 Good morning to all and welcome to this public meeting of the United States Consumer Product Safety Commission We have two Commission agenda items today a briefing on the National Academies of Sciences Engineering and Medicine's report Entitled quote a class approach to hazard assessment of our gonna halogen flame retardants and quote and our second agenda item today is a CPSC staff briefing and General hazard scenarios associated with off-highway Vehicle fires burns and debris penetration as well as any recommendation CPSC staff may have for the applicable voluntary standards body We will begin this morning our public hearing with a briefing on the National Academies Sciences Engineering and Medicines report again Entitled a class approach to hazard assessment of our gonna halogen flame retardants Briefing us this morning, and we extend a very warm welcome To dr. David Dorman who served as chair of the committee to develop a scoping plan to assess the hazards of our gonna halogen flame retardants Dr. Dorman is a professor of toxicology in the Department of Molecular Biosciences of North Carolina State University Dr. Dorman received his undergraduate training in chemistry at the University of San Diego and a Doctor of veterinary medicine from Colorado State University He completed a combined PhD in veterinary Toxicology residency program at the University of Illinois at Urbana, Champaign He completed a postdoctoral fellowship in toxicology at the Chemical Industry Institute of toxicology He remained there as a staff scientist until joining NC State University Veterinary school as associate dean for research in graduate studies Dr. Dorman is a diplomat of both the American Board of veterinary toxicology and the American Board of toxicology Dr. Dorman's research Interest includes neuro toxicology nasal toxicology Pharmaco, excuse me, pharmacokinetics and cognition and olfaction in animals He is an elected fellow of the Academy of Toxological Sciences in the American Association for the Advancement of Sciences He has served on advisory boards for the US Navy the National Aeronautics and Space Administration the US Department of Agriculture and the National Toxicology program where he served as a member of the NTP's Board of Scientific Counselors He has chaired or served on multiple National Academy's committees, and he has a National Associate of the National Academies His current National Academy service includes chairing the committee on toxicology and the committee to develop a scoping plan to assess hazards over gyneohelogen flame retardants and he is serving as a member of the board on environmental studies than toxicology his past National Academy services including includes chairing the committee on endocrine related low-dose Toxicity the committee on predictive toxicology approaches for military assessments of acute exposures the committee on design and evaluation of safer chemical Substitutions and the committee on potential health risks from recurrent lead Exposure to DOD firing range personnel and I should say and exposure to DOD firing range personnel He recently served on the international agency for research on cancer Monographs volume one two three group that evaluated the carcinogenesis Excuse me the carcinogenesis city of some nitrobenzene and other industrial chem chemicals Just reading that Has exhausted me So good morning and just welcome to you Dr. Dorman were so grateful that you came here this morning a company dr. Dorman this morning is dr. Ellen mantis and a s project manager Miss Susan Martel senior program officer for the board on the environmental studies and toxicology and Dr. Gregory Sims executive director of the division of earth and life sciences Again, I want to extend a warm welcome to all of you and thank you for being here this morning and for all of the Tremendous work that you all did on this report presented to the commission Following dr. Dorman's presentation We return to questions from the commission each commissioner will have ten minutes each for questions and we can go multiple rounds if necessary And with that dr. Dorman, I would ask you to begin. Thank you Okay, thank you very much for that very kind introduction. I always joke with people It's sad to hear you're obituary while you're still alive So I just but thank you and also on behalf of the committee and national academies Thank you for the opportunity for me to provide a briefing on behalf of the committee and so What I'd like to do is we did prepare some slides as briefing materials and feel free to interrupt me at any time I know you probably have some set rules But if there's a point of clarification feel free to interrupt me. I'm I'm quite used to that and So just to begin one thing I do want to just remind everyone Especially maybe the public that might be watching this is the role of the national academies as an advisory group And so one of the things to just consider is part of the reason why national academy reports play such a critical role in What different federal agencies do is that they bring together a wide range of experts But then these types of reports also go through one of the more robust peer review processes So by the time a report like this one is actually published There's been a lot of hands and eyes that have looked at that technical material and the conclusions that the committee draws So hopefully I can advance the slides. So this is an abbreviated statement of tasks So consensus study reports like the one that I'll be talking about today Generally have a statement of task that outlines the scope of the project of national academy Scope of the project that the committee is charged with this is an abbreviation Version of the statement of task, but it's the highlights. So one of the first things we were asked to do Me dr. Dorman, I'm going to ask you just to move your microphone just a little bit closer I'm hearing some feedback there was feedback here Yes, and maybe we can get rock to check it out. We'll try it. Let's try that again Okay, and so there is a statement of tasks So one of the first things our committee tried to do is to survey the literature to find out what chemicals actually constitute polyorganohelogen flame retardants We also wanted to provide CPSC with one or more approaches in which a hazard assessment for a class of chemicals Like the flame retardants could be performed And then we wanted to give you a scoping plan that staff at CPSC could follow to try to actually implement the recommendations that are found in our report So the committee just like other consensus committees that I've served on has a broad range of representation There are members of the committee who were drawn from academia industry and governmental laboratories They also represent a broad range of scientific expertise from my own expertise in veterinary toxicology We had physicians with backgrounds in public health and epidemiology. We had several physical chemists We had toxicologists We had informaticists, so we have a wide range of talent that's on the committee That then is shepherded together to try to come up with the final work product Then in addition to the committee members, we were also had two very excellent national academy staff members Susan martel and alamantis that also helped with the generation of the report and Conduct I'll also say that at different times our committee had input from CPSC Early in the process where we received briefings from your staff that was very helpful for us As we started to consider how we would tackle this project and be able to be responsive to the statement of task So one of the questions that we were faced with was whether or not the Large group of chemicals and eventually we identified approximately 161 chemicals that would fit into this class Could actually be treated as a single class And what you'll find is during the course of the briefing and as you look at the report we reached a very Conclusion that said that as a single class is a single class of 161 chemicals That we really couldn't deal with them as a single class for hazard assessments So we would have to break them down into subclasses So although we broke them down into subclasses Doesn't mean that we ended up avoiding a class based approach What it just meant was rather than work with a large pool of chemicals We broke it down into smaller pools, but a class based approach still applies And so really as we go through this briefing The traditional approaches used oftentimes by organizations is to do a chemical by chemical hazard assessment So we do chemical a come up with a hazard assessment chemical b keep going In a class approach what you're trying to do is find a group of chemicals That share different characteristics in which once you start to study Individual chemicals within that class you extend that knowledge base to the entire class And cpsc actually has experience with that with the phthalates in which a class approach was used So that in itself is not necessarily novel So in the hazardous hazard assessment scoping plan that the committee came up with we have three main elements The first task is to actually identify and define what a class or subclasses are And i'll take us through each of these steps The next step because we're doing a hazard assessment and most traditional hazard assessments rely heavily on published literature The next step in the process is to actually conduct a survey And then a scoping and the survey of the literature And then the third step is to actually do the formal hazard assessment based upon the evidence that you find from Surveying the literature. So those are the three steps that get done in series So in plan one the first thing we do is try to identify the viability of a class approach So the committee asked ourselves in essence, can we take Flame retardants and treat them as a single class. So one of the first questions we were faced with Is what are the flame retardants? So there's no Really authoritative list that says these are the flame retardants. So what we did Was in our first exercise is we try to identify known flame retardants that are out in commerce And so what we did was we basically looked at the literature We drew from a number of different sources and we compiled a list of chemicals Then what we did is we tried to identify were they all individual chemicals? Could they be mixtures? Could some of the chemicals actually be duplicates? Because sometimes you have chemicals that are the same chemical structure or named more than one way So what we basically did was we annotated a list of chemicals that we refer to as The seed set of chemicals that we've identified as those chemicals that could be used as flame retardants And what I kind of want