discussants.  Do you want to add?
                  DR. DEBORAH CORY-SLECHTA:  It's hard to hear on the phone because somebody's typing, and so I couldn't hear what Dr. Cobb just said.
                  DR. KENNETH PORTIER:  So this is Dr. Portier.  I asked him about the statement on the probability of defined human health exposures was well with the exception that the assumed emission rates might be underestimated for population exposure, and I think he said that might have come from your notes.  And I wondered if you could expand on that.
                  DR. DEBORAH CORY-SLECHTA:  I'm not sure if I did, but I'm wondering if it's the comment that I had about the fact that these ONUs may not really be distinct from workers particularly in a lot of the smaller dry-cleaning operations and that I think that distinction in a way is potentially artificial and overstated.
                  DR. KENNETH PORTIER:  Okay.  Do you have any additional comments to add to the Section 7.1, 2, 3 questions?
                  DR. DEBORAH CORY-SLECHTA:  No, I don't.
                  DR. KENNETH PORTIER:  Thank you.  Dr. Davies.
                  DR. HOLLY DAVIES:  No.  I think Dr. Cobb captured the comments for these sections.  
                  DR. KENNETH PORTIER:  Dr. Perkins, you indicated you had supplemental discussion.
                  DR. EDWARD PERKINS:  Most of my stuff was recapping things that were discussed in Questions 1 through 6 such as accounting for actual respirator use.  There was discussions on how people may not be actually using the respirators that they think they might be using, and the quality of respirators may be a lot lower than we're thinking.
                  The idea of combining dermal exposure with inhalation or that some of the procedures during occupation, people might have two different types of occupational exposures simultaneously, so that created a lot of uncertainty and then discussion on application of age adjustment factors and how that there may be daycares or children in the area.
                  Then pretty much a number of other things that were talked about created a lot of uncertainty to greater or lesser degrees.  Dr. Cobb talked about the general topic.  Most of the specifics have already been discussed in other questions.  I'll expand upon that in the written element.
                  DR. KENNETH PORTIER:  Thank you.  Dr. Willett.
                  DR. CATHERINE WILLETT:  I think Dr. Cobb captured everything under those three that I had.
                  DR. KENNETH PORTIER:  So we'll come back to Dr. Cobb.  I should open it up to any additional comments, but I'm going to ask Dr. Cobb at the end to kind of point out what recommendations might have come out of this discussion because I'm not sure I heard recommendations specifically.  I heard a lot of good comments but recommendations.  Dr. Blando.
                  DR. JAMES BLANDO:  This is Jim Blando.  Just one other uncertainty, I know Dr. Cobb mentioned something about the estimation of the number of workers exposed.  One other additional uncertainty that might be worth noting in the report, although I don't know that there's any recommendation of what you can do about it, is just the difficulty in predicting future worker populations because, as any economists would probably tell you, that's almost impossible to predict but that might be worthwhile just making a mental note that although you predict the number of workers or you might say there's only a certain number of shops or places that use this, it's really hard to predict in the future what that future demand might be.  Again, I don't know what you can do about that other than just make a mental note that that's something to think about when you're trying to predict impacts.
                  DR. KENNETH PORTIER:  I know in the past the panel's always been worried about future consumer use and the impact of this assessment on driving or depressing future consumer use.  I was talking to someone this morning who said, well, the industry says the future for 1-bromopropane is bright for consumer products.  And I'm like, okay.  How do we incorporate that into this?  We don't.  
                  UNIDENTIFIED MALE:  Maybe we should.
                  DR. KENNETH PORTIER:  Yeah, but where do we get that information and how do we quantify that to a point that really helps EPA?  I think we all agree that's an uncertainty, but it's not necessarily a quantifiable one at this point.  But it's always a concern.  Any additional comments?  Dr. Cobb, can you identify any specific recommendations that came out of that?
                  DR. GEORGE COBB:  This is part of the challenge we had because this question -- and you even alluded to it this morning -- how is this question different than some of the underlying questions for the previous six?  So a lot of the suggestions are probably called out in the first six responses, the responses for the first six questions.  The things that I can kind of grab onto to right now is, make it clearer how the data assemblage, data exclusion, data retention was done.  I know that's kind of a mammoth task.  That gives people a better sense of the objectivity and trying to include data that is useful even though it might not pass every single one of the criteria and maybe use a weighting scheme in there.  And I'm not even sure we got to that.  That might be in 7.4 even.  
                  And then making sure that uncertainty factors or adjustment factors are appropriate for the data set.  For instance, if you've got an N of two for the universal data that's available, there's a lot of uncertainty about that data and having the right adjustment factors or uncertainty factors in those sections in the numerical hazard quotients can help.  Most of the things I brought up are things that could be improved in the report, the different ways that the uncertainties aren't as well-specified, and if you would like, we can try to be specific on where in the document that those improvements could be made.
                  DR. KENNETH PORTIER:  Any additional comments?  I was looking at Question 7.1 and thinking about sensitivity and trying to figure out, is there anything else they could do especially on a quantification sensitivity analysis?  But I can't think of anything especially on the human health hazard assessment.  Then using the model averaging approach kind of was part of a sensitivity analysis and so they've done that.  Other than actually quantifying the assumptions and trying to do some more, I just can't think of anything to add to that.  Dr. Kissel.
                  DR. JOHN KISSEL:  I've been trying to figure out where exactly this fits, and maybe it was supposed on 6.5.  There seems to me there's a lot of overlap between these sections that we're discussing.  But I asked early on in the -- it was either under Question 2 or 3 about how some of the dermal modeling was done.  And I finally had a chance last night to go find P_DER2b and look at it and put it in a spreadsheet and do some calculations, and it is structurally incorrect.  So all of the results that are generated by P_DER2b are incorrect and do not reflect the assumption, so we're not talking about just I'm arguing about what assumptions you're making about the exposure parameters.  
                  As you're using it, you're using the density which is the concentration of the pure compound as the driving force with a permeability coefficient that applies to application in an aqueous vehicle, and so there's a partitioning problem there that is not included.  The consequence of that is that fluxes of up to 13 milligrams per square centimeter per hour are predicted which are 13 times higher than the max fluxes experimentally determined by Frasch and another 30-fold higher than the max flux that you would get if you took that aqueous vehicle permeability coefficient and multiplied by the solubility.  
                  So it's likely that the dermal estimates calculated using P_DER2b are overestimates given the stated assumptions.  It doesn't mean that they're actually overestimates because the stated assumptions can be wrong, but the calculation on the basis of the stated assumptions is leading to too high of a number.
                  Are we in the right place for this?
                  DR. KENNETH PORTIER:  Yes.  We can always move it back to 6.5 when we write the minutes, but it's good to capture this at this point.
                  DR. JOHN KISSEL:  Okay.  There's also some confusing stuff.  In the consumer exposure model user's guide on page 67, there's language which explicitly states that P_DER2b doesn't use load, so it's just calculating from some mystery material that's somehow out there and actual loading doesn't come into play.  But, in the tables which describe the assumptions for the various consumer exposure scenarios, there actually are numbers about amount of load.  And I don't know whether those are just stuck in the table and not used or whether there's some disconnect there, but I couldn't figure that out.  So that's not clear to me.  
                  This is a minor point, but in the description of P_DER2b -- I really don't like that acronym but -- P_DER2b says to select a surface area in a bodyweight ratio which is one of the exposure factors that's in there, go to Table B2 in the appendixes to the user's guide of the consumer exposure model.
                  DR. HOLLY DAVIES:  Can I just ask one quick question?  Is the dermal model that was used just for consumer use or was it the dermal model used for occupational?
                  DR. JOHN KISSEL:  P_DER2b is the permeability model that's used for the consumer case.
                  DR. HOLLY DAVIES:  Okay. 
                  DR. JOHN KISSEL:  The occupational stuff used to fix fractional absorption.  So the appendix says, go look in Table B3, and it's actually in Table B2, so I had to hunt around for that.  
                  Now, if I could say something a little more general about the -- so many people, mostly toxicologists, said, we really don't like flipping back and forth between permeability models and fixed fraction absorption models.  And, although I don't like fixed fraction absorption models, I'm not so sensitive about that if you do both of them at will.  If the predicted outcome is a reasonable result, then I don't care how you got there quite so much.  The mere fact that they're different doesn't bug me quite that much.  
                  But, generally, fixed fraction absorption models are not well-grounded in physics because they only apply -- whatever fixed fraction that you observed experimentally only applies to those experimental conditions.  And, once you extrapolate to some other condition, it's not a physical constant.  It's not like molecular weight.  And so you can't just go off to anywhere else in the world and say, oh, we were always going to get 0.0016 fraction of the material absorbed, so that basically doesn't make any sense.  This is a pretty volatile compound, and so the competition between volatility and absorption is important.  I do think that the way the experiments were done were under higher-velocity conditions -- wind speed -- and, therefore, more rapid evaporation when you would see in a normal case.  
                  And so, therefore, I think there's a good chance that the dermal exposures are underestimated, and, actually, there's a second reason for that as well.  Whereas, the consumer exposures overstate the actual -- the intended exposure.  Now, in that world, there's how much stuff do people actually come into contact with?  And I don't think we have a very good handle on that, so there's always a possibility that you're overestimating from an underestimating starting point.  So you might be getting to not an unreasonable result.  I can't really say.  Without biomarker data to back up what's going on, then you don't really know whether you're making a good estimate or not.  
                  The other piece -- and there was some confusion about this because I brought up that, if you put on respiratory protection then you can't necessarily neglect direct dermal uptake from vapor.  And that's not in here anywhere. I don't know if Dr. Anderson's is on the phone, but in his comments, I've detected some confusion about how the dermal absorption stuff is being done.  And the dermal absorption that's presumed is exposure to liquid, and nowhere in this document is there any exposure to vapor.  
                  And I did a back-of-the-envelope calculation using the slope factor of the 0.006 milligram per kilogram per day extrapolated inhalation to dermal slope factor for carcinogenicity.  And my back-of-the-envelope calculation is that via vapor alone at 20 parts per million ambient, you've got about a 10 to the minus 3 lifetime excess cancer risk; and that's not in here anywhere.  
                  Potentially, the dermal vapor uptake is not negligible for this compound.  And that's based upon -- if I had done that calculation two weeks ago based upon the max flux estimated from water solubility and epoxy-type permeability coefficient, I wouldn't conclude that.  But I've adjusted that calculation for the observed experimental fluxes shown by Frasch, and those numbers are much bigger.  And when you plug in those potential max fluxes, then the reduced flux at something less than saturation in the air and it's a lot less than saturation.  So you're at 20 parts per million, and saturation's about 200 thousand parts per million.  But you're involving two square meters of a human being when you're doing vapor uptake.  So it turns out that I think direct dermal vapor uptake is not negligible, and that's not in here anywhere.
                  DR. HENRY ANDERSON:  I'd say thank you very much.  You're right; I was confused.  I didn't realize it was liquid that the estimate was done on, so thank you very much.
                  DR. JOHN KISSEL:  In the last part here, I was a little confused by the conversation yesterday about we can't do aggregate risk because we don't have a PBPK model.  I understand that, if you're trying to match biomarker data to inputs and you have to work backward, then you need a PBPK model to do that.  But my experience with EPA risk assessments is that pathways get added together, and aggregate doses are then multiplied by a slope factor.  And that's been happening in the Superfund world for decades, and so I don't understand why it can't be done here.  It may be a little imprecise if you're taking gross systemic absorption from dermal and one and adding it to ingestion because of the liver first-pass phenomenon.  But, in this case, ingestion isn't really in the mix, and I don't understand why we can't add systemic dose from dermal and systemic dose from inhalation and estimate it a total risk.  So that's more of a comment than a question.  
                  I guess, something else that I would add is that what I've discovered is that not just what somebody has done for this particular risk assessment but the actual -- the text in the published consumer exposure model user's guide is incorrect and misleading, and that's a problem going forward.  And, if P_DER2b is wrong and there's a possibility that everything's in there are wrong also, other dermal models are wrong also.  
                  It's a fairly common mistake to think that once you've calculated a permeability coefficient -- and some of the people that produced them are not very helpful and the work is taken from Wiltenburg's (phonetic) website and it's not at all clear.  He doesn't point out that he's produced an estimate for an aqueous vehicle; it's not clear there.  But you can't take a permeability coefficient for an aqueous vehicle and apply it to a mixture of any other kind than an aqueous vehicle, and I see this in the literature all the time.  It's not like EPA has done something that is particularly odd here.  
