 Okay. Welcome. We're ready to get started. This is the fifth meeting of the chronic hazard advisory panel on phthalates and phthalate substitutes. This meeting is a public meeting. If you have not done so, please sign the attendance sheet, which is out in the hallway. So the meeting is being webcast live and will be recorded. Mike, I'd like to start off by having the CHAP members introduce themselves. My name is Phil Murkis. Chris Jennings. This is Holger Koch. Andreas Kartenkamp. Russ Hauser. Bernard Schwetz. Paul Leoy. And I'm Mike Babich, the CPSC project manager. Thank you. We are all here today. One of the first items that we wanted to discuss this morning, we're now looking at a final product, which is our report to CPSC. We wanted to discuss the issue of peer review of our product and whether we want to have peer review and if we do what form it should take. So I open that up to the committee for discussion. Tell us how that has been done in the past. Yeah. Well, there's no requirement for peer review. The previous CHAP was not peer reviewed. I'm not sure about the ones before that. I think it's up to the panel. If you want to have some kind of informal peer review or something a little more formal, we can certainly set that up except, however, a more formal process might add a little bit of time. But that's really up to the panel. If you want to have certain people show parts, different parts of the report to different people, or if you want to get a group to review the whole thing, however you want to do it. To me, with the right reviewers, it always is good to have feedback. Yeah. And I think that the... I was going to say the only issue would be the timing. Yeah. It's back for a deadline in December to have it reviewed by April. That's not going to be... Yeah. And if we set up a formal process, we have a contractor actually proposes reviewers and it just takes a little time. Another way, you know, if you want to do it a little more informal process, I think we have a couple of people in mind to review the exposure scenarios part who are willing to do that and have expertise. So, you know, if you had some people in mind to review the other sections, I mean, that's okay too. I think I'm in favor of a more formal peer review this time. Do it. I agree with Andreas. This makes me feel good. Why do we disagree that? By formal, you mean have a contract to do this. No, not necessarily contract, but have external experts who look at it. Not necessarily by contract. Well, we have a contract set up where the Versar will propose a list of reviewers and do a whole process. I mean, they have to make agreements with the peer reviewers. In that case, that takes a little bit of time. But we can... That's how we... If you wanted a more formal one, that's what we would do. And then we, or the panel, certainly, I mean, you usually have to make a list of reviewers. And then we say, gee, that's a good idea. Or, you know, why don't you get somebody else or something like that? So, well, try to think of the process. I mean, if we start now, they can have everything set up in advance. It's a matter of timing. How long do you want the people to have to read it? And if it's going to be a long report, do you want everybody to review the whole thing or just their expertise? They've got to sort of compile all the comments. I mean, how many for a report this size and this broad, I mean, how many peer reviewers want to have? Not more than five. And I think I'd rather have them review the whole thing, because if they don't, then we're going to have people who are going to be halfway between never, never land in reality. Can I ask about... I think, procedurally, it would make sense if we suggested some, but it would make me feel a little better if then there was an instance or a group of people, maybe in CPSC, who would then look at it and draw in other people as well, if they so feel... Is there a formal... So we have a really neutral process of selection of peer review. That's fine. And if you, again, it's up to you. If you wanted to have people at some of the other federal agencies or something like that, we can bring them in, too. It's really up to whatever the panel wants. Mike, can you... Having gone through the lack of peer review of the first CHAP report on THAL 8, can you... Do you have a feeling that there would be a better document, a better process in place if there had been a peer review the first time? What are we trying to fix by having all of this extra effort of a peer review? You know, in the case of the previous CHAP, I don't think it would have changed much. I mean, peer review always makes the report a little bit better, but on the other hand, I think here the justification might be just because it's such a... Because of the implications of the report, it's a huge undertaking, and it's going to have, I think, implications probably well beyond CPSC. If we get comments from external reviewers who have the same level of expertise as the people around this table, we'll get a lot of comments. And my question would be who's going to negotiate through all of those comments and transform the report to reflect what was said? Would we have to have another meeting to discuss? We're going to take these 29 comments, and we're going to ignore these 42, and we're going to have to do more review, research kind of effort. We'll be able to address these comments. I think it's best to decide ahead of time are you going to actually write responses to the comments. You don't necessarily have to, but it's good to think about those things. Hopefully the comments won't require any major changes, rewrites, but it's always a you never know. I guess part of my question is, and I don't know how this is normally... I don't remember how this is normally done in the IOM or other groups that go out for formal review. Some of the rest of you have been closer to that than I have in the last few