to do is I'm going to step away from the chemistry for a moment And just talk for a second or two about taxonomy So I think a lot of folks are used to the idea of animal classes. So being a veterinarian I'm going to try to really kind of simplify what we did. Okay So as a veterinarian from a taxonomy perspective I might ask myself What are cows? So what are cattle? So there could be features of cattle that we would want to identify them as a class So what we would do is identify what do we think are classes of cattle? And we would start to identify what are those characteristics that that group of species of animals that taxonomic group of cattle share Then what we'd want to know is is there something distinctive about being a cattle? Is there something about being a cow that distinguishes it from being a giraffe or being an elephant or another animal in a taxonomy approach? In order to be able to do that comparison what we have to do is compare it So what we did as a committee is we took our seed set And then what we did was we said All of these chemicals share a halogen. They contain bromine or chlorine or fluorine And then we said let's grab all the chemicals all the animals that share those features And then what we did was in the next step we then said But that's not very that's a crude way. So let's refine it and ask We're going to compare our seed set to other critters basically that share different properties So those are our analogs and then what we did was analyzed Is there a nature to being a cow so to speak distinct from all the other animals that we identified? And what this analysis what we did Was what we found was the seed set when we compared to analog structures. We couldn't clearly separate them So in other words, they shared features that don't allow us to say This is a unique set of chemicals that have certain properties that allow us to distinguish it Taxonomically from another group of chemicals. So that's in essence what this slide is trying to explain to you So it's a very technical terms and we used a lot of technical approaches But at its heart we grabbed a group of chemicals Expanded that list and asked could we separate them out? So once we came to that conclusion that brought it brought us to the conclusion That these are not a unique group of 161 chemicals that could be discriminated from other chemicals That share other similar features. So in order to do the next step in the analysis What we did was broke down The ones used in commerce into 14 different subclasses. So we're still doing a class approach But I I found when I've talked about this report to colleagues That kind of taxonomic approach hopefully helps you get through that that portion of the report But what we did was we broke the larger set of 161 chemicals down into 14 subclasses The committee points out that the way they get broken down into subclasses different groups might actually come up with a slightly different subclass approach Because we constrained our analysis in a couple ways. So what we said was We don't want to have a subclass with just one member So we didn't want to divide it and keep dividing and dividing and dividing it into smaller and smaller subclasses Because ultimately you'd end up with subclasses that really were not informing So the 14 subclasses every subclass had A small group of chemicals that shared certain physical chemical properties And or biological properties. So that's how the subclasses were created So again having it be subclasses what we I think what the committee would encourage cpsc to do is not necessarily go through that same exercise I think from a cpsc perspective the 14 subclasses we created Are a viable place to start with a hazard assessment for different subclasses of chemicals And that we would also encourage cpsc to not more narrowly define them because again You end up having where you might have a subclass with say two In a lot of ways traditional approaches for subclasses of one or two. You just would default back to a traditional approach Okay, so that's in step one. So we've completed step one We've created our subclasses in step two What we would start to do is ask what does the literature what kind of data is out there And so what we wanted to do is survey the literature So what we did was we actually did a number of computer programs now exist That allow you to look at the literature that's published And be able to understand is there data in epa databases? Are there been other types of data available go? I apologize There is some concern that the hearing isn't being broadcasted online yet And we just want to make sure before you go any further with your testimony Good luck You can try Because people right now they're halfway through trying to figure out why i'm talking about cows and that could be disturbing for them Probably right Okay, we've been assured that it's broadcasting. So you may proceed. Thank you so much No, no worries And so the second big step is to try to survey the literature and so in this initial survey You're just trying to get a sense of for a subclass how much data is actually available We're not trying to go to individual studies at this point We're not trying to interrogate all the data. We're just trying to landscape it Like so is there a lot of data on certain chemicals or is there like certain classes with virtually No toxicology data that's available in a public domain And so when we're surveying the literature, you're again, you're just trying to get your Kind of grips to what the data inputs might eventually be And in our language within the committee report Surveying the literature goes beyond just peer reviewed publications So it may be technical reports you have access to it's just you're just trying to grab a lot of different types of data You'll also notice from this slide that we're looking at different what we refer to as evidence streams So some of the evidence might come from human studies Some of the evidence could come from traditional animal studies or other types of animal studies Some might come from in vitro studies mechanistic studies So again, you're trying to grab all of this data at this point to just survey what's out there So you then would want to map the literature and we we kind of Didn't want to present some of the literature maps, but if you turn within the report you can quickly see Several figures that kind of show you what that mapping exercise will look like So for example, I'll draw your attention to page 26 in the report in which Different horizontal lines represent Different groups of chemicals and then the vertical lines represent databases And what you can quickly kind of appreciate is that just visually looking at this data There's some classes where there's not a lot of bars Not having bars means there's not a lot of data there Okay, other chemicals what you might find is there may be one or two chemicals within a subclass that's actually data rich And then others don't have any data. So again, you're you're why the survey becomes important is because It's now going to have down the road some implications about How do you kind of do this from a policy perspective because the science won't be robust for every chemical So you're going to have to make some choices Will we use say a data rich chemical? To draw inferences for a data poor chemical and we'll talk more about those scenarios here in a minute But that's what this survey and map is actually trying to do in the report So then in step three now you're going to do your deep dive So what cpsc might choose to do is pick one or more Of the different subclasses for a variety of different reasons. They may be the most important from a commerce perspective However, those decisions are made but once you've made that decision to look at the deep dive for data Now what you're going to do is actually start to do a more rigorous Look at the literature. So what you're going to do is actually search it Find the literature that's relevant to what you're trying to do the hazard assessment Meeting the regulations that you're working off of and then you'll extract that data So you'll pull the data out of the primary literature Another literature sources And then you'll have that data available so that you can start to make hazard assessment decisions Again, you're trying to do this as a class So you're simultaneous looking for data for say 10 members of a chemical class So the survey again is just a broad understanding the literature search the deep dive And there's a variety of different ways that a deep dive literature literature search could be performed So for example other academy reports have mentioned things like narrative reviews or systematic reviews So there's a growing literature base that says here's how you systematically look for the literature. We didn't say Tell the cpsc that there was one Method that had to be used we were just recognizing that at this point though What ideally you would want to do is have a search That others could replicate so if I was looking at how your staff did the search I could understand what search terms did they use what databases did they look at so that I would have an idea of What was the scope of the actual literature search that was performed? so in a literature search That that can be very time consuming can be you know Actually quite a quite an effort in and among itself and extracting lots of data So sometimes what you would want to do is Maybe make some decisions ahead of time that you're not going to try to search for every outcome that will focus on say cancer Because the survey indicates cancer may be the driver for that subclass where it may be cancer and reproductive effects So again, you're trying to have Different steps in the process you're having to make some expert opinions and make some expert judgments Then not just go as you know and use everything One of the other things we're recommending and we'll talk about here in a few minutes is that Because there's a paucity of traditional data for a number of the classes and individual chemicals You may need to rely increasingly on In vitro studies that have been performed for different chemicals So you may have to make the move from traditional animal based or human epidemiological based studies to doing decision making Based on in vitro or non-traditional rodent or non-traditional animal tox data So again, that's a has policy implications But that could be an important thing you'd want to do at this step If you're not going to rely on that type of data So you make the policy decision We will not use this type of data to inform the hazard assessment then Quite honestly, you wouldn't go looking for it But today most of the guidance coming from national academies to