                  But working in the dermal risk world, this is a really common mistake.  The problem here is that it's embedded in an EPA document, and that's actually been embedded in prior EPA documents that I have criticized.  But at some point, you need to hire a team of dermal people that actually know what they're doing and work on your protocols because I'm going to be saying this at every one of these meetings is that the way you're doing dermal is not correct.
                  DR. KENNETH PORTIER:  Dr. Kissel, can you add in a reference for that aggregate risk.  You mentioned the Superfund document.
                  DR. JOHN KISSEL:  Well, it's probably in regs which is the standard document.  I'd have to go look, but I'm sure I've seen that many times.  In many EPA documents that multipathway -- if you have exposure via ingestion of food, ingestion of dirt, ingestion of water, they'll get added together ultimately.  And you don't have to isolate all that so that we can't do aggregate because it came in a different vehicle or it came from a different route.  That's not historically how EPA has operated.
                  DR. KENNETH PORTIER:  I think Dr. Cobb and then Dr. Davies.  Dr. Cobb.
                  DR. GEORGE COBB:  Thank you.  I really have a question.  So your assessment was the error in the P-DER2b model overestimated the exposure by dermal uptake.  Is that correct?
                  DR. JOHN KISSEL:  P_DER2b is exposure to product, and I think it's overestimating because the driving force is the density and not the aqueous solubility.  So, if you have pure compound -- thermodynamically, at least theoretically, with damage to barrier function, things can change.  But, theoretically, exposure to pure solvent is the same thing as exposure to saturated liquid.  
                  So, if you had pure solvent on your skin, then you could take an aqueous vehicle permeability coefficient multiplied by the aqueous solubility and that would give you an estimate of what the max flux that you would expect.  Now, in this case, that estimate, that traditional estimate of max flux, is much lower than what was actually observed in the Frasch experiments.  I mentioned this last time; you're making estimates based upon physical/chemical parameters and those things can be off by one and half logs which is what we have here, and so that number can be wrong.  But, with respect to intention, the consumer thing is too high.  Now it's possible that there's other errors build in there that that overestimate actually isn't all that awful, but that's a different issue.  And that's not error at all.  I'm not asserting that the air is overestimated because the air is not even considered.  So it's by definition underestimated.
                  DR. KENNETH PORTIER:  Dr. Davies.
                  DR. HOLLY DAVIES:  So having John dig into that model just brings up another issue that I want to mention some point explicitly on some of these models that were used here that we didn't dig into that seemed reasonable for this use that I didn't want to make it seem like we were kind of peer reviewing them for all uses or blessing them in some way more than we are because we haven't really dug into them.  And I just wanted to make sure when we write up our report or the way it comes that it's qualified in how they were used or that they were being used for this.  And maybe that goes in Question 7 with the characterization of uncertainties and assumptions.  And it's also kind of how they get peer reviewed and what constitutes peer review because we haven't -- because I think John's the only one who really tried to run any of these and dig into it and only the dermal one.
                  DR. KENNETH PORTIER:  I would say things like the BMD package.  It's pretty standard, and they put that right.
                  DR. HOLLY DAVIES:  I saw in the EPI Suite.  
                  DR. KENNETH PORTIER:  But in some of the -- 
                  DR. HOLLY DAVIES:  Right.  Some of them are well established.
                  DR. KENNETH PORTIER:  Yeah.  The one like the unloading model that I questioned earlier on.  I was like, I wanted to kind of dig into it, but I had a hard time getting down to the actual details of the model.  And I think we recorded that a couple of days ago.  Any additional comments?  Oh, Dr. Blando.
                  DR. JAMES BLANDO:  When I read this question, I was thinking more of sort of global statement that I felt like we were being asked to make.  So I'm going to make my global statement that, after reviewing this and spending a few days talking about a lot of the subtle details that are really important and really and crucial for EPA to consider to improve the assessment, I just wanted to make the statement that for me, personally, as a member on this committee as a scientist working in this field that I did feel that this risk assessment was objective and is a quality product, and that as a public health professional, I would feel comfortable utilizing this document to make decisions.  And I just felt -- that's what I thought we were being asked for to confirm that, and, in my personal opinion, I just wanted to confirm that.  That's not to say it can't be improved.  It's not to say that all the very important suggestions that have been here in the last few days that EPA really does need to consider, but I just wanted to affirm that I think this an objective and quality product.
                  DR. KENNETH PORTIER:  I think that's a good lead into Question 7.4 because I have a feeling that's what 7.4 is asking us.  So why don't we go ahead and read 7.4 in.
                  MS. TAMUE GIBSON:  Dr. Barone.
                  DR. KENNETH PORTIER:  Yeah, okay, before we read it --
                  DR. STAN BARONE, JR.:  I'd like to get some clarifications first.  I just want to be sure we understood things correctly, and I want to make sure we get the feedback in the report so that to really make these recommendations or what I'm hearing as nascent recommendations actionable.  
                  I do appreciate Dr. Blando's comments.  Sometimes I think the committee has really great ideas and has criticism, but sometimes these criticisms can be interpreted by others as showstoppers.  And so we really need to know when things are showstoppers or prioritize what these efforts could actually entail because some decision-makers would use some critiques as, oh, you can't make any decisions.  Oh, you can't do anything with this.  So that's why I'm drilling down a little bit, so I want to preface my comments or my questions.  
                  So I heard the comments about occupational emissions as being incorrect, and it wasn't really clear.  When I hear occupational emissions or occupational emission rates, that's a big deal for our modeling efforts.  So I need a little bit more information about what you think is wrong with our assumptions or the data that we're using for occupational emissions.
                  DR. KENNETH PORTIER:  This is Ken Portier.  I didn't quite hear that and that's why I asked the clarifying question because I think what they were saying was, and if I get my notes right, that general population exposures -- and maybe you're talking about indoor air consumer use exposures -- might not have been properly estimated, and I'll ask Dr. Cobb to clarify.
                  DR. GEORGE COBB:  Again, that was asked earlier and I appreciate the question and I was kind of the scribe here.  That was from one of the other discussants, and I'm not sure which discussant that came from.
                  DR. KENNETH PORTIER:  Can you read that section again?
                  DR. GEORGE COBB:  What I'm reading here is probabilities -- and, Stan, correct me if I'm wrong with this what you're getting to -- differing exposure scenarios were well characterized with the exception that the assumed emission rates for occupational exposures may be inaccurate.  Is that what -- and lead to an underestimate of actual risk.  I'm struggling to find out who sent me that comment.
                  UNIDENTIFIED MALE:  That may have been Cory-Slechta.  Yes, Cory-Slechta. 
                  DR. HOLLY DAVIES:  Yeah, I checked my notes, and I don't see it in my notes.  It doesn't sound like me, and I didn't accidentally send it.
                  DR. STAN BARONE, JR.:  So, if you can figure out what the basis for that is, we would like to be able to correct that because that, at least from our vernacular, that's a big deal.  
                  DR. GEORGE COBB:  Understood and actually you can't see it, but I have that italicized, or actually bolded, in something that I need to go back and clarify.
                  DR. STAN BARONE, JR.:  Okay.  And, again, emission rates for general population, general population wasn't assessed here.  We scoped that out for regulatory Nexis reasons.  There was also mention of ADAFs again in Dr. Perkins's comments.  And, again, we are not looking at daycare and adjacent buildings.  ADAFs only apply to a mutagenic mode of action.  
                  So, again, depending upon the recommendations of the committee, and then what we decide to do, we sort of need to have that ground tree thing of, you know, when do apply ADAFs and when don't you apply.  You don't apply ADAFs for every carcinogen.  It's only with a mutagenic mode of action; and it's only for children's exposure from 0 to 16.  It's not for gestational exposures.  We do not have guidance for gestational exposures, unlike Cal. EPA.   
                  So I just wanted to put that out there.  The comments on P_DER2b, again, the more that can be detailed out in the recommendations, the more actionable that will be for us.  And we do have ongoing work on dermal modeling.  And, again, this has been brought up several times, so we're working fast to try to resolve some of these issues.  But the more detailed comments -- and we've yet to get a report from the committee.  As soon as we get a report from the committee, we hope that we can come back and talk to you next time about the specifics in the report and have revised estimates based upon our ongoing work.  So this is a sort of moving target we recognize, but we really need the committee's input on this very important issue.
                  DR. KENNETH PORTIER:  Oh, Dr. Blando.
                  DR. JAMES BLANDO:  This is Jim Blando.  What I thought I heard, just to clarify regarding occupational exposures, was the comment that Dr. Cory-Slechta made on the phone which was her concern about the occupational emissions was that for small shops, small operations, the distinction between an ONU and a worker is not that clear, and that perhaps it's too artificial, and that maybe an ONU and a worker that distinction between their exposure concentrations because it's not that distinct and that maybe their ONUs and worker exposure should be.  So that's what I thought I heard, but she's on the phone, so she can clarify if that's what she meant.  That's what I heard.  That's what I thought I heard.
                  DR. DEBORAH CORY-SLECHTA:  Yes, that's what you heard.
                  DR. STAN BARONE, JR.:  That's much clearer to us and more conscribed.  Again, in the assessment, we're making the distinction between ONUs and descriptively and the actual user but, in effect, the exposures are virtually on top of one another because neither -- we're not considering PPE at all in that particular condition of use.  And again, because of the activities, we are recognizing that spot cleaning and the dry cleaning are aggregated in some cases.  So that is part of the calculations so further support or --
                  UNIDENTIFIED MALE:  (Inaudible). 
                  DR. STAN BARONE, JR.:  Pardon?
                  DR. KENNETH PORTIER:  Who is that?
                  UNIDENTIFIED MALE:  (Inaudible). 
                  DR. KENNETH PORTIER:  Is that Dr. Anderson?
                  UNIDENTIFIED MALE:  (inaudible). 
                  DR. HENRY ANDERSON:  No, that wasn't me.
                  DR. KENNETH PORTIER:  I think we're getting some bleed through here, but anyway, okay, good.  I think we need to add that kind of clarification, and Dr. Cobb will work Dr. Cory-Slechta to get that wording right and clear.  
                  From what I've heard this morning, I think the one key thing you're going to have to work on is the dermal stuff, and I know Dr. Kissel always provides us good background on this stuff, so that'll be in the report.  Ms. Rudel.
                  MS. RUTHANN RUDEL:  Thanks.  I wanted to echo what the previous reviewer said about feeling like this is a good, carefully done document that has a lot of really difficult, complicated, and helpful sort of focused analyses to answer some very specific questions.  
                  I think in Question 8, I might talk about that issue of the scope and how narrow you are and trying to address a recommendation for this one in particular.  I think that the Question 7.1 is -- I'm interested that you're specifically asking about objectivity of the underlying data.  I think I would just echo something that I said for PV29 which is that the one area where I really felt like I didn't understand the data gaps, or they weren't completely described is on how much is produced and where it goes and where it ends up.  So that sort of big picture about this, without knowing that, is really hard to evaluate the reasonableness of all the rest of the assumptions.  But for each of these specific scenarios that you've picked out, I feel comfortable with the way you've described things.  
                  I have a couple of comments.  I think, about how easy or hard it is for people to understand what's happening when you talk about MOEs and excursions above or below them and above and below benchmarks and that's more, I think, for Question 8.
                  DR. KENNETH PORTIER:  Mr. Kaufman.
                  MR. ALAN KAUFMAN:  Hi.  I was going to cover this when we got to 8, but, since we're already on the topic, I'll go ahead and mention it now.  I would agree with Dr. Blando and some of the other speakers that I think it's an objective report.  I don't see anything to quibble with in a major way, I guess I would say.  I know there may be some adjustments that need to be made.  
                  I think where there may be some improvements that could be made is in clarifying some of EPA's assumptions and thinking and I will say -- I think, somebody mentioned it yesterday -- the presentation, particularly the PowerPoint presentation, was very, very helpful in sort of understanding a lot of that.  And what I would recommend is maybe incorporating some of that, particularly some of the figures into the executive summary because I read through the risk evaluation, but it became must clearer as we went through the presentation.  Someone just picking up the risk evaluation is not going to have the benefit of that.  And so I think incorporating some of that into the assessment itself, I think, would be very helpful.
                  DR. KENNETH PORTIER:  Thank you.  Any additional comments before we go onto 7.4?  So let's move to 7.4.
                  DR. KATHERINE ANITOLE:  Question 7.4 Please comment on whether the information presented supports the findings outlined in the draft risk characterization section.  If not, please suggest alternative approaches or information that could be used to develop a risk finding in the context of the requirements of EPA's Final Rule, Procedures for Chemical Risk Evaluation under the Amended Toxic Substances Control Act (82 FR 33726).