years. But does the host organization, CPSC in this case, go through and make changes to the report based on the comments of the reviewers? No, I think... We have to have discussion about what we're going to accept and what we're not going to accept. Yeah, I mean, I think the panel has got to make any changes. And I don't see any other way to do that. It wouldn't be if CPSC made revisions, it wouldn't be your report. Or maybe, I'm assuming we wouldn't be so interested in editorial type comments. We're more interested in substantive comments. If somebody said this paragraph would be better here, that's not really... But that's also easy to do so. If you're going to have peer review, you have to take into account what people say. I don't think we have to formally respond to every comment, but I think we have to state which ones we accept and which ones we reject. And I think it's fair to do that. Since this is not normally always done on CPSC's watch where you have a peer review, we just have to figure out what's the most efficient approach for us to handle the situation. Clearly, there'll be some divergences in terms of the way they felt we approached it versus the way somebody else would approach it. But if we can maintain our integrity about what we did and why we did it, or adjust it accordingly, then we just state it. I have no problem with the peer review and with the utmost care and scrutiny. Does the panel feel that we could do this by email? Good. I mean, the peer review itself are responding to the comments. Most NRC reports I've been on. We do most of our response by mail. I think it depends on the nature of the changes. If we're going to make editorial changes, that's one thing. But if we're going to make conceptual changes in direction of how we interpret data and that we're including new data, I don't see how we can do that by email and have that be considered a public process. Conference call. If it's something really serious, we can have a conference call. Conference call that's open to the public. Of course. Yes, absolutely. Well, thanks to mine. Would the comments be public? They might have to be. Well, maybe not. But being that the report isn't public until it's released. Can we not adopt a procedure equivalent to what NRC does? It's very straightforward, like Paul has described. Yeah. Peer reviews done by email, responses are done by email. Yeah, absolutely. And the NRC reports aren't public either before they're approved and go through peer review and approved by the panel and everyone's happy. Yeah. And the review is unnamed. NRC, they are. Yes. But you have to ask them beforehand to make sure that you have. Yeah, you just need to know in advance are their comments going to be made public after the fact or at least FOI-able? Well, I don't think the comments, but I think the fact you named the peer reviewers is fine. Okay. I mean, you have to define it. You have to define a protocol. And I think just call the NRC because those guys know what they're doing. My concern is the timing still because it seems to me on the phthalates NRC that took the whole summer. I remember correctly. And it was a beautiful summer. Sorry? It was a beautiful summer. Yes, it took time, but I think it's worth it. We can impose timelines, deadlines, et cetera. Do you have a major problem with that, Chris? No, I'm just concerned about, you know, if this backs us up to say we need a final report now by December, I'm not sure we're set to do that. If we have until February, maybe that's better. If we have until, you know, so it's the timing part because the whole thing has to be done by April. Is that right? That's the deadline. I think a draft report by February would probably allow for peer review and still meet the deadlines. That nicely gets us into the next topic, which Bern is going to chair. And that's how we're going to get to the final product. And I think we need to have some serious discussion about that. Thank you, Phil. I think it has become a little less clear in my mind now than I had before this discussion about peer review. Let's work backward from April until now and see what we have to do in the next three meetings. And then we can try to fit peer review in there or not. I'm assuming that today we're going, today and tomorrow, we're going to be talking more than we have in the past about use of the exposure data. And we're going to talk more about hazard index. And we're going to, one by one, touch on the various components of the report to just check progress on these and see how they're fitting together. And see if by the end of tomorrow's meeting we have identified some gaps that we need to fit yet into this or if we're satisfied that it's coming together in the pieces that we have already had progress working on. So I see this as still kind of a relaxed meeting in terms of timing to be able to talk about process and product to some extent and to hear updates. So that's kind of my vision of what today and tomorrow will be. I think by November we need to start talking about what are the recommendations that we're going to make to CPSC. So we will have the results of the exposure information. We'll have the results of the hazard index work and the rest of the information from the human epidemiology, from the animal studies, reproductive and developmental in both species. We'll continue to be filled in. And if you need additional information on no whales or whatever, we can continue to work those out. But I think by November, and we have tentatively agreed on November 3 and 4 as the next meeting, I think we need to begin to come to grips with recommendations because we're running out of time. And we have to compress this effort of evaluation that we've been in to ask what specifically are we going to recommend on the phthalates that are already restricted and regulated and those that are alternates and those that are out there without much review or