different organizations is That type of data is going to become increasingly available traditional studies are less likely to be used So most organizations are going to have to come with groups with how they'll use those types of data In the regulatory process and we're encouraging CPSC To kind of become a leader in that in that area So you extract the data So again, this can be time intensive and so you collect the data that you need Chemical information biology information What were the outcomes of interest? And then finally what you have for a single subclass is you've got you've brought together all the data You think is relevant for you to do the hazard assessment At this step you're doing the hazard assessment for a class of chemicals not an individual chemical And so what you're going to find is that we identified three possible outcomes for that class based hazard assessment You might decide that the data points to the class being hazardous The data might point to the class being non-hazardous And the data might be inconsistent what we refer to as discordant data in which it's not black or white to say it is Or is not hazardous What we found just to forewarn you what we found in our two case examples Was discordant data and if I had to predict that won't be a rare outcome It's not always going to be black and white this subclass non-hazardous this subclass hazardous So we'll spend a few moments talking about how do we handle discordant data because that's One of the outcomes that we identified in our scoping plan So we tried to identify a few scenarios that your staff may be grappling with At the hazard assessment plan and it's important for me to also point out The committee was not asked to actually perform a formal hazard assessment So please be aware of that nowhere in our scoping plan Was national academies nor the committee asked to do a formal hazard assessment or risk assessment So what we're trying to do in our report is illustrate Likely outcomes. They're all database outcomes. They're all plausible outcomes, but it's not a formal hazard assessment So I just want to make sure everyone again is reminded of that fact So we were illustrating different scenarios And so one scenario we mentioned in the report for example is you might have a really data-rich subclass We've got say 10 members of that subclass and five of them have been well studied And a couple others have got some data And all of the data for all of the members of that that subclass are all pointing that there's a signal a toxicity hazard from that subclass They're all carcinogens So let's say hypothetically of the 10 chemicals in the subclass Five are known carcinogens And five have say data for mutagenicity that would make you suspect that they might also be carcinogens In that scenario, that's a relatively straightforward scenario You could treat that group as a hazardous subgroup and move on and make a determination And in a lot of ways that's kind of what you did with the phthalates You kind of found a common mode of action common chemical structure And you said this is a group of chemicals that could be treated the chap for phthalates kind of drew that conclusion The other scenario which I've already kind of alluded to is that for some subclasses you may not have a lot of data at all So you've got a few choices So one of the important things the committee wants to make sure That we communicate clearly Is the absence of data should not be considered a subclass is safe So we don't want to get into an what we would refer to as an unfortunate situation where we do an unfortunate substitution Where we might take a compound that's somewhat hazardous Replace it with one where there was really no data at all. So if there's no data, we can't consider that to be non-hazardous So that's an important caveat So you may need to generate data And so now here's where a few decisions have to come in So do we generate data for traditional data for all the members of the subclass that would be very time consuming and expensive Do we make some policy decisions that say we'll study a couple of animals or a couple of chemicals And then get traditional tox data for one or two members And then what we'll do is extend that knowledge to the other members of the subclass Or will we say survey using in vitro approaches that are less expensive So there are several different scenarios here Another possibility would be to say Maybe we should just take that subclass and merge it with another subclass So that that would be another viable option We might also use that broader set of chemicals we identified as analogs Maybe there's a closely matched set of analogs that we can now use to inform it But these are largely driven by some policy decisions as opposed to scientifically based decisions They can be informed by the science But they're largely different options that have some policy implications in them So how would we generate data again consistent with other national academy reports? We would suggest a tiered approach to testing that from a cost effectiveness basis You would want to start with lowest kind of systems cheaper systems in vitro test systems non-traditional animal models like zebrafish and others And then work your way up into more traditional testing as you deemed necessary So the third scenario is where we have Coherent data on a few members and so again you could make a decision that says We're going to use say the most data rich member of that group to drive the hazard assessment for the rest of the members And that that's a policy decision That in a essence what we refer to in the report is those would be our anchor chemicals within a subclass They're well studied the other ones seem to be following in the same guys And so what we can do is we would extend that knowledge to the rest of that subclass And then the other option you would have is again There may be individual chemicals of interest because they're heavily used in commerce or other reasons Where you might generate additional TOCS data for one or more members additional members of the subclass But here you have at least some data and the data seems to be coherent And the final scenario which rarely was the one I think in a lot of ways Our committee grappled with the longest because it's the one we identified in our two case examples Is you have discordant data so you have positive results for one chemical In a subclass negative results for the same endpoint for another member of the subclass So that reaches kind of the broadest suite of different options that you might need to have to consider So for example, what you might do is just say Conservatively what we'll do is we'll treat the most toxic one in this subclass We recognize they're not all behaving identically, but we'll recognize the most toxic member of the subclass And we'll base a hazard assessment on that member You might also decide That what you're going to do is you might have to actually parse that subclass into a few different even smaller subclasses Because it may be that there's a common metabolite that some produce and some don't produce So that's guided by expert judgment about why is it that half the group Seems to be behaving this way half the group is not behaving the same as the other fellows within that subclass And then what you might do is some additional targeted analyses to try to elucidate why there's discordant data And then of course you've always got the option of trying to generate new data Okay, so as again in the two case examples we had we were faced with discordant data But we tried to illustrate a couple of ways that cpsc might try to move forward in managing the discordant data So a lot of what i'm mentioning is pretty time and resource intensive And it requires some expertise that crosses a lot of different disciplines And but in a way of time and resource intensiveness What we don't really recommend is going back to the traditional one chemical at a time approach It's just too time consuming too resource intensive So coming up with policy decisions to help guide A class approach for your organization Will be expedite the process it'll protect the public And it'll cut down costs and it'll increase efficiency But there are those are driven by some key policy decisions and I Can't stress that enough And so you'll need to make decisions about would you accept non-traditional tox data? That's a challenge not just for you, but for other organizations I've chaired several committees now in which those have been recommendations to do and other organizations So different federal partners you have are moving in that direction And part of the reason for that is again cost efficiency And we've got a lot of chemicals out there with very very little toxicology data So it's one way for us to get some confidence or make some inferences And then you'll need to make some decisions about what data Like minimal data set would we need for a class approach because again the traditional approaches We've always used for one chemical at a time. We kind of know the data we need But now in a class approach To have it all break down to just every individual chemical we've got to do the exact same thing It's just isn't a viable approach anymore. Is your material well aware? So some concluding remarks that I want to just kind of bring back to your attention So again, thank you for the opportunity to give you a briefing. Hopefully I haven't You know disturbed you too much I I'm actually quite proud of the report on behalf of my committee members and really thank National Academy staff. I think we tried to be as responsive as we could to your statement of task We you know, I think in a lot of ways what we did was we tried to give you a blueprint The scoping plan of what you might try to do I know in discussions with my colleagues at academies academies is here to help you In any way they can in the next steps And I look forward to meeting with your staff here at lunchtime and throughout the rest of the day But I think I'll be more than happy to take questions and try to To maybe clarify points within the report that May not be as clear as I would hope they'd be at the end of my briefing, but thanks again for the opportunity Thank you very much dr. Dorman, we will now begin the commission's round of questions each commissioner will have 10 minutes To begin to have questions and ask questions of you and the panel And I will begin the round of questions I wanted to just Go to the very basic premise in terms of We talked about the traditional approach chemical by chemical and then I think essentially what you What was handed to us is you can't do the entire class of organohalagen flame retardants So break it down into subclasses Is there