                  DR. KENNETH PORTIER:  Dr. Cobb.
                  DR. GEORGE COBB:  Thank you.  So, again, I'm reading this response.  In essence, the environmental data that are presented from environmental aspects of the assessment are inadequate for a robust risk characterization and do not support the conclusion of no unreasonable risk to the environment.  The risk characterization for occupational exposures to 1-BP is supported by data in the assessment.  However, the risk characterization for consumers excludes several indoor air sources of 1-BP such as wallboards and other things.  Thus, the assessment underestimates risk to the consumer.  
                  We also feel that the lack of the general population exposures excludes a vast extent of the U.S. population and should be considered, and we realize that's been deemed outside the scope.
                  Evidence of widespread exposure to the chemical that maybe 1-BP based on biomonitoring data is noted, and, although this biomonitoring data is not definitive proof of widespread 1-BP exposure, it does add to the uncertainty estimates for workers both directly and indirectly exposed and consumers.  This does raise concerns about broader exposures that may need to be assessed or recorded in the uncertainty analysis.
                  Two assumptions in the human health exposure require some further consideration: one is the assumption that seems to overstate things relative to dry cleaning operations where it's more likely that all employees are actually occupationally exposed which is somewhat different than what the Agency has done.  This could certainly affect the number of people exposed or the extent to which they're exposed.  And this is actually Dr. Cory-Slechta's remarks, the part I just read.  
                  And then finally, the discussants identified some challenges in evaluating Table 5.1.  I think Dr. Davies, perhaps Dr. Willett, are going to speak to that.  And there are uncertainties related to the exposures via inhalation that Dr. Willett is also going to address in comments that I've captured here in this writeup.
                  DR. KENNETH PORTIER:  Before we move on, on your first statement about the environmental risk, are you going to kind of refer to what needs to be done to correct that.  I mean, we've had that discussion, but I think we need to, at least in this section, refer back to what the remedies might be for that.  And you mentioned something about including more data of lower quality, but do you have other recommendations that --
                  DR. GEORGE COBB:  So the first recommendation is with the paucity of data to go back to that reference that I provided just a few minutes ago about the assessment factor or it was actually the acute-to-chronic ratio and the species distribution that has been done to demonstrate what you have to do to do that acute-to-chronic ratio for a single species and then figuring out what to do about the inter-species adjustment factor is important and that will specifically will be called out.  There are also probably other ways to get to a worst-case scenario for where it leaks into the water, and there are varying things that you could do, but we'll make some recommendations there.  
                  But those are the two big pieces.  Actually, you've got one data point for exposure that's marginal at best, and you've got two data points for effect and that simply has so much uncertainty that to just put it in a normal package just is definitely not protected.
                  DR. KENNETH PORTIER:  Okay.  Let's move on.  Dr. Cory-Slechta, do you want to add to these comments?
                  DR. DEBORAH CORY-SLECHTA:  No, I don't have anything to add.
                  DR. KENNETH PORTIER:  Dr. Davies.
                  DR. HOLLY DAVIES:  Yes, I was going to talk about Table 5.1 which is where it's all kind of rolled up, and I actually do have a table that I made in trying to understand this.  I keep going back and forth as to whether this is kind of -- I think of Question 7 as being more substantial information and Question 8 as being kind of presentation overall and go back and forth.  
                  So I spent several hours trying to figure out Table 5.1.  At the end it all worked, and it seems like it was all substantiated and the unreasonable risk all seemed reasonable, so that was good at the end, but it all came together.  But it did take me a while to get there in trying to follow everything.  And I made this table, and Dr. Willett thought it was helpful, so I asked Don to show it.  I thought the rest of the committee might find it also helpful because I wanted to through how EPA had come to the either unreasonable risk determination or not unreasonable risk determination because that seems to be kind of the crux after all of this.  And, of course, the actual determination is appropriately an Agency determination, so the risk is what we're looking at with all the numbers.  So you don't need to look at my notes on the side.  You can make it bigger, Don.  They don't need to see all of it.  They don't need to see the M column.
                  So we looked at risk.  So I was kind of color-coding it, and I thought it would be kind of easy to pull out extra columns off to the side of when did they see cancer risk and when did they see non-cancer risk and who did they look at and obviously there might be mistakes in this because I did this quickly.  And it was interesting to see.  And then the green on the very left, it was the Agency determination of unreasonable risk or no unreasonable risk with green being no unreasonable risk and red being risk.  
                  And so kind of going through it, you can just see the different categories and how they did it.  And what goes to Question 8 that we can talk about more in the next one is how it is hard to kind of follow and link through all the tables.  But I do think the Agency did a good job with going through, and it's confusing and complicated.  But going back, sometimes you have monitored end modeling with the APE, without APE; with controls, without controls which is why it ended up not being a simple table.  They know this.  I'm sure they've tried this where you end up looking at five tables that go into this or, of course, there's multiple endpoints for the non-cancer.  And, I think, they picked the lowest MOE for the highest one is what it looked like.  So, after following them, I was like, okay, they didn't say what they did, but I think that's what they did.  So it all ended up making sense which is good and what they looked at and what they didn't look at.  
                  I sent you the wrong version because I found that was a mistake of mine that cancer that's red, that shouldn't be red because I thought that was red at first but it's not.  I was looking at the wrong table because it's confusing.  So they were right, and I was confused, which is great that they were right.  
                  If you go down, you'll see some didn't look at ONUs and some of them were mixed where they made the determination.  If you keep going down, you'll see -- keep going, keep going -- and then they didn't look at PPEs, of course, as we saw, and Dr. Willett will talk more about PPEs.  They didn't look at those for some of them, and then you -- keep going  -- find consumers and bystanders for all the consumer products and then disposal at the end, which I assume is in that order.  For some reason, that makes sense to them, but it links back to Table 1.4, which is nice.
                  I just wanted to bring it up here, after going through it all, it did all link up and kind of hang together, and it seemed to be supported by the data in the previous chapters -- and if that table helps anybody else.
                  DR. KENNETH PORTIER:  Any questions of Dr. Davies on her table?  How is this table kind of different from the table in the last chapter?
                  DR. HOLLY DAVIES:  They have everything in words where it says go back to this table, go back to that table.  They just have many more words.
                  DR. KENNETH PORTIER:  Or is it just a quicker scan?
                  DR. HOLLY DAVIES:  I was thinking of it as more of the communication tool.  How could you do it?  The table, which is an end, that table has to be there.  It's the words, right, under this part of TSCA, we found unreasonable risk to the -- it has to be there.  And it links to all the -- I shortened a lot of these things which they can't do, but it has to be there.  But, in addition to that, I was like, was there an easier way of saying, well, we found this because what we saw was -- and it's just so many tortured words like, it was above the or below the MOE, but then with PPE it was -- it was just so tortured.  It's complicated and there's just too -- and I just was finding it's like, well, what's an easier way to present it, and like, the no with PPE or this one was no, which seemed easier.  And we looked at all of these and they all are agreeing in this way.  Okay, great, we found no unreasonable risk, or we found risk in all of these endpoints for all of these people, so we found risk, so kind of showing how that all kind of rolled up was more clear to me.
                  DR. KENNETH PORTIER:  So for those ad hocs on the committee that haven't sat through the previous three meetings, the permanent panel members continue to struggle to get the high-end view at the end of this whole process to where -- and what Dr. Davies is saying is that last chapter and it's got that very formal table, few of the panelists see that as a good communication device.  It's a requirement, I think, in EPA's regulatory environment, TSCA environment, but for the rest of us, as a communication tool, it doesn't cut it for us.  So this is the more general view, and this is kind of what I had in my head at the end of the reading the whole process, so I kind of could not see that.
                  DR. HOLLY DAVIES:  But I didn't have it in my head because I couldn't get into my head because it was so -- I needed to write it out and could you add other columns, and the words in the executive summary were really hard.  I started there, and it was like, these are okay and some of it is the MOEs versus the cancer.  Which ones you want to be below and which ones you want to be above?  And it's just there's so many double negatives.  There's just so many double negatives.  
                  And it's the audience, so this where we go into Question 8.  It's the audience and some things are written for the audience with the double negatives and the legalese, and some things are written for a different audience where we can get rid all of that.  
                  But, as I said, we've gone over this, and, I think, EPA tried to make clear and they really did make it clear who they assessed for what and they repeated it several times, and I can see that in this.  And sometimes they didn't have data or sometimes they said, we don't think we should assess them who they looked at cancer.  So sometimes that's not clear to us, and Dr. Barone has said that several times.  It's in the guidance that we're not always as intimately familiar with.  It's always good to refer to that which might not be something we always agree with, but it's definitely something EPA will follow.  And there's a reason that EPA will do it, and so it's good to see all of that what was assessed when.  
                  DR. KENNETH PORTIER:  So, Dr. Davies, actually A, B, and C are really the conceptual model that EPA puts back at the beginning.  And the only thing kind of missing is this are those conditions that are considered out of scope that we continue to talk about even though they're out of scope, things like internal air for consumer products and external air for population exposures.  So that would be the only thing I might add to this is a couple of situations where we would just have a line that says, determined to be out of scope, so that a user who's looking or a reader who's looking for that condition of use needs to be identified right up front.  You're not going to find that in this document because it's scoped out.  That's the only thing I would --
                  DR. HOLLY DAVIES:  But I'd stop looking.
                  DR. KENNETH PORTIER:  Huh?
                  DR. HOLLY DAVIES:  Some people know to stop looking.
                  DR. KENNETH PORTIER:  Ms. Rudel.
                  MS. RUTHANN RUDEL:  Yeah, I think that's an especially important suggestion in order to -- because of the whole state preemption issue.  So if you are explicit about what is not in your scope, then it might make it possible for states to take action on things that you didn't cover if they end up being an issue.
                  DR. KENNETH PORTIER:  Dr. Doucette, do you want to do this or you're moving to another topic?
                  DR. WILLIAM DOUCETTE:  No, I'm still back on -- 
                  MS. TAMUE GIBSON:  Use the microphone.
                  DR. WILLIAM DOUCETTE:  Ooh, sorry.  Bill Doucette.  I'm still back on George's comments in terms of the unreasonable environmental risk.  So it comes around again to that, at least in my mind, we're continually fighting against this; what is the minimum data set that would allow us to be confident in making that decision?  I view that and, since I was on 4.1 and 4.2 and a lot of this is repetitive or just regurgitating some of that in a little bit different way, I felt there's always not going to be enough data.  What is that minimum data set that allows me to make that decision with some level of confidence that I can sleep at the night or whatever?  I'm not particularly worried about an environmental impact unless it's a spill situation of 1-BP based on what I've seen.  There is probably a scenario where I took the total year's or a single plan or a pubic -- a production facility that lost it all at once and it certainly would be an impact to the environment within that area, but I'm not sure it's a -- to make a statement that I can't say that it's unreasonable risk to the overall environment.  I'm struggling with what data I would need to be able to do that, and I'm not sure we can answer that.  Or how would you answer that?  I guess that's my question.
                  DR. KENNETH PORTIER:  Dr. Cobb.
                  DR. GEORGE COBB:  So, if you had release data from the industries, then that would give you the robust data set that you would need to assess what the releases are, where they are, what the strength loads are, what that looks like, so that would be one for aqueous.  I know that we're not talking about air necessarily here.  You tend to get the same information from air.  
                  So that could solve the aquatic component largely, but then that's the exposure side.  Then you've got the effect side.  And to evaluate the universal species out there, I guess EPA has to FIFRA, a process that has a certain amount of fish and a certain number of invertebrates.  I'm not suggesting we necessarily need to go there, but if you had several species of fish that represented the universal species where you've identified the proper -- the releases might be highest -- maybe two or three in that particular area, then you could say, I'm pretty confident that we're not going to be adversely affecting environmental resources.  
                  I shy away from even mentioning this, but you'd run into endangered species.  I think that's a whole other can of worms, which I don't think we want to get into here, but, if you layer all that on top, you do kind of have a mess.  But, to protect the environment, I think you need more than a couple of fish studies to do that.  And it might be three or four fish studies and a couple of aquatic insect studies, and that would probably do it for fish, but there are people on this panel that do that sort of thing much more than I do.