regulation. I would hope that a work piece on the table next time would perhaps be a draft of some recommendations. And we may have to have a subcommittee of this committee get together, at least by telephone, to think about structuring that, to get a framework for that so that it isn't going to be just one person's opinion by November. But that's what I would see for the November meeting. If we stick to this timing, we would have another meeting in January and February time frame. One more meeting before the final one in April. So in January and February, we would perhaps finalize what we think our recommendations should be to CPSC. So we will have discussed them in the November meeting and argue if need be what they might be, but then have another chance to think about it for a couple of months and in January or February be prepared to either buy into or fail to reach agreement on what those recommendations might be. So then we have to have more writing assignments to have a document that perhaps we could look at by January and February. The parts from Russ and Phil and I can hopefully be pretty much finished by that time. But the final thing that's going to tie into the recommendations is the work that the rest of you are doing that's going to give us the basis for saying, well, here are the three reasons why we should restrict this one or here are the two reasons why we shouldn't restrict that one, kind of things. So I think that we'll all have to fit together by the end of the February meeting so that perhaps we would have an intermediate deadline for a draft report that we could review before the April meeting. So then in April, what would be left would be one last shot at saying, do we agree or disagree with our own recommendations? And the report would have been fleshed out to the extent that we would be able to leave that behind as well. There will be final touch-ups that have to be done in wording but not in concept. So we don't have to have something that we're prepared to sign at the end of the April meeting, but we have to be in agreement of what it says as revised and then it can be put into final form. And Mike, I'm not sure what the final process for CPSC is to get agreement of all the committee members on a document. Maybe you can comment on that. Well, in the past there's been a signature sheet. I don't know if we're doing the final editing and so on. I don't know how we would address that exactly, but FedExit around or whatever. We'll worry about that. So then how do we retrofit a peer review into this? Conceptually, I agree with a peer review. But I'm afraid it might take us for sure one meeting. It might lengthen this process by one meeting, if not two. Because for us to gather the information and discuss it, I would assume that the information that we will get back will be fairly voluminous. It won't be measured in a few pages. It'll be measured maybe in inches that we might get back on this. Although this open comment period is strictly for those that handful of reviewers. It's not for everybody. Okay, so that should cut it down from inches to maybe half inches. But it'll still take us some time to discuss the concept behind some of these recommendations for change and whether or not we want to engage in that. So I don't think that's a one-hour discussion. And whether we are prepared to ask CPSC if we can go from April to July for the last meeting or maybe September. That's the consequence of extending this through that period of time. That can be determined only by Mike and his colleagues within CPSC. Andreas. What's the final deadline according to the law? Well, it's two years. The first meeting was I think on the April 14th. I think your appointments were officially like the 12th through 13th. So it would be mid-April 2012 is the official deadline. If we go beyond that, we would have to ask the commission for permission to do that. And that's usually not a problem because, you know, if the report's not done, not quite done, it's not quite done. Could we not argue we are still sort of within the spirit of the law if we get, according to Bernie's timetable, the report finalized by April with final touches and then at the peer review process? Well, in that case, it would still, I mean, it wouldn't be final. So technically we would still, yeah, it might be the spirit, but technically we would have to still have to ask for an extension. Would you see that as a problem? No, because I think it would be a defined period of time. Well, the alternative is to get ready with the final report, final for peer review by the January-February meeting and then submit it to peer review and limit it to an April meeting and then finalize. I don't know how feasible this is, but I think. If we plan to have a version that's ready for peer review after the February meeting, presumably after the February meeting, we would, somebody would have to do some cleanup, you know, not changing of the report, but just typing things that we suggest to be revised. So you're talking the middle of February if we have the meeting early in February, before the copy would go to the peer reviewers, and if we give them what's the normal thing for a peer reviewer, 60 days maybe? I think we would probably want to do less than that. Maybe 30 to 60 days, 45 days, something in there. Then the April meeting in our part would, we don't have, we wouldn't have a chance to second guess our recommendations because they would have already been out there for peer review. But we could use the April meeting to discuss the comments that we got from the peer reviewers, and thereby hold this to a one meeting extension, not a two meeting extension, but that cuts our time a little bit short for being able to formulate recommendations, think about it for a couple of months, think about the implications of those recommendations, and then bring them up again to discuss. Now, this time we have to make a