anything that's lost or does it become less Um I'll say accurate or less informative by forfitting The the individual chemical by chemical versus I think probably where we are somewhere in the middle in terms of what you've suggested to us Sort of a compromise between this approach and the entire class approach Is there do we lose anything do we gain anything could just speak to that a little bit and why this why that was chosen right so The the breaking it up into the subclasses was science-based So we asked we you know the committee really did try to treat it as a single group But we couldn't do it from a biological perspective or a chemical perspective. It's just too diverse a group So from a science defensible position breaking them down Into the different groups is a much more defensible scientific approach to do and it's more in line with what Others have thought about for a class approach for toxicology what you probably lose is Time and resources in some ways If you had them as a single class as a mega class that you could just make some policy decisions about Then that would obviate the need for breaking them down into the subclasses But we didn't see a non You know those that would be a non science-based approach that more policy to do that but from a science-based perspective Breaking it down into the subclasses more robust more defensible Thank you. Um, I want to go now to um so On one of your slides the caveat of class approach took uh, ofr's and this is just Probably what you just explained a little bit maybe you could clarify So this the committee concluded that best approach is to define subclasses as broadly as is feasible Is that can you just speak to kind of what that term means? So for us what that meant was again, we we when we were parsing these large class in the subclasses We constrained our thinking so we said we don't want a subclass with two chemicals So we want subclasses large enough to make a class approach feasible And so you could parse it Lots and lots in different ways until you eventually come up with in essence 161 classes each represented by an individual chemical So our thinking was constrained to say let's have a large enough data set that you could do it as a subclass but not parse it Till the end of time so to speak Thank you But again the parsing is based on They're all are sharing some chemical features They're also sharing some biological activity or predicted biological activity So it's not it's not a random thing that these subclasses were created the subclasses were based on The evidence we have about their chemistry physical chemical properties and predicted by a biological responses Thank you. I want to go to um the No relevant data piece and then So You gave us three scenarios If there's so if you're saying there's a class of chemicals and there's no relevant data Relevant to that class. So you're giving these options. Is am I understanding that correctly? Okay And just because and what I think you made very clear is just because there is no relevant data That does not mean there's not a hazard associated with the Okay So for option two or three I was curious to reclassify the subclass so that data poor members are distributed in other rich data subclasses Can you maybe elaborate on that a little bit? What does that mean? And again, do you lose anything in Moving it to a different subclass, right? So we we have a cautionary tale of again, like you could get a different group of scientists together that might have subclassifications in a slightly different way So there's not There's more than one way you could create your subclasses What we don't think is a viable thing to do is every time you hit a headache Reclassify because then you're just Changing the wheel and kind of kicking the can down the road so to speak for how you're going to do a class based approach So what we're recommending is that there may be times when you have to consider reclassification But they should be done judiciously And what's that's kind of the analogy going back to my simple analogy? It's close enough to a cow that we're going to act like it's a cow type of analogy So we're going to kind of although it's ideally should be here We're going to move it into another subclass and like next closest neighbor so to speak With with those chemical properties or biological activity But what we would not encourage you to do is just keep doing that Endlessly because that just will delay your analyses and we'll just kind of confuse and Not have the transparency that we need Thank you And then my last question for this round. I just want to talk a little bit about the discordant data And so the options that you gave us and I thought and I just maybe you could elaborate a little about an option for Generate new data that could increase clarity and the scientific base for the decision. Can you To non-scientist. Um, can you elaborate? What does it mean to generate you just Because I thought An exhaustive well relatively exhaustive search had been done. So how would you generate new data? So for example, I'll use a relatively straightforward Point where we may have say a class of 10 chemicals And so chemical a has been identified as a carcinogen Chemical b in that subclass Has gone through a cancer bio assay And has come out clean. It's negative. So it does not produce cancer. So that's an example of discordant data They share certain chemical properties. So one thing you might ask is What was the cancer driver say a mutagenic effect? Did it affect dna and could we maybe look at other members of the class where mutagenicity data wasn't collected Now we generate new data If they're positive we would treat them with the cancer positive one And it may be that the non cancer one we could explain for some other reason why it was negative in the bio assay Maybe it got metabolized really rapidly. So we might need some pharmacokinetic studies So that's we're generating new data where you've got discordant data. You could ask yourself wait a second There's like three members of the class are all acting the same. These two are not Let's have some targeted studies that might explain that discrepancy You could also come back to the different options and just say Let's not generate new data. We're just going to make a policy decision that says We're going to use those most toxic ones For example, we'll drive the hazard assessment for the rest of the members of that class. So that would be another option Thank you very much commissioner edler Let me begin by thanking you and the committee for the Excellent work that you've done. I'm not a scientist But there were great portions of it that I could understand and some portions that I couldn't understand But I do thank you and your colleagues have a reason to be very proud of the study that you've done And I can't thank you enough for doing that So I wanted to start with one question We had a couple of days of hearings on organohalagen hazards and one of the questions I kept asking I realized you didn't do a hazard assessment But in those instances where you found it to be data rich Did you find any of the organohalogens that you looked at and said This is not a problem. This is not going to ever present a public health hazard or Um Anything that in effect would be exonerating one of these subclasses Right. So again, I'll come back to we had a scoping plan not a formal one So the committee was never charged with that. I know I understand just asking if you or your colleagues I'd have to say it'd be almost inappropriate for us to try to answer it and that guys because we didn't do that type of analysis I tried And I thank you and uh, also I I also Want to tell you how impressed I am with your understanding of the problems that we face as a regulatory agency We're not a pure science research agency. We're a policy driven agency where we're supposed to take the scientific judgments that you make and then apply them to policy and I'm so glad that you understood that So let me share one of the miseries that I think we face which is we're a very very poor and tiny agency And I will only repeat this example one or more times But when you look to see fda's request for its budget last year They asked for an increase The increase that they asked for in their budget was about five to six times our total budget And so I see the figures that you've given us in eight and a half years and 14 and a half million dollars That should be easily Uh addressed with other agencies for us. I'm not saying we couldn't do it But it would be a strain and so I guess one question I have is Do you have any suggestions to us if we were to undertake this or something similar? Whether there are other bodies that could provide us or assist us in funding something like this? For example, epa if you've thought about that, but we did So I'll draw your attention for example to one of our committee members was john booker who I saw his name Yes, associate director of the ntp at that time and so For example, one of the things you can do is partner with agencies like ntp to develop The in vitro data that the committee talks about that's part of why we highlighted in some ways the in vitro data because There are federal programs going on in which chemicals get nominated and testing gets done So there's a nomination process that you could work with ntp and other federal agencies to do that The I think the key thing though to really answer your question is You could be Any of the federal agencies with all the money in the world to be honest and you were trying to then do a chemical by chemical approach traditional You'd probably exhaust your resources trying to do that So what we tried to be was pragmatic and giving you what I'll call policy outs so Make a policy decision on say precautionary principle will do this type of thing That's option one say of just treating extend the most conservative estimate to for hazard to the entire class That those policy decisions in essence shortcut the need for all of the data And they may still remain public health protective And so we recognize that that's part of why the committee actually put as much time and effort into trying to come up With policy choices for you was trying to be pragmatic Just understanding you're going to be faced with these questions. And so what are those shortcuts that you might be able to take? Well, two quick responses. First of all, when you say precautionary principle, I could see Hair standing on end across the room because that's one of those controversial approaches using that strictly as an example No, I understand that but you did you did mention mr. Booker from Nihs and one of the things I did yesterday just out of curiosity. I looked to see whether Organo