                  DR. WILLIAM DOUCETTE:  Just a follow-up and I think I had this discussion with Dan because I was dealing with the -- basically, we had one piece of solid information on the fish, and could I extrapolate from that one key piece of information all the way across multiple species and all sorts of things?  And that's a constant struggle.  And I'd like to hear somebody else, in terms of a toxicologist, in terms of that extrapolation because that's the key piece of data for the environment.
                  DR. KENNETH PORTIER:  Dr. Schlenk.
                  DR. DANIEL SCHLENK:  I think I mentioned this the other day.  The first version of this had absolutely nothing.  So, in my mind, this actually is an improvement, and I was going to say that when we get to 8 actually.  I actually appreciate the fact that there's an effort that's being placed to sort of address that, which wasn't in the 2016 document.  So it's kind of is your glass half empty or half full kind of thing.  The Agency's put in a really tough place because there's no data.  It's not like FIFRA.  FIFRA mandates you have multiple species that are done, and even that, there's uncertainty associated with that.  You still have uncertainty with that as well.  
                  So, in my mind, given the data that the Agency has, I think they're doing the best they can with this, and you've got to go.  You can always say that we need more data which is great; we need more data.  My only sort of comment would be, it's always hard to prove a negative, and so, therefore, it just seems to me, at least in the uncertainty section, you have to be a little bit more descriptive about the uncertainties that are there.  Obviously, it boils down to lack of data.  But, again, to talk about other uncertainty factors that you can add to throw in, here's another possibility of this.  
                  But, using the standard practice of the day, the Agency did what they're supposed to do.  I mean, the stuff that I am familiar with in my expertise, they did what they're supposed to do.  They took those uncertainty factors and they took those AFs and made a number and got a number off of that.  It was less than one, so they went with it.  Could you have thrown in an uncertainty factor because you only got an N of 2?  Sure.  But I don't know.  Does that really provide an extra amount of safety?  I don't know.  I mean, again, it's kind of what the standard practice is.  And then, do we think outside the box and go to something in addition to that?  Again, I don't know the answer to that question.  If you're following standard practice, they did what they're supposed to do.  If you back up and sort of looked the forest for the trees, then, yeah, maybe you can throw in an uncertainty factor.  You can throw in these other things.  You've got to be careful when you do that because are we going to do that for every compound down the line?  Again, as far as I'm concerned, I think they did what they could with what they had.  Is there more that's there?  Do we need more data?  Yeah, we need more data.  But, as far as what they had, they did what they were supposed to do.
                  DR. DANIEL SCHLENK:  Just to follow up to George's point, so if I had release data, I could estimate concentrations.  I could estimate all those sorts of things, but, without actual measured exposure data, I'm still just estimating.  So could we propose -- are you thinking about a worst-case scenario in a model like a level three fugacity model to estimate the concentrations in the environment and see if they correspond with all the appropriate adjustment factors toxicity to organisms?  I still don't see how the release data's going help me unless I have a model that looks at the overall fate within that particular release scenario in environment unless I have measured concentrations, so I have monitoring data.
                  DR. GEORGE COBB:  So I would agree that the monitoring data would be better.  But, I mean, the model would be perhaps the best way to proceed in the absence of monitoring data.  I think where this whole program needs to go is to monitoring data in advance of these assessments so that you can inform the assessment.  But, again, that's a different discussion.  Yeah, I think a fugacity model or some type of -- whether it's a WHAM model or something in a stream bed that's more specific for the stream could be used.
                  DR. KENNETH PORTIER:  Dr. Blystone.
                  DR. HENRY ANDERSON:  I could comment a bit on the fish, I think.  The challenge there -- I mean you can do a generic for their stored fish samples that EPA has on Lake Michigan, but that would largely show probably negative.  But the issue would be is there a discharge into a river or stream somewhere and some settlement in sediment and then you'd want to sample the fish that are actually resident of that area rather than fish in multiple areas.  That's kind of what's done with the Superfund program when you have a waste site where you may have some discharge into a stream.  You then pass the fish around that pond or wherever it is, but that's quite an undertaking.  The only thing would be, if there is a known discharge into a given stream, you could then query the state where that is and see whether their fisheries program has done any sampling in that area and may have stored specimens that could be analyzed.  
                  So that would be kind of a gray literature kind of thing that there's a lot of.  For instance, there's a lot of monitoring done in Wisconsin that isn't really published, isn't necessarily sent to EPA either, so that's a concern.  That would be a group of people upfront you could query to see, do you have samples in this given region?  And do you have stored tissue that then could be analyzed?  So that could be relatively quick if there is stored specimens to do that.  That's basically what EPA has done with a lot of the emerging chemicals, is they go back to the tissue bank and see.  
                  Now, with Lake Michigan, there just happens to be a single long-term species of lake trout that may not be very sensitive to this particular compound.  Again, as water runs down the stream, there's been quite a bit of monitoring of discharge into the river.  And many of these chemicals, when you have running water, you have higher rates of evaporation from the surface of the river.  So all of the compounds evaporate pretty quickly.  As I said, it's a bigger task other than to have a system to query if somebody has done that that hasn't gotten into your data collection or data identification process.
                  DR. KENNETH PORTIER:  Thank you, Dr. Anderson.  Dr. Blystone.
                  DR. SHERI BLYSTONE:  Yeah, I just wanted to make a couple of comments on the hazard side of things: number one, to remind everyone we're talking about industrial chemicals, not pesticides designed to have an effect.  I think the data that we have for this molecule, I think of it from the DHS hazard classification side, clearly puts it at moderately toxic -- clearly category three, possibly category two.  So the categories go less than one concentration limits that cause effects to populations, one to ten, ten to a hundred, or greater than that, right.  Ten to a hundred seems accurate.  It's about category three, moderately toxic.  I don't know that any additional data's going to give you anything different than that.  Typically, you test the aquatic species because they're more sensitive in most cases.  And then if we're talking about the conditions of use, we're going along that lane.  From my perspective, going out and trying to do several more species isn't necessarily going to get us any further along on the assessment.
                  DR. KENNETH PORTIER:  Any additional comments?  On the eco side, the only kind of strong recommendation I heard was the additional that more recent reference Dr. Cobb presented to a different adjustment factor that might be used.  I didn't see a consensus for going out and getting more environmental data at this point given the bulk of chemicals that we're going to be looking at.  That's the other thing I guess we have to think of as cost benefit on some of this stuff.  Dr. Barone, any closing comments?
                  DR. STAN BARONE, JR.:  Just a couple of clarifications, the reference that was referred to, we definitely would like to get that reference about the assessment factors and the discussion around the assessment factors.  
                  We've mentioned this before; for TSCA, we do not have a minimum data set.  This us unlike what our sister program for food-use pesticides with CFR part 158 where a minimum data set and a rather extensive database is actually required for food-use pesticides.  
                  So, again, this is part and parcel to grandfathering in the TSCA inventory.  We have what we have.  We do have authorities to collect data.  But, again, this is part of the first ten, and the first ten we were told to do risk evaluations essentially from the day that the bill was signed, which did not allow for robust data collection.  
                  So that's the situation we're in.  We'll do the very level best we can to collect and acquire data, and, as you've seen, read-across and make inferences and apply the appropriate assessment factors.
                  Mention of spills is sort of an important context but not a condition of use under TSCA and not something we're assessing specifically, not that the information couldn't inform emergency management and risk communication, but that's not a condition of use.  So, just to be clear, that's out of scope.  
                  Again, I'm hearing you guys struggle.  We have the same struggles internally, so one of the things that we're looking to you is prioritization of the recommendations.  What are the must-dos versus the nice to-dos?  So again, we have essentially less than six months to finalize this risk evaluation to meet our statutory deadline in the buffer in this estuary deadline.  So the more you can help us with identification of data or surrogate approaches, yes, we can go on to do aggregate and add things up.  We've got a lot of criticism for trying to take that approach previously.  If you think that's a strong recommendation, well, please let us know.  
                  Just putting this out for context because some of these same themes are going to be coming at us again and again.  We're dealing with a data-poor environment, in general, and we do need to make decisions in the absence of information.  That's what risk assessment is.  And we're doing hopefully our best in being clearer, and that's what this section's about -- being clear about what those uncertainties are -- and we're making those assumptions.  If you think we can bolster or make more clear what those assumptions are, please let us know in your recommendations.
                  DR. KENNETH PORTIER:  Dr. Barone, I have a question.  Something that Dr. Anderson said and your comments on spills, I know moving forward, not with these legacy chemicals but, EPA usually goes out for a general request for information from the public, and I was sitting there thinking, well, you usually ask for gray literature, but what we've been talking about is gray data.  In fish collections where they may have done the analysis, it's sitting in somebody's -- I can see all the uncertainties that come up with that.  I understand, but sometimes some data's better than no data.  Bad data may be better than no data.  Spill information, while not required, might be available that would help inform.  So I'm just wondering if that kind of request for gray data goes out in those or can go out in those future requests for information.
                  DR. STAN BARONE, JR.:  The request for data with both a scoping exercise and the problem formulation was relatively broad.  I think a lot of stakeholders and a lot of the public were not as clued into that opportunity to provide information that they may have, but many did.  What has happened with the naming of the 20 high-priority chemicals for the next round, we've already started receiving data in the public docket.  We will probably also receive CBI data and there's ongoing discussions about that for some of those 20 chemicals.  
                  In addition, the next public comments, which started a week ago Friday, we're getting additional comments and submissions based on the dossiers and the conditions of use that are presented in those dossiers.  So we hope to make this a more interactive process where we get information in a voluntary way and also have more lead time to obtain data that we don't necessarily get voluntarily.  The input from this committee and what we have experienced with the first ten risk evaluations will help us sort of focus in on critical data needs for the future.  It's informing us.  We're also looking at how to prioritize those efforts, so that's an important issue as we look at our pipeline and downstream because it does take time to generate information.  
                  But it also is important to recognize that some of this data is existing data.  And how do we get that existing data without impacting CBI and proprietary issues?  These are all legal consequences.  It's not just the technical issues that I deal with day to day, but there are issues that we have to, of course, consult with our Office of General Counsel.  
                  So it's complicated.  I get it.  I appreciate the comments and discussion about the conclusions, the risk estimates that go into the risk determination.  It was a very important exercise again that Dr. Davies did.  It's very similar to the worksheet that we have.  We have additional things that you didn't identify in that worksheet.  I appreciate you made that effort, and I think it helps display the transparency.  
                  From the last comments from the last meeting, I also want to say, we've been working on trying to make that presentation.  We didn't get into this risk evaluation, but we do plan to present that summary table with the risk inclusions, the risk drivers, for every condition of use so that you see how many different decision points there are within the condition.  And we're putting numbers in there, not yes/no, because we think the numbers actually tell the story because there's uncertainties associated with each of the numbers.  And we're talking about high-end versus central tendency estimates, and you've already discussed with us all the different factors that are included.  
                  So, hopefully, you'll see that in the very near future in the next round.  We're sorry we couldn't get that into this assessment for you, but again, with timing, there's a lot of issues with the timing.  We're trying to get these through our internal review process and get them out to the public for public comment.  And, hopefully, your comments are still affecting things that we're trying to make sure that we track back to.
                  DR. KENNETH PORTIER:  I think Dr. Willett and I just realized that I didn't make it down to the end of my list, so we needed your comments.  Sorry.
                  DR. CATHERINE WILLETT:  Yeah, that's just what I was going to mention.  So we didn't quite finish 7.4.  But my comments are not substantive, but I did want to chime in a little bit and support the idea that this risk determination description is of the considerations and the conclusions are, in my mind, a vast improvement over the previous two that I read through and are well supported but very, very difficult to walk through as Holly mentioned.  It just took hours and hours and hours to really check and go through.  I don't how everybody's going to check and go through everything, but that was sort of our job.  
                  And it's, as you mentioned, extremely complex, so I'm really excited to hear about this table that you're putting together with the numbers and everything.  
                  And my comment was really, again, just to try to get at this issue of including assumption of PPE use in the risk determination and trying to tease that out and maybe figuring out how to separate those two issues: the risk management and the risk determination.  It's a reiteration of what we had in the previous discussions or the discussions of previous chemicals.  But I have a long paragraph that talks about that, which I will not read because it's very detailed as there are so many different conditions.  I just wanted to get that particular thing on the record and also, basically, just support the direction that you guys are going.  I think each one is improving and that's it.
                  DR. KENNETH PORTIER:  Thank you.  Let's go ahead and take a 15-minute break.  I have 11:42.  I mean, 10:42.  We'll come back at 11:00, and I'll start of Question 8.
                  
                                    (BREAK)

                  DR. KENNETH PORTIER:  Dr. Cobb tells me he has answered some of the questions that we had earlier before and if those were going to go into the notes.  So we're good with Question 7 so let's move on to the general content and organizations Questions 8.  And I've asked Dr. Anitole to just read both parts in and we'll move forward.
                  
                               CHARGE QUESTION 8
                  
                  DR. KATHERINE ANITOLE:  Question 8.1:  Please comment on the overall content, organization and presentation of the draft risk evaluation of 1-BP.  Question 8.2:  Please provide suggestions for improving the clarity of the information presented in the documents.  
                  DR. KENNETH PORTIER:  Thank you.  We asked EPA to move this question from Question 1 to Question 8 so that we could, in a sense, look at this as what additional comments on content, organization and presentation, and what additional suggestions that we haven't discussed up to this point in the previous seven questions, because there's a lot of suggestions there.  So, we've asked Dr. Twiss to lead this discussion.
                  DR. MICHAEL TWISS:  Thank you.  Want I'm going to do is directly respond to Question 1 very briefly.  And then Question 2 has more of the criticisms, both positive and negative.  First of all, the oral presentation made by Dr. Anitole was really very clearly made and well-organized.  Many people on the panel here said it would be an excellent template for the Executive Summary.
                  The content of the draft risk evaluation falls a template for chemical risk evaluation in each section that's complete as possible, notwithstanding the scientific arguments for more information, which are stated in the document and which we discussed earlier in Question 7.  In several instances, reviewers have shared examples of improvement in data presentation compared to previous risk evaluations, and so they consider it to be a better risk evaluation than the previous one done in 2016.
                  There was widespread concern among both panel members and public respondents with the shortened comment period.  Consideration of public comments during peer review preparation and its critical function and continues, truly, in challenging under the timeframes used in this risk evaluation.
                  Now for Question 2, and I did a bit of crowd sourcing here by requesting folks on the panel to provide information, and I'll state their names if they're not associate discussants.  And so, there is a request for improving the clarity of information presented in the documents, and the presentation techniques.  For example, the use of hyperlinks is useful, but a direct link to a document rather than to a database where it can be sought if proper references known is not really that helpful.  And so, hyperlinks are nice.  If there's a reference, you hit the reference hyperlink and you've got it right there, rather than going to a database where you don't have to go mine for it.
                  Catherine Willett stated with reference to question 5.2.  Although description and the weight-of-evidence, considerations for noncancer endpoints seem logical, the actual rate of evidence process is unclear; for example, what the weighting system is and how each datapoint was assessed and the result of that assessment.  And the table in Appendix I-2 was not that informative.  And perhaps, the second table showing the weight-of-evidence analysis for each endpoint with endpoints chosen would be helpful.
                  Steven Bennet mentioned something, but he'll be speaking afterwards.  Holly Davies stated a number of issues, some of which were mentioned earlier, so I'll just talk about them in more of an overview rather than getting into a lot of the details.  And it states that the document goes through smaller pieces for environment, occupational, consumer uses in this repeating loop, which makes it hard to follow in each major section.  And the presentation by the EPA on Tuesday was easy to follow when one topic was presented from start to finish, and that's referenced to Dr. Anitole's extended summary.  And so, the document would be more readable if it followed the same format as that presentation.  
                  The Executive Summary was too long, confusing in spots, had too many acronyms.  It didn't seem to have a consistent audience in mind, and I'll be speaking about that at the end of my comments here.  All right.  There's a number of comments that are made, and I won't repeat them here, but they'll be included in the written comments.  They're both requests for added clarity and comments on various tables, which were well executed.
                  There are some requests for better clarity in expressions.  And so, in some cases and throughout certain tables, scientific notation is not used consistently, and it really doesn't need to be used at all.  And so, when you do those kinds of things, you have to be careful that you're not making errors, and so the reading it is difficult.  
                  There was also a comment on use of colored titled information in tables.  And besides issues of inclusivity since a significant proportion of people are color blind -- so to our most computer printers.  And so if you're printing this off and handing it off to someone to read, they might not be able to get the same information that was intended.  And so, having color to bring out a point is really nice, but there can be an error in doing that.  So a different approach as an alternative should be sought.  There are some cross-referencing errors, and again, this will be in the written comments.  And some consolidation is required, particularly the WOE discussion is mostly rehash of hazard discussion that could be combined into one section.
                  Risk communication issues, again, we have reference to Question 7.  Table 5.1 was hard to read, and Dr. Davies did present a modification of the table, which I think would be a nice thing to incorporate.  There are a number of typographical errors.  I'm listing some, but it's not complete.  Definitely, a careful draft reading is required.  And finally, for my comments here, I'm going to state that this document is written purposely for us, a peer review audience.  So I went into the public comments and mined in and looked at some -- just your basic folks who are writing in on comments of those people who weren't representing an NGO or company, and there are clearly some individuals that are clearly uninformed about the properties of the chemical.  Some people may consider all chemicals to be bad and some of these public commenters may be in that group.
                  Nevertheless, the EPA does need to provide an educational service by composing some summaries in lay language.  And I suggest that summaries in lay language, for example, at grade six level -- and even at a primary language level, like a grade 2, should be included.
                  There's a strong correlation between literacy and health, and the situations that are different from medical health concerns and the need to communicate with its constituents.  And so this production of a lay summary would foster better communication with the public and help gain its trust that this institution has relied on for evaluating safeties of chemicals in our society.  And that's all I have to say.
                  DR. KENNETH PORTIER:  Thank you.  I wanted to make sure you had this missed link on page 153 and 181 that's referenced to Appendix H, which should be Appendix I.  It occurred twice.  If it had occurred once, I'd have disregarded it, but it occurred twice.
                  DR. MICHAEL TWISS:  Yes.  That cross-referencing error is noted here.
                  DR. KENNETH PORTIER:  Good.  Dr. Bennet?
                  DR. STEVEN BENNET:  And I think he captured it quite well.  I made the comment about the discussions of the uncertainty, and we've had a lot of conversations throughout the past three days about making sure that the uncertainties are well described.  I particularly like the consumer exposure piece.  That was the area I covered, but I did like that discussion within there.  So I think there may be some lessons that can be repeated in there to make sure those things are addressed.
                  I think one area is to make sure that the information is comprehensible when you take discussions within sections and then you summarize it in a table.  One example in Section 2.3.1.8, there in the body of their text, one of the processors manufacturers noted Envirotech.  But when you go to the table that summarize that particular section, they were not noted.  That may be because they're captured with in the CBI portion of it, but I'm not sure.  I'm not really sure that information is in fact correct and accurate, so it represents both sides of it.  But certainly, learning by doing is definitely occurring as the evaluations go through, so I appreciate their efforts on that part.
                  DR. KENNETH PORTIER:  Dr. Blystone, I think you're next.
                  DR. SHERI BLYSTONE:  I agree with everything he said.  There's a few typos and stuff I'll send along, but I've nothing else to add.
                  DR. KENNETH PORTIER:  Then we swing back to Mr. Kaufman.
                  MR. ALAN KAUFMAN:  Okay.  Yes, I think a lot of this has already been cover.  I think as a general comment, yes, there are areas of uncertainty.  I think there are always going to be areas of uncertainty.  I think Dr. Blystone said it well, particularly about the environmental piece, but I think it applies generally.  I don't know that we're going to resolve that, and I don't know that additional work would necessarily change the conclusion, so I think it's a solid piece of work.
                  So I think where I do have issues is in presentation.  And I would echo some of Dr. Twiss' comments that when you think about a lay audience, they're probably not going to read any further than the Executive Summary.  And so I think making sure that that's very clear, and the language is at a level appropriate for that audience would be very helpful.  And lots of color and graphs always is helpful.  But generally, it's a very solid piece of work and I wanted to commend the EPA staff for the work that's been done here.  When you sit through one of these meetings, you get three days' worth of people criticizing what you've done.  But I think the issue is more one of presentation and explaining decisions well so that there's an understanding of how you got to where you ended up, and I think that will also help allay further criticisms of the evaluation.  Thank you.
                  DR. KENNETH PORTIER:  Ms. Rudel?
                  MS. RUTHANN RUDEL:  Thanks.  I have fundamentally three suggestions, and I'm going to just talk a little bit about why I think they're important.  But they relate to general population exposures, availability of data about when chemical is used, emissions, discharges and also the use of MOE on presentation, and a couple of presentation data questions that could make the report more accessible to people who aren't specialists.  
                  But I want to reiterate that I thought the report was clear, and that EPA has done a nice job of estimating risks associated with these named conditions of use using information they have been able to gather on worker and consumer exposure.  And the toxicity database is fairly complete, and it does show the chemical is neurotoxic, reprotoxic immunotoxic carcinogenic with reasonable evidence for a mutagenic mode of action based on the electrophilic activity.
                  So, it's important that EPA moves forward sort of quickly to reduce risks that have been identified in the risk evaluation.  Worker and fence line community and neighborhood exposures to this chemical had been recognized as a problem for over a decade.  Earlier, people were talking about OSHA and not doing anything, but OSHA's enaction on the problem isn't a reflection of their lack of concern, I think, but rather of social and political realities that have made these risks hard for them to address.
                  And similarly, this chemical was proposed a decade ago to list as a hazardous air-pollutant with still no action.  And it seems optimistic to think that that's going to all of a sudden change.  And it's odd to have this report assume that that's going to happen and be effective.  And so this TSCA is what's in front of us now, and it does provide EPA with an opportunity and responsibility to reduce worker and general population risks, and I hope that that will happen.
                  So I've heard you say obviously you're not going general population risks, but I don't really think that's consistent with the law maybe because of the poor quality of data about chemical use and emissions.  So the document doesn't really support the decision not to consider general population exposures, which could occur through groundwater contamination and subsequent drinking water or vapor intrusion or indoor concentrations from consumer products, as well as air and neighborhood emissions.  But if you do maintain the narrow focus that you have currently, to be very explicit about what's in scope and what's out of scope to facilitate space, to take action without running into preemption barriers.
                  So for this and all the priority chemicals where you don't have good information about how the chemical, where it's produced and used and disposed and so on, I recommend doing some kind of data request that's going to give you that information because that's a really big source of uncertainty that underlies the whole assessment.
                  A presentation thing to help make it more interpretable and to help people think about general population exposures, I personally liked Table 5-1.  I found it helpful in understanding what you had done and so on.  And what I was looking for that I thought would be helpful would be, what would be the target indoor air concentration associated with a cancer risk of one in a million, associated with an MOE of a hundred or a thousand, whatever you set it for for a few of the key noncancer effects.
                  So to come up with what would be a risk-based screening level air concentration for a 24/7 residential child/family exposure scenario, and then also to come up with an eight-hour time weighted average worker exposure that's associated with those target risk, 10 to minus 4, in the MOEs for workers just to be able to have some air concentrations as benchmarks.  I think it will help people understand what's the magnitude of range that we're talking about within this acceptable risk range for a short term or worker.  And also, if new data come if somebody does air testing in a home, if somebody does air testing at a fence line or in a workplace or personal air, to have those concentrations for comparison is always really helpful.  So I know they're in there somewhere, but that they were being more brought to the floor.  And I know that that doesn't consider dermal, so it's just for inhalation, but I still think it would be useful.  
                  And then my last comment was about the margin of exposure presentation.  And in my experience -- well, that that margin of exposure are the two -- essentially the RFC -- not to the point of departure.  You've got the point of departure, then you've got these uncertainty factors of a hundred.  But the margin of exposure, as I'm seeing it, you're doing it to the point of departure.  Unless I missed it, that's what it looked like to me.
                  DR. STAN BARONE, JR.:  So the document does -- in several places -- actually it's both in the hazard section, but also in the risk determination.  It describes what the margin of exposure is.  It's the point of departure for the hazard divided by the exposure estimate.  That's separate.  That's the MOE for the risk estimate, which is separate from the benchmarked MOE, which is made up of the concomitant uncertainty factors.  That's not to be confused with an RFC or an RFD because the RFC is actually a different equation.  You're taking a point of departure and you're dividing the uncertainty factors into that point of departure.  Exposure is not considered in the RFC or RFD, so it's not a risk estimate, whereas the MOE is a risk estimate.  So there's a big distinction between the two.  There is a commonality, and the commonality is the PODs are an enumerator in both estimates.  But one's a hazard estimate; the other one's a risk estimate.