halogen flame retardants are A project with NTP apparently they've never been nominated So I guess my question would it be appropriate for us just in the interest of expanding the partners to nominate ofr's to NTP I mean, so This is kind of answering as dave dorman not as the committee member But I'll answer and the guy is that I was also on the board of scientific counselors So I we the board of scientific counselors for NTP Oftentimes gave feedback to NTP about which chemicals to pursue So again, what you might find is that there's say a commercially viable or important subclass that has limited tox data That would be a very appropriate thing to do would be to nominate that chemical Thank you very much And I just want to go back to a point you made because it was also one of the points of contention In our earlier hearings you have unabashedly aggressively endorsed the idea of a class approach And because otherwise we get into this doctrine known as regret regrettable Substitution so could I ask you just to speak a little bit more about why it's so critical that we adopt This class approach that you're recommending So the class approach has a couple of advantages One is once you start to make hazard calls for a subclass If I say a commercial manufacturer of a flame retardant now bring into the marketplace A chemical that's very very closely similar to the ones you've already made a hazard determination for Now your staff would be asking does that next member Is it doesn't belong in that subclass Whereas the current way we again, we would just be thinking about them as an individual chemical by individual chemical And again the cpsc to your credit when your chap looked at phthalates That was part of the drivers was trying to understand where this is different The landscape is a little bit different than phthalates Is that the phthalates shared a common mode of action? And it was really kind of a data rich set of chemicals But still had its own challenges of trying to do a class approach So that's one way that the public gets protected is you can start to ask the question is this next newest member Similar enough to ones we've already made a determination of that we would say yes or no that it should enter commerce So one of the things that instigated our attention to this was a petition That asked us to address organo halogen flame retardants in four product categories Which I won't run through but i'm pretty sure you know what those categories are But when I was reading your report it dawned on me that you're not necessarily Limiting your Recommendations and your thoughts to just these four product categories. You're addressing organo halogen flame retardants is a broader class and therefore would it be appropriate to say that Any study that we would do even if it were to be Constrained or guided by the petition would be in effect addressing hazards across the board with respect to organo halogens Correct. I mean we took a broader approach I'll also say that this may give you further guidance for other chemicals that are beyond the organo halogens I mean we we were also very cognizant that we didn't want to create a scoping plan That would be one off only for organo halogens So although this plan Deals with your that that current problem You may be faced with similar concerns in the future with other classes of chemicals And the scoping plan with some modification could be applied then more broadly Also, I would say that what your uh committee did Excellently as far as I'm concerned was a scientific analysis I'm guessing you didn't have any lawyers on the committee Is that fair to say appendix a provides the biographies and to my knowledge we did not yeah And and it's not a criticism, but I did note that uh when you were in your report You were talking about a proper approach to this and one of which is you start with the finding of toxicity And then you move to quantitative risk assessment And the only point I would make is that We have we actually have a set of guidelines about addressing things like carcinogenicity, mutagenicity, teratogenicity And quantitative risk assessment is certainly one of the tools, but it's not something that's mandated By our guidelines or by our statute and so my question is In terms of the scoping Did your scoping necessarily include The quantitative risk assessment was at a major part of the cost No, our our scoping plan Recognizes there's a follow-on step, but our scoping plan really is at the hazard assessment stage Where the next step in the process the risk characterization could also come in though earlier is when you're scoping the project In other words, you may start to inform which subclasses you might want to study next might be informed on for example Exposure information that you have so that they're they're distinct But some information may actually help guide some of your policy decisions Some of that information may guide how you select a subclass for hazard identification My time is expiring, but I want to thank you for a very very clear explanation Such that even I could understand it. So thank you again. Thanks for the opportunity commissioner k Thank you, madam chair and thank you to the panel for all the work It really is a tremendous public service in a much needed area. So we're deeply appreciative of what you've done So a few questions about how we would potentially go forward and what work might already be going on out there Are there other mentioned ntp may not be doing ofr research But is there other ongoing research that the committee is aware of that might help provide some of this new data? So when we so for example, I'll talk a little bit about just the zebrafish because we highlighted that as one example And again, I just want to illustrate that that was an illustration We didn't say you have to use zebrafish. We just wanted to illustrate that as one data stream And so what we're we didn't try to scope all of the literature But what we are is aware of that there are grants out there that are looking at comparatives So you're going to be in a situation in which in the next five ten years You'll have more data from academic and other labs where they're doing comparisons somewhere to the studies We've already reported. I mean we were aware that some of those studies were on but again That wasn't a major focus for us We were trying to look at what data is available now In the arena of in vitro data Kind of non-traditional approaches just knowing where the field of toxicology is going and again speaking as Dave Dorman Not on behalf of the committee That that type of data is going to be generated through You know in academic labs largely over the next decade or so So positioning yourself to be able to deal with that data and apply it to hazard assessment Is what would be helpful for you as an organization? Okay, so it's good to know it sounds like even if we're not driving Additional research there is additional research that is going on it might not be enough But it at least continues to some extent and again, it's largely you know because there is human exposures So NIH is funding studies. There's human epidemiological studies. There's biomonitoring studies So we allude to some of that we point to but some of those studies are long term Where some of those cohorts of people will be reevaluated over time. So they'll be increasing data The the challenger again faced with is we tend to study what we know And so there's a lot of the organo-halogens where there's not a great deal of data So those may continue to have a Pocity of data even in the near future. I see so that might be a role for us to Maybe encourage some of that research correct or again like make decisions Let's say we'll live with our data rich ones and have to extend it to the to the others And why does some just had a curiosity? Why does some OFRs end up driving data and some don't? Could be because of human I mean again we as a committee we didn't try to look at that background But again based on my own experience and dealing with a variety of different chemicals a lot of that comes down to It's broad use in commerce. So some of the chemicals are found in commerce more than others And part of it also is they have positive signals So toxicology you tend to publish positive results, especially academic labs So once you find a positive You tend to get more positive studies because other labs are trying to replicate your results and things like that There's a little bit of that that goes on as well And you mentioned new approach to methodologies and the encouragement that we are open to that if we move forward from a time and resource At some point how novel would those approaches be relative to what epa or ntp or other agencies are doing? So I think where all of those federal agencies are kind of having to work hand in hand trying to come up with a way to Use those so I think partnering with groups like ntp fda epa about how regulatory agencies are trying to use that type of information would be You're at the forefront. I mean if we go back to academy reports Advocating those the use of non traditional approaches those have all been published in the last decade So everyone is kind of moving in that same direction Got it. Okay. That's obviously something we can follow up with our staff and then I just wanted to Clarify something that commissioner adler said I think many of us really do want to move away from a policy system Where there is a regrettable substitution, but I'm not hearing that moving to a class approach solves that problem It just changes the basis of it. So it might be instead of going regrettable substitution chemical by chemical There's still the possibility of going regrettable substitution class or subclass by subclass, correct? That's possible Okay, so this this is while this is certainly a better approach This does not solve the ultimate problem of chemicals are put on the market without any assessment of what the effects are in humans Especially children we have to wait usually decades to find out about the negative effects And then by the time somebody publicizes that and they're accepted We move on to another chemical and repeat the cycle over and over again So I'll again drawn a little bit different experience as was mentioned by commissioner berkel I served on the chair of the committee that looked at the chemical alternatives of assessment That was a project that was done through national academies and that report was published a few years ago One of the really really important Take-home messages was again the absence of data Does not imply safety does not imply no hazard And so that's where many of the regrettable substitutions have occurred is that we've got a well studied chemical