                  MS. RUTHANN RUDEL:  Yeah.  Well I think that it could be misleading for some people.  When you go into the point of departure, some of those effects are serious.  And I'm not sure that it's easy for risk managers to understand that you're talking about all of your uncertainty.  And if there's co-exposures and anything else, it's got to be within that margin of exposure.  The two uncertainty factors are -- I don't know.  When you say there's unreasonable risk, it's because the exposure is the point of departure divided by a hundred divided by the exposure.  No; it's just the point of departure compared to the exposure.
                  So it's kind of tight.  It doesn't have a lot of room for error, I guess.  And the uncertainties are big and that's not a very forgiving calculation, in my opinion.  Your target was to have the exposure be a hundred-fold lower than the point of departure, basically.
                  DR. STAN BARONE, JR.:  Not always.
                  MS. RUTHANN RUDEL:  No?
                  DR. STAN BARONE, JR.:  Not always.  The benchmark is not always a hundred.  Sometimes the benchmark is 10, sometimes the benchmark is 30.  We have benchmarks up to a thousand.
                  MS. RUTHANN RUDEL:  Wasn't in a hundred in most of them in this?
                  DR. STAN BARONE, JR.:  In many, the benchmark was a hundred, but there's also several where it's 30.  So the benchmark relates to the endpoint and relates to the study type.  It's not always the same.  Again, the uncertainty for that margin of exposure -- and that's one of the reasons -- and this is different.  We present multiple PODs for the effect domain and we present what we think is the most sensitive and robust for that domain.  And then when we look across the risk characterization -- you see this in the risk characterization section -- we actually show there's a whole range of effects.
                  MS. RUTHANN RUDEL:  Yeah, I saw that.
                  DR. STAN BARONE, JR.:  And what we're trying to do in the risk characterization is actually inform you, inform the risk managers between the most sensitive and the next effect domain what's the magnitude of difference.  Are we within the order of magnitude?  Do we have five different adverse outcomes with the same order of magnitude?  In fact, that's the case we have with 1-BP, so I think we've made that clear.  If you have some suggestions to make that more clear, please help us.
                  MS. RUTHANN RUDEL:  All right.  
                  DR. KENNETH PORTIER:  Thank you. Dr. Schlenk?
                  DR. DANIEL SCHLENK:  Yes.  I was fortunate enough to be on the 2016 document, and I got to say this one is way better.  So it's encouraging to see, when you make comments, that they're actually incorporated into the next iteration, so I thought that was fabulous.  I thought the document was fabulous.  I'm with Jim on that.  I just thought you guys did a fabulous job in terms of the presentation overall, so kudos for that.
                  The only recommendation I would make -- and we talked about this before -- is, again, the uncertainty aspects.  I think that needs to be expanded, particularly when you're saying there's not an effect.  I think that's critical to expand on that.  And honestly, the only topic that I brought up in 2016 that was not put into this document was using adverse outcome pathway paradigm, particularly in the weight-of-evidence approach. 
                  I know you guys did mode of action in there, and the last figure is nice, but I just think the AOP approach provides a way to look at uncertainty and the data gaps that are present, and it allows you to describe those a lot better in the uncertainty analysis.  Some people actually proposed using it in the problem formulation step as well to actually identify the gaps that you have there as well.  But since we're not part of that, obviously you can't make comments on that.  But at least in the characterization section, having that in there is a real plus.  
                  And we're very fortunate today to actually have the guru of AOP sitting right next to me here.  I know he has some really nice examples of slides that can be utilized here.  And particularly, for this compound, the narcosis AOP which is pretty solid in terms of its science and in terms of what predictions can be made from that, I would totally put that in your eco evaluations, particularly for other compounds that are going to go through that which, according to the list I've seen, they're going to be very comparable in terms of their mode of action.  So what Ed is going to present here can be very useful for future documents that are coming down the pipe.  
                  DR. KENNETH PORTIER:  Dr. Perkins, do you have something to present?
                  DR. EDWARD PERKINS:  Yes. 
                  DR. KENNETH PORTIER:  Great.  
                  DR. EDWARD PERKINS:  And the computer's turn down.  Well there goes my problem.
                  DR. KENNETH PORTIER:  While they're bringing it up, let me add in my comments while he's bringing that up.  And I just emailed these to Dr. Twiss, so he's got them.  My comments are around organization, and they are similar comments to what I've made previously.  In the introduction, Section 1.1 really should answer the question of why this chemical is on the TSCA priority list.  It's really not answered, but that really should be right there at the beginning before we talk about chemical properties.
                  The second thing is a comment we made yesterday about potential exposants susceptible to populations.  They're raised out in 2.4, but then they're kind of hiding in Section 3, and hidden even deeper in Section 4.  And I recommend you have a 2.4, a 3.3, a 4.4 and a 5.2 that actually highlight and discuss specifically past -- you're already doing it; it's just moving it up in the outline in some of those subsequent chapters.
                  And finally, in the scoping Section 1.4, I looked at that and I go, why do they present the conceptual model and then talk about the conditions of use included?  I think you really should present the conceptual model, Section 1.4.1.  Then Section 1.4.2 says, "conditions of use included in this risk evaluation, risk justification," which you already have.  And then Section 1.4.3 should be "conditions of use excluded in this risk evaluation," and justification for that.  And that kind of answers a lot of questions that the panel have had about why did you do this and not do that.  
                  And sometimes you're not telling us what's in the model and what's excluded.  I think it's in there somewhere.  It's in that Section 1.4.1.  You just need to kind of pull it out and highlight it separately.  So those are just some practical organizational suggestions that will improve the layout.  And with that, I'll turn it over to Dr. Perkins for his comments.
                  DR. EDWARD PERKINS:  So, you mentioned wanting to see how you might be able to use AOPs or put them in, so I pulled out a few from some papers and from the AOP-Wiki and some that we're developing.  So these AOPs kind of demonstrate how you might start organizing general effects that you see proceeding from.  And the first case here, MIE is the molecular initiating event.  And we didn't know what it was at this point, so we had a question mark, so you don't have to know everything about it.  You can just put question marks there and it'll be data gap.  But it lets you arrange things along the order of the mechanism of action up to your adverse outcome point of interest.
                  So in that case, it's very similar to the cancer endpoints that you're talking about here.  And then narcosis is pretty explicitly one of the mechanisms of action.  That's only in eco.  And even in human exposures, you see some kind of narcotic-like effects as seen with the confusion, loss of equilibrium and other things.  So there's a great log KOW of compounds very predictive of that outcome, so you can use essentially a quantitative AOP based on that.  It also lets you take the data and align it with these essentially different categories with in it, and then you can do weighted evidence as Dan mentioned.  What kind of evidence do you have for which biological event will let you proceed down the line.
                  So I'm also thinking it's a good mechanism to look at how you might organize information and translate it into context of how it's acting.  So, next slide.  So in the AOP-Wiki, which is a database that has a number of AOPs in it, there's a few there for liver tumors and hepatocellular carcinoma and liver cancer, some of which are mutagenic mode of action, some of which are sustained AHR activation, one of which is set to E-1 activation where it cycles, a rat gets stuck in a cycle and keeps producing oxidative stress.  So all of these might be part of the multiple modes of action that you've been talking about here.
                  It allows you to sort of build a hypothesis and then see where your data fits in with that, whether it's quantitative data or not.  And then the dose response data can inform rates of going from one key event to the next.  And so you can start using that to build the key event, build the quantitative models.  I liked it mostly because it helps me understand things in a more intuitive way.  It takes all the descriptive tables and information you have and put it more in a flow-chart manner, help to understand things better.  So I'm kind of presenting it in an interesting way to organize different things.  Next slide.
                  And of course you can get more complicated.  You can put as much detail as you want in.  Next slide.  Or you can have the whole mess right there and have networks.  But part of the point is, is you can keep it as simple as you want to keep the communication down, or you can expand it out to include all of the different measurements you have, whether it's the chemistry, whether it's electrophiles, protein misfolding, cellular evasion, different things like that.  You can start putting them together in more of a logical context that might act as a way to organize your thoughts on how things are functioning, so I thought I'd put that out there.
                  DR. KENNETH PORTIER:  Are you using a concept mapping tool to do this?
                  DR. EDWARD PERKINS:  Cytoscape; so it's a freeware thing.  And we have a tool called AOP Explorer where these networks are available free, and you can draw your own.
                  DR. KENNETH PORTIER:  I used to do these and then people would just laugh.  And I looked at this and I'm thinking, okay, and then the EPA communication people can write this up at sixth grade level.  That's the problem with the communication recommendation.
                  DR. EDWARD PERKINS:  Right.  And so, then you'd back up over to the simpler ones.
                  DR. KENNETH PORTIER:  Right.  You got to the two or three previous. 
                  DR. EDWARD PERKINS:  Right.  And then the nice thing about these for risk assessment, these are all causal linkages that if one activates, the next one happens.  So you don't put in things that you know don't happen, so it helps you winnow out all things.
                  DR. KENNETH PORTIER:  So where do you see something like this going in this document?  I mean, I'm just trying -- 
                  DR. EDWARD PERKINS:  Well, no, it can go multiple places.  So there's was a lot of uncertainty around the effects and how it was causing different toxicities in the human risk assessment component.  So it could act there as different hypotheses of how it might be acting, and then you can talk about which data aligns where.  And you can have dashes in between airways you think it's going; and then you can build a weight-of-evidence based on that.  So what's the likelihood of that hepatotoxicity goes to liver cancer based on this chemical and things like that.
                  So each of the AOPs, when you look at a specific chemical, it's a hypothesis for how that chemical might cause cancer or whatever endpoint you want.  And then each of the boxes essentially is where you have the evidence for that happening based on literature and experience.
                  DR. KENNETH PORTIER:  So this image here, which is liver cancer, you'd have it in the human health hazard liver cancer section.
                  DR. EDWARD PERKINS:  Yes, you could do that.
                  DR. KENNETH PORTIER:  Dr. Schlenk?
                  DR. DANIEL SCHLENK:  Yes.  I would put it actually in the characterization uncertainty section.  That's normally where you see it.
                  DR. EDWARD PERKINS:  That's pretty good too.
                  DR. DANIEL SCHLENK:  In the weight-of-evidence, I'd address the uncertainty that's there.  That's kind of where most people have proposed to put it, is in that section.
                  DR. KENNETH PORTIER:  Okay.  But when we make this recommendation, that they look at this and use this as part of the presentation or the modification of risk evaluations.  It's nice to be able to say, well we see this as a tool in discussing uncertainty and laying out the weight-of-evidence.  I understand that.  
                  DR. DANIEL SCHLENK:  Also, the other thing, it helps when we're talking about mutagenic and non-mutagenic modes of action on that other question, because it actually lays out the gaps that are there and the uncertainties that are there.  But in a figurative way it allows you to make more certain statements, at least with regard to those mechanisms.  I mean, I see it as a way to get at that as well.
                  DR. EDWARD PERKINS:  Yeah.  Also, if you have a quantitative AOP such as the narcosis ones, you can take that actually and see does your chemical -- is it predicted to cause narcosis based on the statistics that we know or mechanisms that we know that we have.  So in the characterization stage, it could be used for that too.
                  DR. DANIEL SCHLENK:  I mentioned this the other day, and I'll put it in the comments for Question 4.  But this is how, when you don't have data, you can actually use this to predict at least acute tox data from that mode of action for solvent-like compounds that have long KOWs between, like two to five, two to six.  It's pretty solid in that range in terms of predicting, at least LC50 numbers proficiently.
                  DR. KENNETH PORTIER:  Dr. Cobb, you want to follow-up on this?  I see Dr. Davies is -- 
                  DR. HOLLY DAVIES:  If he has a follow-up, he can go.
                  DR. KENNETH PORTIER:  Yeah, if he wants follow-up, yeah.
                  DR. GEORGE COBB:  So, a couple of things.  First of all, I'm assuming that the low long KOWs are higher in narcosis, is that correct, low log KOW gives you a high narcosis -- 
                  DR. DANIEL SCHLENK:  No.  Usually they're directly proportionated.
                  DR. GEORGE COBB:  Directly proportioned, okay.
                  DR. EDWARD PERKINS:  Yeah, if you want to put it to that actual in the slide.
                  DR. GEORGE COBB:  It's okay.  
                  DR. EDWARD PERKINS:  Yeah, because they're accumulating the membranes. 
                  DR. GEORGE COBB:  So it's an accumulation in the membrane.
                  DR. EDWARD PERKINS:  Higher would mean more in the membranes.  
                  DR. GEORGE COBB:  And then would you envision a presentation of this having highlights around the boxes that you have high confidence in, or just presenting it and writing text around it?  How would you envision that?
                  DR. EDWARD PERKINS:  You can do a lot of things.  You can do different colors and different sizes.
                  DR. DANIEL SCHLENK:  Different lines.
                  DR. EDWARD PERKINS:  Different lines, thicker lines, dotted lines.
                  DR. EDWARD PERKINS:  Dotted lines.
                  DR. DANIEL SCHLENK:  Dotted lines.
                  DR. EDWARD PERKINS:  Four-dimensional lines.
                  DR. DANIEL SCHLENK:  Polk-a-dot lines.
                  DR. KENNETH PORTIER:  I'm still envisioning this.  So how many of these pathways have you already identified?  I'm thinking in terms of, well, you know, a new chemical comes up, I'm going to take this box and put this in and this one in, and I'm going to talk about them.  It's a limited number, or is this created each time a chemical comes up?
                  DR. EDWARD PERKINS:  Well that's why they say they're supposed to be chemical-agnostic, so it's really about -- okay, what happens if I get something that accumulates in the membranes?  What cascading effects does it have?  What does it lead to?  So, any chemical that would accumulate and activate something.  So for the SIP 2-E1, if something interacts with that and it gets stuck in cycling and causes excess of oxidative stress, any chemical that would do that would cause the same pathway to be activated.  
                  So right now, there's probably 100 or 200 with multiple species at the AOP-Wiki, some more in publications, but they're meant to be reused and recycled.  As you can see if you go to the next one, basically those are three different ones, but they have some of the same things.  They have cytotoxicity or they have regenerative proliferation, self-proliferation, cellular proliferation and hyperplasia.  So those are being really reused within different things.  So they got different ways to get there, but the actual mechanisms -- the descriptive things are still being reused.  But for your particular chemical in a regulatory context, you have to overlay your data that shows that your chemical actually does it.
                  DR. KENNETH PORTIER:  Any questions on Dr. Perkins?  Dr. Willett?
                  DR. CATHERINE WILLETT:  Well maybe I have some questions, but I wanted to just stress and point out that this is a terrific example.  And I'm so glad that you brought this up and that Ed was here because Ed has this really great visualization tool for the AOPs.  But even so, this is a really interesting application that would be great to include because it would demonstrate the utility of this information base in assessing the weight-of-evidence for a decision, even when you don't have a lot of information about the chemical.  So what people don't really understand is the utility of the biological information that's in this information base is relevant to any chemical in any context if you can map your chemical activity on that wonderful picture that Ed showed, which there are some end-routes there.  
                  So the whole discussion that Dr. Eastmond had around the mode of action, he could substantiate that in this pictograph.  You could show it there and it's a lot easier to grasp or to just see, oh yeah, thick bars, thin bars, and here's the biology that's supports it, and you look at all these gaps.  So that's an amazing example of how to use this database.  And there are other aspects of it that would substantiate some of the other biological discussions that we've been having.  So, I'm actually very excited and I really support the inclusion of some discussion of this in the weight-of-evidence.
                  DR. KENNETH PORTIER:  I was going to ask Dr. Davies if she wanted or -- 
                  DR. HOLLY DAVIES:  I have a different topic, so if you just remember that I have a different topic that we can get to after the AOP discussion.  But I'm also excited that we're having the AOP discussion because I think it's great to include.
                  DR. KENNETH PORTIER:  Dr. Eastmond?
                  DR. DAVID EASTMOND:  While I like the concept of having figures, it helps me follow things.  The challenge with this chemical is, your predictions are really nice, but the in vivo evidence does not support the predictions, and that's the fundamental problem.  So you can go through and say, yeah, a direct-acting alkylating agent will bind to DNA and that will cause a mutation and this.  And you can support this in vitro, but when you go in vivo, which is the higher, more reliable studies, those all fall through.  
                  And so, you can put forward all these predictions, but fundamentally you're going to have to say at some point, reality hits and now we have to make a decision based upon our predictions or what we see.  So I see that as being useful in one sense, but it doesn't answer the key question for me here.
                  DR. KENNETH PORTIER:  Down to comment point now.
                  DR. DEBORAH CORY-SLECHTA:  Can I make a comment?
                  DR. KENNETH PORTIER:  Dr. Cory-Slechta, do you want to step in here?
                  DR. DEBORAH CORY-SLECHTA:  Yes.  I want to highly concur with the comment that was just made.  I think we have to be very conscious about these.  There are so many limitations of how this data was derived and how it predicts or doesn't, what actually goes on physiologically that I'm concerned that we're over excited about something that isn't ready yet for what it's being proposed as.
                  DR. DANIEL SCHLENK:  To counter that, I look at this a little bit differently because I think it makes predictions that allows you to show that the in vivo data doesn't match that.  It's stating a hypothesis that it should go this way, but your in vivo data doesn't support that, and it gives you a mechanism to show that it's not supported.  And it provides you with an alternative hypothesis in terms of whatever mode of action that you're getting for that effect, if that make sense.
                  DR. DEBORAH CORY-SLECHTA:  You're going to explain all that in the documents?
                  DR. DANIEL SCHLENK:  No, you already are.  I mean, the documents are already there.  We just spent half the day yesterday talking about it.  So, to put it in a format that allows you to see that graphically makes things simpler to understand.
                  UNIDENTIFIED FEMALE:  I think Dan wants to respond to AOPs.
                  DR. KENNETH PORTIER:  Dr. Barone, do you want to comment or ask questions?
                  DR. STAN BARONE, JR.:  Yes.  So in my previous life and several previous lives, I've worked on mechanism in the AOPs and developing AOPs.  And again, I think some of the practical questions that have come up here from the committee are key concerns for us.  And I mentioned earlier, quantitative, that AOPs are really where we need to go, and that's where the rubber meets the road.  Again, we have to have a defensible presentation.  What's the predictive validity between the A box, the B box, the C box, or the A to C box?  And I think your narcosis AOP is fitting that bill, because we've got to be able to defend the stakeholders, hey, we don't have this data, but we can use surrogates.  We can use this other biological information for X chemical. 
                  Again, I think that's one of the important approaches.  We talked a little bit about this in BP29 with the lung overload in our approach to use surrogate information for lung overload.  We'll probably be presenting some of this in future assessments, but again, it's an important research tool.  It's an important hypothesis testing tool.  We're still, from a practitioner's approach, still behind on implementing this into our risk assessments.  
                  Our risk assessments are based upon apical endpoints, apical data.  What we've struggled with 1-BP here this week was, here's the cancer incidents, here's the tumor incidents.  We have multiple hypotheses that get us to cancer.  We have multiple hypotheses to get us to neurotoxicity and developmental toxicity.  At the end of the day, as a risk assessor, I'm going to make a risk guesstimation (phonetic) or a call based upon that dose response for that apical endpoint. 
                  We would like to have more confidence, more relevance, more robust explanation of the relevance, and I think the AOPs will help us in that regard, but I don't think it's a panacea yet.  There's about less than two dozen peer reviewed AOPs in the Wiki, quote/unquote, validated the last time I checked.  There's about 86 that really have robust documentation the last time I checked.  And it's been a little while but, again, Dr. Perkins, it's not the universe of all the hypotheses that are proposed in the literature.  And the Wiki is a wonderful platform.  And at least in my experience, it's sort of been an effort where other duties as assigned, there's not a fully funded effort of the Wiki supporting this.  So that's something that EPA has contributed significantly to, and several other agencies and academics, but it's not this greatly funded effort.  So, if you guys think this is important, let us know.
                  DR. KENNETH PORTIER:  Dr. Perkins?
                  DR. EDWARD PERKINS:  Yes.  No, I totally agree with your comments on that.  The Wiki is an ad hoc effort pretty much and driven by need.  However, I think that the idea in showing you these was showing you the potential to try to organize information and organize it in a way so that you can get some weight-of-evidence for what potential mechanisms might lead to your outcomes of concern.  
                  If you have specific rulings on how you monitor exposure, whether it's mutagenic or non-mutagenic, then that can be very helpful.  That can be very helpful in determining what kind of exposures you're considering, like that.  Then also, if you start extending it to population level effects, then you can start incorporating things.  It's not meant to be a cure-all really.  And I think you're absolutely right; there's not very much out there.  There's certainly very little on risk assessment and how to use these.  So one of the challenges is how would you use this and how could you include this.  That's why offering it more as an organizer for information, except where you can develop enough data and connectivity or relationships where dose response information to make a quantitative model based on that.  But then still you've got another model, at least, that's going to measure your variables and what you're trying to account for in it.
                  I think there's a number of ways you can use it, and I don't think AOP-Wiki is going to have a lot of that in there.  So, there's still a lot of work to do.
                  DR. DANIEL SCHLENK:  Just to add, I do think the narcosis one is ready for primetime though.  I think you can definitely use that to estimate at least LC50 data off of that, especially when you don't have that.  And it gives you at least something to bounce your measured laboratory data, if there is any, against to see whether you're in the right ballpark or not.  So I think at least that one, and particularly with the compounds that we're going to be dealing with down the pike, those I think will fit into that.
                  In terms of the cancer-based stuff, I agree with that.  I think it's an organizing tool, at this point, in the weight-of-evidence section that I would see it as its benefit.  But I agree it's not a panacea.  We're not there yet, but it is a step forward in terms of how to think about these things.
                  DR. KENNETH PORTIER:  And I also worry about what happens as we move away from generating a million data into a lower file of data, and that's all you have.  You know?  Today you have some animal data.  Five years from now you may not have that animal data, or at least mammal data.  You may have a little fish data or some zebra fish and some in vitro information, and that's all you're going to have, and somehow you're going to have to extrapolate that to a human health outcome.  I see this as the only way to do that, at this point.  And there are a number of panels that have looked at that and made that point.  Dr. Pessah, do you want to jump in here?
                  DR. ISAAC PESSAH:  Yes.  So I think that the pathway approach, especially what it includes, all of the data that's available, whether it's molecular, cellular or in vivo, can be highly informative.  Clearly that's been essentially proven, or it's as close to proof as you can get in genomic analysis.  And when there are discordances between what the pathways tell you and what outcomes at the population organismic level, it tells you quite valuable information.
                  My concern -- and this is a question for Dr. Perkins -- is what happens if you do not include 80 to 90 percent of the available data into the AOP approach, which is really essentially what EPA has done.  I think we've beaten this horse quite a bit.  I don't remember the exact numbers, but of 800 or so available papers in the literature, 24 were used for the assessment.  And this has come up over and over again.  Using the AOP approach in the Wiki, how sensitive would it be to inclusion and exclusion criteria?  And should that be a point that's also made along with the points you wanted to make?
                  DR. EDWARD PERKINS:  I think that's a very good point.  I'm expecting that there will be some criteria to make sure people have actually looked at a sufficient number of papers and what their exclusionary criteria is for why they exclude certain datasets from building their AOPs.  If we want to have these to really support risk assessment in a regulatory context, we have to have some sort of understanding of how hard did they look for data, what kind of data did they search for and where did they search.  It'll be exactly like some of the same issues that were here on exclusionary things, but maybe only finding one paper.  Why did you find that? Did you just not look hard enough?  
                  I think the AOPs all have to get examined under the same context as the data here.  So, it would affect it.  If you only searched for one thing and you exclude other useful information that was contradicting that could increase uncertainty, then you're going to have too much confidence in what you're doing where it's not really deserved.  Does that answer your question?
                  DR. ISAAC PESSAH:  Yes.  Thank you.
                  DR. KENNETH PORTIER:  Dr. Davies, I think we've discussed this.  Dr. Barone, you wanted some final?
                  DR. STAN BARONE, JR.:  Yeah.  I did want to come back to Dr. Pessah's point, and I think he actually hit on something that's an important clarification distinction.  In our systematic review, a lot of our searching efforts, we're looking by chemical name and by CAS number and by outcomes.  And again, a lot of these mechanistic studies are not necessarily informing us about the apical endpoints that we use in the risk-assessment.  So they become put over here into another category and essentially they're excluded.
                  So there's a lot of literature in that realm.  And by using AOPs and informing more systematically -- and again, this is a larger effort.  We tried to do this in the IRIS program, and we saw it was a huge effort and resource intensive.  But including this into the assessment, it's a bigger deal.  So again, having this effort integrated into our risk evaluation process is going to be incremental.  I think it's going to be focused.  I think it has to be targeted because we have limited resources and limited time to complete these evaluations.  They are not 10-year research projects; they are three-year evaluations.