We're concerned about it. So let's Regulate it on that basis. Here's a second chemical that's out here That's available that could say replace that chemical in its application But this one's not studied So I think again, what we would strongly encourage you is to to really take all those steps not to do that type of a process No data doesn't mean that there's no hazard Correct. We really appreciate it. Thank you again Thank you commissioner b. Oko Thank you. Dr. Dorman. I appreciated your summary I think I got it and um, I I do have some questions that I want to follow up I took some notes and I want to make sure that what I wrote down was correct and and thank all of you for the good work that you did more to commissioner adler's point the fact that the CPSC is a small agency and doesn't have the Finances or the size that the FDA has does not relieve us of our Obligation to make sure that we get to scientifically sound conclusions We just need to do them in the most efficient and effective way possible, which is what I think I understood you're telling us by doing The the class approach and and I I just want to make sure that I got that right And and commissioner adler asked you a question about you know, you said it was Why it was so critical to adopt a class approach I don't remember you saying it was critical What I understood is that what the committee did was give the cpsc a scientifically viable place to start That there were some the way you grouped the subclasses there were some common Scientifically establishable common characteristics that we could go from there. Is that correct? That's correct. Okay, that's correct So we don't need to do a 101 161 different ones, but you gave us a starting point. So again, what I would suggest is that the subclasses we've formed was using Strong science available now. It's been peer reviewed And if your staff was charged with trying to do the hazard assessments for those subclasses that be a very logical Offshoot of this report and would be very defensible Okay, that that helps and and I also I I wrote down here that I took away that you said by grouping the chemicals into the subclasses that you did you gave A signals. So we might these are things that might also have the properties, but they're not conclusions. They're just signals From which to proceed, correct? Okay And to address a point that all the other commissioners have raised I also want to raise because I think it's an important one The no data doesn't mean it's safe, but it also doesn't mean it isn't safe, correct? Correct. We just don't know You just don't know You don't know so in the class approach that the committee is recommending there may be policy decisions in which you sweep together compounds And make a hazard determination in which you actually didn't know that's that is one of the outcomes that could occur When you say policy decisions, are you I mean I translate that in my head to These are the margins of errors that we might be willing to accept Can you tell me what you mean by? policy decisions So what I mean is that there's certain so for example, let's say that a decision that's reached this And I'll have my slide here say the first option Of to extending the most conservative conclusion that all of the members of that subclass That's not a science the science helps to identify which was the most sensitive ones Might identify the outcomes of concern But the decision that we're now going to extend it to an entire subclass of chemicals That's to us as a policy decision. It's not it's informed by science But it's not one that science on itself can make. Okay. I think that's that's important. Thank you and and to just take that a little further when When you talked about having I mean we can use data-rich chemicals to draw inferences from data poor chemicals, correct? We can draw inferences, but they're not necessarily conclusive, correct, right and then We we aren't able to draw those inferences Any further than the basic inferences in other words? We can't tell whether the basic inferences that are scientifically Viable to draw from would actually be The same in different applications in other words if it a if we put this chemical into a child's product that they mouth May be different than putting the same chemical Into a casing of an electronic, correct? We're just talking about properties now So again, we're the hazard assessment is looking at the inherent toxicological potential of the compound The scenario you're now talking about is the difference between the hazard and the exposure and the risk And so but the inherent hazard Is independent of that application the risk? Exposures can be widely different But okay, that's an inherent property So I think maybe what I was trying to get at was more of a so you take a chemical and when you have it interact with An individual we could have just that I mean it's like having an addictive We have a we have a chemical that has addictive properties But it doesn't mean that if I use that chemical or my colleagues use that chemical We're all going to react to you know from an addiction point of view the same way, correct? So as a toxicologist I can't predict how any individual is going to react to the chemical But the way a hazard assessment's done is what you're trying to do is look at the body of the evidence to make Into termination as to whether or not that chemical has the potential to to cause harm And then what what at what level and at what level and then we get into the Exposure back to my example of child's products versus casing an individual variability So example, so for example exposures to an adult may be something we handle But an exposure to a developing fetus you could have a very different outcome So again, there's what in a hazard assessment what we're trying to do is Identify what are those inherent properties of the chemical that could cause harm? But that's where we are then so what you gave us is a place to start We still have to make those determinations that we just discussed, correct, okay, and That parsing from here forward still requires And those decisions still require a lot of expertise. Would you agree with that? That's correct. Okay, that's correct Okay, so I see you did a great job. That's very helpful. I appreciate that And what I want to do is just spend one moment about the expertise The expertise is challenging in a couple of ways One is that there's a lot of physical chemistry here and again, I would not recommend you necessarily duplicate that wheel Okay, but where it also requires some expertise is just all of us trying to understand and when I mean all of us the toxicology community Of trying to move from Analyzing traditional data to the less traditional data. So that's something as a community. There's Scientific expertise that's emerging in that arena as well. Okay That's all I have right now. Thank you so much. Thank you Thank you commissioner feldman. Thank you and thank you all for being here today This has been an interesting discussion And and thanks obviously also to your colleagues over at the national academies for their work and contributing to the study that you put forward Um, I had a question about the class approach and then I also had a question about the hazard assessment assessment scoping plan that you put forward um With the class approach that you've identified Taking a look at the subclass distinctions that are there Do those distinctions that are based on shared physical and biological Properties of of the chemicals themselves lend themselves to classification of Future chemicals that may come in to commerce are uh, those core subclass Properties the kind of chemical and biological markers that industry appears to be innovating around If I can't speak to the innovation issue because that was again beyond the scope of our committee But in answer to the first part of that question. Yes. So what we're finding is that what we're Stating in the subclass approach is that there's different functional groups that are in the chemistry so that Members within that subclass that all share that functionality oops I can keep going in the dark. I mean I'm a college professor lights. No lights awake or not. I can just keep going so But um, so the issue really is is those functional groups help drive the biology. So I'll use a simple example that Relatively straightforward. So for example We can have compounds with a double bond that double bond can get oxidized by the body And in some circumstances can produce what's known as an epoxide And what we know from those epoxides is those are very reactive molecules and they tend to be mutagenic In a number of carcinogens that share that double bond epoxide Production can ultimately be carcinogens. That doesn't mean they all are but that's like it's a predictor So now what you would do is if you've got say your subset that share that chemical feature If now dormant industries is coming in and creating another chemical that shares that feature Then that would be We would consider that to be a member of that subclass because it's being driven by that functionality And it may also be hazardous just like the other members of that subclass So that's a very kind of black and white example that toxicologists are very familiar with in the case of the phthalates That was you know, the ortho phthalate had compounds. They shared a certain chemistry. They shared a certain biology. So This is my understanding that the r&d piece of this was a little outside of the scope of your inquiry Let me try to ask it another way When you took a look at the 161 chemicals that that sort of make up the Universe that you you were examining that are the 161 chemicals that that are in commerce Is that is that a universe that's been fairly stable over time or are there New chemical entrance into that class that right so I'm trying to get a sense of how dynamic that field is So I my guess is it's probably dynamic and the reason I say that is for us What we tried to do was we went to a variety of different sources to ask What had other groups identified as organo halogens? And what we actually found was there wasn't a lot of similarity between different sources So source a may have identified 20 chemicals source b 50 chemicals and the overlap was actually fairly minimal And so that to me it's indirect evidence But that evidence would make me think that it's not necessarily stable as to what represents the organo halogen flame retardants Okay, thank you With respect to the hazard assessment scoping plan You talked about that that for certain sub classes that you looked at and and sort of Previewed that this was likely something that we were going to encounter That there are subclasses and will be subclasses