                  So having this modular, getting this fit for purpose and ready to launch is going to be critically important.  But I do agree, incorporating more of the clinical information and other biological information in some context, AOPs is one of them, is going to be important for us as we move forward.
                  DR. DANIEL SCHLENK:  Yeah.  What I had thought of in terms of Dr. Pessah's comments are -- again, this can be used in the problem formulation step in targeting the data that you need, in terms of that exclusion process.  So again, it's a way to set hypotheses of data that maybe you do need in the evaluation document.  So again, just something to think about in terms of the problem formulation step, which I realize we're not a part of.  But it is, I think, a tool that can be used in that capacity.
                  DR. KENNETH PORTIER:  Dr. Cobb?
                  DR. GEORGE COBB:  So one thing about this when you mention the narcosis -- and I've been pounding on the positive of data in the fish model.  Maybe this can be used to ground truth that very specifically in this assessment.  Just go ahead and do that and see if it matches what you've got. 
                  I'll also say to Dr. Willett's point, I think this is a good visualization tool, even the prancing reindeer that we saw at the end there, the last slide.  If you highlighted where you have data and what it looks like is happening, that can give our human brains a visualization of maybe what's happening.  So, that's my thoughts on that.
                  DR. KENNETH PORTIER:  Any additional comments on this?  Dr. Davies, another topic.
                  DR. HOLLY DAVIES:  I think they're smaller topics.  I had just wanted to go back to something that Dr. Twiss mentioned in his original list.  In the human health hazard, there was a nice summary of all the papers.  And then in the next section was the weight-of-evidence discussion where it went through, again, summarized all the papers.  And it just seem like those could be put together because it was the same paper summarized within a little bit of the weight-of-evidence discussion since they are both summaries of the same papers.  
                  And we've talked a lot about the Executive Summary and that seems like a good place for the Federal Plain Language Guidelines, not the rest of the document which obviously is much more technical.  But the Executive Summary, since they are the Federal Plain Language Guidelines.  And something I've mentioned before; it would be really helpful to have next steps.  And obviously you don't know what's going to happen.  You can't give us detailed next steps, but something more generic where you've made a determination of no unreasonable risk, what does that mean?  That means that this chemical is now available.  It's on the inventory.  It's available for people to use in this way, or whatever it is.  
                  I don't know what the legal ease is.  You know what the legal ease is.  You know what that means.  Where you have made a determination that there is unreasonable risk, there's going to be a process after that.  It'll just be helpful to have a short -- obviously we don't know what the agency is going to decide to do with that, other than there's going to be another step.  You're going to go into the next phase, and just the outline that would involve rulemaking, that involve these steps.  Again, you can't tell us what the outcome is.  But I think it's helpful to the public and to everyone else to say that there are next steps coming after that, to the extent that you can.
                  DR. KENNETH PORTIER:  Dr. Kissel?
                  DR. JOHN KISSEL:  I have a general comment.  There was effusive phrase for the overall document.  But as I noted under Question 7, I think the dermal exposure assessment is sub-fabulous, so I don't see how the document as a whole can be ready for primetime.  And I also have a specific comment which gets alluded to in many places here.  But there's language repetitively in here, for instance, on page 260.  This appears at least twice.  There is no unreasonable risk when PPE, APF equals 10 is used.  That's the stock phrase which gets cut and pasted all over.  
                  I think that would be enormously improved if it said, there is no unreasonable risk when PPE is used in a manner that achieves an effective APF of 10, which then acknowledges that one can use PPE and not actually achieve the designed protection.  And it also doesn't imply that EPA is endorsing other people's occupational hygiene practices.
                  DR. KENNETH PORTIER:  Dr. Eastmond?
                  DR. DAVID EASTMOND:  Yes, just a brief thing.  This is following up.  I do think that attempts to trying either your figures or tables to try and clarify it would be helpful.  It's kind of unusual that there isn't an appendix summary of some of the in vitro gene tox data.  There's an error or two in that that I'd probably point out in my comments.  But there wasn't anything for the in vivo data, and that's where you can see the totality of what's there, and that influences the way you look at it.  So I would recommend putting a table in for the in vivo data as well.  And in other areas, looking at that, and possibly doing some sort of simple figure to illustrate how you think things are happening as far as the mechanistic or mode of action.
                  DR. KENNETH PORTIER:  Was somebody on the line?  Dr. Anderson, Dr. Cory-Slechta, Dr. Pessah, did you guys want to add comments?
                  DR. DEBORAH CORY-SLECHTA:  No comment.
                  DR. ISAAC PESSAH:  No comment from me.
                  DR. KENNETH PORTIER:  Any additional comments in the room on this section?  There was something Dr. Davies said that fired off a neuron, especially when I was looking at all the supplemental files and the assessment that goes into the review.  And at some point I thought, at the end of this big table, I'd almost want some kind of comment on the quality of this study for this thing.  But you've got comments on every item, but not an overall comment.  But I'm not sure that would add anything.  I have to think about that, next meeting.  That's the nice thing.  And those of us on the permanent panel have another opportunity to do this next month.  Any additional comments?  Dr. Davies, is your flag still up?  I just didn't know.  Dr. Eastmond, I guess you've made your comment.  Any additional comments?  I look to EPA.  Any additional insight?
                  DR. STAN BARONE, JR.:  No, I don't think so.  I really, really appreciate the robust dialogue of the committee and the input from the public.  And stakeholders, you've commented.  I think you all who were on the previous assessment recognize we've put some significant work into this.  There are still some data gaps.  There are some methodological issues in our prior assessment.  We didn't have eco; we didn't have dermal even considered.  We've gotten some good suggestions and I think that's a significant contribution.  
                  We will be looking forward to getting your report and recommendations.  We very much would like actionable prioritized recommendations for what you think is most important.  We really have to move expeditiously to finalize this risk evaluation because this is a draft.  And so your comments will be incorporated, as well as the public comments, into a revised risk evaluation.  Risk determination is part of that.  And then a response to comments document which will synthesize our responses to both the recommendations and comments of the peer review panel as well as the public.  That's the final piece of the puzzle for the peer review and for the risk evaluation.
                  And then if there are still unreasonable risks -- as was said in my intro slides during the orientation, that triggers an order for risk management for those conditions of use.  And then the risk management piece, that process starts.  And the risk management for those conditions of use will of course include outreach efforts.  It will include different kinds of data collection, information collection and option development, impact assessment and so on.
                  I can't really say anything more about risk management at this point.  We can't really say anything more about what an unreasonable risk call means because it's very, very different.  It's fit for purpose and it's on a case-by-case basis depending upon the condition of use.  There are, of course, multiple option selections or option development that goes into that as activities, and it's really a broad range of considerations.  
                  And again, economics and impacts, other than risk-based criteria, are evaluated during that phase.  That phase, according to TSCA, is supposed to be completed in two years.  So, we have three years to do the risk evaluation.  We have two years to do the actual impact analysis and option selection for risk management.  That's what we have with some degrees of freedom around that, but that's basically what we're required to do under TSCA.
                  DR. KENNETH PORTIER:  Dr. Davies?
                  DR. HOLLY DAVIES:  And one reason I was bringing it up is because the finding of no risk does seem like it's a final -- it seems like when EPA finds that there's no risk, it seems like we do know more.  That was part of the -- the states are preempted, and it's permitted to be used.  As you said, there's more uncertainty in how we go forward with risk management.  It does seem like certainty in the no-risk.
                  DR. STAN BARONE, JR.:  So to Dr. Davies' point, the process is, if there's no unreasonable risk found in the final risk evaluation, that's a reviewable decision.  That's the legal context, and preemption and other factors are also part of that reviewable piece of the decision.  That's about as far as I can go.  I'm not with our attorney's office, but I do know that much.  That is a final reviewable decision, so there are two bifurcates.  No unreasonable risk, reviewable decision.  Unreasonable risk starts the whole risk management process, and then it's not until the end of that that it becomes a reviewable decision.
                  DR. KENNETH PORTIER:  I had one last comment.  And in the risk evaluation doctrine, you have a section called Acknowledgements, and you just acknowledged a team.  And I'd like to see names.  Okay?  I'd like to see how big that team is and who's on the team.  So over time, we can begin to put faces to names.  I know that's not necessary, and you can actually put that as a separate supplemental document.  But no one in the panel have asked for that to understand, you know, is it two people working on this or 20 people?  We suspect it's 20 people, but it would be nice to see names.  And I realize, with a document like this one, which started in 2014 and now it's 2019 and you've got five years' worth of work in there, there's probably a lot of names.  Some did early work; some did late work.  You can pick a cutoff point and say since 2017, these are the people who've worked on this document.  But I think it would help us.  Otherwise, all we see is stand-alone.
                  UNIDENTIFIED SPEAKER:  It's hard.
                  DR. KENNETH PORTIER:  I know it's hard, so it's just a suggestion.
                  DR. STAN BARONE, JR.:  Please, please, please; there is a huge team behind me.
                  DR. KENNETH PORTIER:  They're great.
                  DR. STAN BARONE, JR.:  Figuratively speaking, the risk-assessment division, our chemical control division, our economists help us, and there's a lot of technical input.  It is not just one individual doing these risk evaluations, by all means.  It's an interdisciplinary team.  I totally believe in what you're saying.  Authorship, ownership is an important thing.  Acknowledgment for both career development and permission is another important aspect to this.  We've had a lot of internal discussions about putting names on these documents and there have been differences of opinions, both at the staff level and at the leadership level, and I think we're coalescing around a position.  We take your point under advisement.
                  DR. KENNETH PORTIER:  And with that, I think we're done.  Mr. Kaufman; didn't see your flag there.
                  MR. ALAN KAUFMAN:  Yeah.  Actually I just had one quick question for Dr. Barone.  You had mentioned that a finding of unreasonable risk would trigger the risk management process, but that's a formal process, as I understand it.  If there is a situation where you find no unreasonable risk, that doesn't preclude the agency from making recommendations because some of the underpinning of no unreasonable risk here, for instance, is the proper use of PPE.  In other words, I would assume that you would be able to make recommendations to OSHA to say we recommend that, in the case of this chemical, that appropriate PPE be used, let's say as an example.
                  DR. STAN BARONE, JR.:  I'm going to leave that one alone.
                  UNIDENTIFIED SPEAKER:  It's a sensitive subject, yes.
                  DR. STAN BARONE, JR.:  Well, no.  It's part of the reviewable nature of the decision.  I'll leave that alone.
                  DR. KENNETH PORTIER:  So I guess that brings up the question.  I know you confer with them at the beginning of the process.  The question might be, do you confer with them at the end of the process.  So you've got a final document.  Do you sit down with your federal partners and say, here's what we've concluded, what do you think?  I just wondered.  Or do you at least say, here's what we concluded; go make of it what you want?
                  DR. STAN BARONE, JR.:  Actually that's a good question.  And in fact, that is part of the process.  So after we revise the risk evaluation just before we release it to the public, we have a consultation within the federal family.  We have a consultation internally within the EPA offices, and then we have a consultation within the federal family with the other agencies, and it's fairly broad.  And they review the draft evaluation and the supporting documentation.  
                  When we get to the penultimate final risk evaluation, the response to comments and the risk evaluation, we'll also take that to the same group, and we'll be talking with them about our conclusions and whether we address the peer review recommendations.  And they're the checkers.  Do we address the peer review recommendations?  Because we don't come back to you to say did we answer all of your peer review recommendations.  And believe me, for those of you that have been involved in Government service that are familiar with this process, it's an important check and balance.  People look.  So that is part of the process before we go final.  It affects the timeline.  We have to incorporate that into our timelines as well.
                  DR. KENNETH PORTIER:  With that, I think I'm going to close the formal part of the meeting and turn it over to Ms. Gibson, our DFO, for final comments and discussion of how we're going to wrap this up into our report.
                  MS. TAMUE GIBSON:  Thank you, Dr. Portier.  First I would like to thank members of the public as well as the committee.  Thank you so much for your participation and your robust discussions.  This was very successful, and I greatly appreciate your time and efforts here.  I would like to thank our Chair, Dr. Portier.  Thank you so much for keeping us on track and allowing us some time to come together and break for some who have long days, so thank you again for your service.  At this point, I will close the public portion of the meeting. 
                  UNIDENTIFIED SPEAKER:  Are we going to meet to follow-up?
                  MS. TAMUE GIBSON:  Well, yes.  We can meet for several minutes just to follow-up to have some ending comments and directives, if you would, just for a couple of minutes.  And I know that some individuals are on some timeframes, so we will officially adjourn this meeting.  Thank you.
                              (MEETING ADJOURNED)