where there's a paucity of traditional data. I think were your words That that put us in a situation where we're likely going to have to rely on in vitro versus other hazard assessment methodologies Why would we want to use traditional data versus in vitro is there a risk in Facing or inquiry on in vitro that we're going to miss something So you run the risk with traditional data as well So there's you know, I'm I'm a veterinary toxicologist. I've worked in the arena of public health for my entire career Traditional models. They're an imperfect model as well. So part of it is asking Kind of going back to what level of uncertainty do we have so doesn't necessarily mean that a rodent bio assay is any more predictive Than say an in vitro assay. We just have learned to accept it We've accepted the uncertainties around using traditional approaches And we're in a point in time in the toxicology scientific community Where those similar kinds of assumptions will probably have to be made and apply to the in vitro data Neither one is perfect and those are assumptions that are being made of necessity In in many ways. Yes. Yes Um, but they're also informed by the science So it's not I mean, it's an emerging field We have some data. So for example, um, the report talks a little bit about a pedotoxicity So for example Colleagues over in fda have found that some of the in vitro assays are actually more predictive than rodents because some of the rodent models Are so imperfect with respect to certain liver injury So it's not So part of it's out of necessity But some of it is also based on a better improved understanding of the biology We have tools to interrogate biology today that we never had 20 years ago Okay But those decisions then aren't aren't solely based off of a cost savings associated with in vitro And you may actually have in vitro studies that are more expensive than than other ways of looking at the assessment That's entirely possible What we do know is the in vitro assays are faster Faster faster or uh those discordant subclasses. I mean when you're looking at it and you get a uh, uh You're able to make a determination that the subclass is non-hazardous That seems to validate the subclass decision and subclassification that you've made when it comes back discordant, um, it would seem that there's then A drive to further segregate the subclass into sub subclasses So that's one of several options when Is there a risk that in Taking a look at a discordant subclass where you're seeing certain chemicals in the subclass behave a certain way and other chemicals behave Another way that that's going to call into question the the overall thoughts and assumptions that Went into making that subclassification Determination in the first place So we do recognize that could be iterative in other words the sub Classifications right now are based on an underlining assumption that these physical properties or chemical groups within the chemistry And what we think are predictive biology were all used to inform that subclass And so then in science we sometimes make mistakes So it is possible that data may be found in which the assumptions of predicted biology actually turned out to be A fallacy so that's that's an outcome that could occur Okay, and there could become a point where you sort of made that those further distinctions to the point where perhaps the the correct course of action is to Just begin taking a look at individual chemicals You may end up having to default to that It's just again that that's an approach you're already very familiar with but again We were charged with having giving you guidance for a class based approach. I understand I'm running low on time But I would want to open this up to the rest of the panel Who sat there very diligently throughout the briefing today You know, uh, dr. Dorman is do you or anyone else in the panel? Is is are there any other broad takeaways that the that haven't been raised that the commission needs to be paying attention to You based on the research here No, but i'll turn over to my colleague at national academies. He may want to talk about next steps for academies Please Thanks. Thanks, dave Yeah, uh, you know, we don't have As as dave mentioned before, um, you know, the academies are stand ready to help in any way we can You know, we don't we don't have a specific proposal, but You know in terms of next steps but We will be continuing the conversations with your staff and and And one way is just to you know, we can bring in Day back and other members of the committee back just to You know as you know, there are a lot of details in the report itself and We'd be happy to continue the conversations and come back and and talk to you further if you have further questions But I think that's the natural next step And explore options and if there is more targeted You know, uh, you know kinds of, uh interactions that we can take place over over weeks or months Oh, yeah, uh as well as more if you get to a point where you do need a another formal Consensus study at the you know similar or you know to the one that we already carried out You know, we can talk about that as well. I appreciate that miss mantis and miss mortel. Is there anything that either of you wanted to add Okay, then I have no further questions. Thank you very much Thank you very much. I think we will Just go another round of questions. I think a few of us have maybe some follow-up questions I wanted to raise one point and that is I'm going to quote my colleague that you unabashedly endorsed looking This class approach But I that's not my sense when the petition and the petitioners filed this petition. It was to ban a class of chemicals And our staff said We're not comfortable with that approach and to your point Having just gone through the phthalate issue and dealing with that They were not comfortable with Looking at an entire class of chemicals and so that's why we turn to you And and the study to give us some guidance as to what's feasible and how we look at this So I just want to be clear that And and I think we've already covered this we're looking at some place in between Chemical by chemical versus the entire class of non polymeric organohalagen flame retardants This subclass approach I would agree with that I mean the what I the subclass Is still a class approach, but it's not trying to deal with the large set 161 You would be trying to deal with one group with 10 one group with 11 those subclasses. So yeah, okay That that's all I have right now commissioner edler Well, I shouldn't resist. I should resist, but I'm not going to I'll just read your concluding remark Although the challenges to a class approach might appear daunting the alternative Individual assessments of hundreds of chemicals is unrealistic the only possible Practical approach for a set of chemicals as large as OFRs as a class approach But that's just a a quick response. I do have one or two Questions that I'm that I'd like to ask If we were to for those data poor sources turn to in vitro Costs would that then add to the budget that you've laid out the 14 and a half million in the eight years? If they if there were to be Considerable numbers of in vitro testing So I want to just clarify which budget you're discussing. I'm talking about the Eight and a half years and 14 and a half million Dollars for further study of OFRs that was spelled out in the NAS recommendation But the actual report as opposed to Yeah, just to clarify, you know that that particular number was not a formal part of our report The committee didn't you got my attention? Yeah But at the request of the staff we did at the staff level did do those those those estimates and it And part of part of the reason that the committee didn't provide those numbers directly was Just because of all of those all of the decisions that that Dr. Dorman went through that would affect those estimates and it was just impossible to run all the scenarios So we made some logical assumptions things like out of the out of the 14 classes that four would have insufficient data and that And that That the data exists to perform the assessments on 10 of those We then suggested one path forward would to initially do a pilot study that would let You know You go through all the steps for for a single class a single subclass And then follow that up with the lessons learned from that pilot Into three subsequent studies that would look at three subclasses At a time And so in terms of you know your question about the the additional in vitro work That was not part of that those cost estimates that was additional And we we had a separate table that again gave a range of costs for animal studies and And in vitro and in silico testing as well. So it would be additional So For anybody on the panel based on what your assessment is so far of the hazards associated the known hazards associated with organo halogen flame retardants Is this a project worth pursuing in other words, especially in the areas where we don't have that much data? Uh, is it your sense that for Somebody like us, uh, who are sitting up here as regulators that this is something that should be examined and pursued So What i'll what i'll try to do is i'll draw your attention. It'll take me a moment to find where in the report I'll just draw your attention to but for example, we tried to map the existing data For a variety of different assays That are in tox cast and other assays So for example, if you turn to pages 68 and 69 and a hard copy of the report that you have Which you'll see is what scientists refer to as a heat map Okay, and so gray in the heat map means no data okay Green means some level of activity And so what you can quickly see is that within the whole group of chemicals that we're talking about These are different biological assays that have been looked at in vitro. So using the tox cast system So that's the cooperative System that epa and ntp and fda developed in which in vitro testing is done So what you'll find is some of the classes have green So what you might want to be able to do is are there biological signals here that Can help guide and inform the hazard assessment In other cases where there's no data Coming back to tox cast and asking it's really critical for us to be able to do our work To have data for a variety of these different chemicals You can again nominate the chemicals and have that data generated that generation of data can occur fairly quickly Relatively compared to traditional approaches So I think having cpsc position itself so that it can use these types of data streams is actually going to be important Because you're going to encounter this type of data Increasingly for not just this class of chemicals, but other chemicals Thank you, that's a very helpful response and I just make another quick comment is that in terms of my task as a policy maker is I'd like to combine as much Material that's scientifically justifiable in order for us to make a quick and but meaningful and defensible Policy judgment. So I think you've really helped us enormously and again. I want to express my appreciation Thank you commissioner k Nothing further. Thank you madam chair. Thank you commissioner b. Ako. Thank you So let's pick up where you left off So we have 14 classes subclasses, right? Four of them you said had insufficient data. So we really know nothing and there's nothing to base a study on is that right? Well, I'm not sure we I'm not sure that's a correct characterization. Okay, that's what I'm asking specifically What I did say is certain class subclasses Will for it based on our survey because we didn't do the deep dive for every single subclass But based on the survey, there are some subclasses with very minimal data So in other words what I'm hearing is start where we have some data But if it was so the recommendation I think I'd give again as dave as opposed to the chair But it's based on the work we just did try to separate those two people they are they are but this isn't my first rodeo So what I would recommend is you start with the two case examples We had because it illustrates kind of the policy questions that you're going to be faced with And if you and it also builds on what the committee did So if you work with for example the halogenated organophosphate flame retardants and just ask yourself What do we do next so we have discordant data? What will we need to do next and what maybe what are some of our interim decisions? We might want to make in order to be able to move forward to me That's a very logical place to start because rather than try to look at the entire sweep of all 14 subclasses Start where there's some guidance here And work with academies or others to try to as needed But I think a lot of times what that'll do is very quickly once you start to ask the questions with it Which of these scenarios are we comfortable with that will provide guidance for how you deal with the other remaining 12 Subclasses that you've got to work with that's helpful. Thank you going back to The paucity of the traditional data versus the in vitro um You mentioned In response to some of commissioner feldman's questions that you run the risk of with traditional data because there's unreliable There's some unreliabilities, correct So with all data streams, they're okay, you know, all right So and and I think what what I heard you say was You know, it's still valid data But there are certain uncertainties that we have learned over time to accept right There are some assumptions that we traditionally make that say for example traditional data collected in a rodent We make some assumptions that that data could be predictive for human hazards. Okay. So now we're getting to in vitro data and Is is that for lack of a better term? Is that better or more measurable data? So it will depend again going beyond the report But it will depend in a lot of ways on the biology that you're trying to study So we'll in vitro if let's say we do in vitro testing, for example Does that help us do two things one measure the uncertainties that we were willing to accept with the traditional stuff and let us Know how well we did will allow us to do that But let me let me kind of step back because I I'm a little confused. I'm gonna step back for a second So CPSC and your staff already deal with in vitro data In the way that you traditionally deal with in vitro data, for example is mutagenicity data So for example, if you're trying to make a hazard identification for a group of chemicals or individual chemical You might rely for example on a traditional two-year cancer bio assay that's been performed in multiple species That's a traditional data Along the way, you might also have mutagenicity data. So you know that this chemical is reactive with DNA That's an in vitro assay that's done in salmonella and others So we're already as a scientific community and as regulators There's already a history of using in vitro data, but it's been grounded in our understanding of the process of carcinogenicity What I'm saying is that there's going to be in vitro data that will as it's emerging will also be grounded In our best understanding of the human biology or rodent biology to allow you to make informed decisions That's an emerging field of science. It's not new. It's not absolutely new. I mean mutagenicity the aims assays Been around for a long time and we kind of feel comfortable with that one So over time we'll become more and more comfortable with these nontraditional data sources. Okay, that helps. Thank you very much I I'd misunderstood and I'm gonna ask you a really tough question You said something um, and I don't even know how to ask this in response to commissioner Feldman where you said not necessarily stable ray and organo halogen class And I just wrote I had no idea what that meant. Do you remember? Do you remember say I don't remember the context I was hoping that so I'll try to put it into context So the question my recollection of commissioner Feldman's question is how stable is that set of 161? So in essence if we redid the inventory survey tomorrow, might we have found 162 Or might we find 170 next year? So my answer is that the instability points to a future group May have identified say three or four more that have come into commerce in the next year or six months I mean there's no way to predict that but it just doesn't feel like the 161 Is set in stone so to speak that it's actually a fluid group. Okay. Thank you very much. That's that's all I have. Thank you Thank you. Commissioner Feldman. Just one more question. Uh, and I'll be brief Looking through the the toxa the the talk studies that you've done Why would you make a decision to go uh zebrafish over mammalian So we versus in vitro. I guess right so We're looking at developmental effects Those oftentimes require an integrated biological system So what we want is to have a biological system in which the brain development Organ development is occurring because it's a very developments a very complex process So when you think about it, we started off as a two-cell organism and over time very quickly become a very complex Three-dimensional organism that chemicals and other types of factors can disrupt Okay, so the advantage of doing a zebrafish over traditional assays There's a couple of things one is the zebrafish tent those are short-term assays The other huge advantage that's increasingly becoming aware of in the tox community is there's a move away from using traditional vertebrate models To try to go to lower vertebrate models. So from an animal use perspective The history of toxicology is moving away from higher organisms to lower ones. Why? There's a lot of pressure there's increased costs of doing animal studies So it's cheaper in many ways to do a zebrafish study compared to a rodent study But there's also a lot of pressures to try to look at replacements And reduction of animal use and that's that's something that's been emerging over the last couple of decades So the zebrafish model also has some other real advantages And that the zebrafish model allows you very quickly to look at the morphology of the zebrafish And we can raise not we but scientists doing zebrafish assays can raise large numbers of embryos very quickly Doesn't mean it's a perfect model doesn't mean that there aren't discordant data from the zebrafish we illustrated that But what we're saying is there's going to be increasing use we don't think it's going to go away in the next five years So you'll have increasing zebrafish developmental data for this class of compounds So it may help to try to inform the hazard assessment if you think about using that type of data stream But ultimately we're interested in the hazard as it presents in humans and million vertebrates not zebrafish Well, ultimately Correct. I mean ultimately you're more concerned about the human hazard than even the hazard in mice or rats So again, that kind of points back to mice and rats we use as surrogates for people We've done that for decades But we all are very familiar with the fact that there have been case examples in which human hazards were not identified in rodent models I mean, there's some very tragic events where that can occur So it could be that you have to look at different streams of evidence in order to be able to make that hazard assessment Did you want to add something? I just want to add one thing that it's that I don't want the A misunderstanding with the report in that the committee. I don't think was saying either, you know, either in vitro or Animal studies. I mean, it's not the idea is that we use all the data that's available And we then try to integrate the data that we have So it's not as if we the committee was just advocating moving To simply doing what we called our the new approach methodologies, but really trying to use all the data that's out there Also, I'll also add that it could be in a tiered approach So for example, you might find For poorly studied chemicals where you're concerned about development and Effects that you might say do a tiered zebrafish look at how a biologically active is it and then confirm findings in a rodent model So again, there's a variety of different ways in which those nontraditional approaches could be used I understand. I have no further questions. Thank you very much. Thank you Thank you. I believe that all of the commissioners have completed their rounds of questioning so Hearing no further questions again. Want to thank dr. Dorman and the entire team who is here today To present and to answer our questions and for all of the work and I would just want to reiterate that and emphasize The work that was done on this report how helpful it is to the agency and to the commission offices So my deepest appreciation for all of the work that went into it I think you understand how Extremely important it is as the commission considers how to move forward to have the kinds of resources and advice that you've given us and I do look forward to continuing those conversations. I think dr. Sims made some Say movements towards that or suggestions But I do think that continuing those conversations would be a good thing and and in the interest of the agency so Um again my deepest appreciation to all of you At this time the commission is going to take a break and we will reconvene at 130 at which time we'll turn to our final agenda item of the day A briefing from cpsc staff on issues related to off highway road vehicles Thank you very much