decision. Last time we just talked about it. Now that we've thought about it, let's make a decision today. That may be a luxury that we don't need or can't afford, I don't know. It's my sense that you would favor your original suggestion to have the final touch up, the final report in the April meeting, then ask for an extension for peer review and then implement the peer review results. No, Honduras, I hadn't tried to waive the risk. It's a risk-risk decision because our recommendations are going to be out there forever, and I sure would hate to have been part of this committee and have a final report out there that we wish we had had one more shot at discussing the recommendations. I think that in my mind is probably more important than a voluntary peer review, but if our recommendations gel very easily without much concern for further discussion, much need for further discussion, then we won't need that time. I don't assume that we're going to fall into recommendations without much discussion. It'll take some time to work it through so that we really agree with these recommendations as opposed to a last-minute effort before we ran to the airport. I agree with you. I think we should have maximum time for this. So let's finalize, I agree with your suggestion. Let's finalize the report at an April meeting and then see how peer review ties in. That gives us maximum time. Now, a lot of the editing within the committee, I mean, you can all review each other's chapters and do a lot of editing online, I mean by email, so that will help to get the report ready sooner, and that's probably something we should be starting to do pretty soon, is seriously reading the different chapters, or whenever you think they're ready, start reviewing them and sending comments. One lesser important consideration. This report should be as long as it needs to be. And I don't have in mind a goal of having a 100-page report or 150-page report. I think it needs to be as long as it needs to be, but some reports seem to be a lot longer than they would need to be from the past. And it takes a long time to write that extra material and to review it and to reach agreement on it. And what disciplines should we put on ourselves to have these reports be minimalist versus longer than minimalist? How brief do we want this to be? Or how lengthy do we want it to be? I mean, that's a question I think we each have to ask in terms of our writing assignments and what the purpose is. For example, on what I'm writing on the developmental toxicology, what is the purpose of that section? And as I initially started writing it, I assumed that it was to thoroughly analyze the in vivo developmental tox literature with the idea of establishing the most credible noels for the most sensitive end points that Chris and Holger would be able to use in their hazard index analysis. If that's the case, then that my section has to have a fairly thorough analysis of that literature in the text so that people who read it will know how we generated those noels. If that's not the case, then that section can be a whole lot shorter. Well, I think not only is it a matter of what are the noels for those end points, but that's where people are going to look for what really is the phthalate syndrome and what species does it involve and what credibility does this observation have for human risk. And as a result, I think your section on developmental toxicology is far more important than my section on reproductive toxicology because the effects on sperm count or sperm motility or whatever that might define a noel for a reproductive study is of secondary importance compared to that nest of responses that we consider to be the phthalate syndrome. So yours has to be as complete as it needs to be so that that doesn't get second guessed. I think on my part, if there are some subtleties that people agree or disagree with, it isn't near as important as whether or not we missed something on a phthalate response. And for the human data, obviously the question behind this whole exercise is, is there any evidence in humans of a phthalate syndrome equivalent? And then we have to have as many pages as it takes to address the integrity of the information from humans about a phthalate syndrome. And based on that conclusion, what it means that this is primarily a change that occurs in rats to a lesser extent in other species and the risk of making a risk decision based on one species when you have some data that shows that the same thing doesn't happen in mice and rabbits and marmots to the same extent. So here's a one species change that's driving a risk decision for humans. We have to bridge all of that. So I would hope that our space in writing would be taken for making those kinds of, that kind of information forward. Yep. And the reason I bring this up is I think when we discuss our individual sections, which we're going to do, well, it's on the schedule for this afternoon, I hope that each of you have read the updates that were sent around because I, for one, would like your comments about how I put it together. Is that the most appropriate way to put it together? Have I missed something? Have I been in error in what I've written? These are the kinds of comments that I would like to receive. And I think everybody around the table would like to receive those kinds of comments because the sooner we get those comments, make those changes. I think the sooner we can put the sections together and have a draft document of the entire document that we are eventually going to send around for peer review. I think a more important question than how long should the report be is what we put into the front of the report versus what we put in appendices. Because if we have a long flowing document without appendices, it's going to be a lot to work through. Whereas if we can capture important points in the upfront part of the report but put the data slogging parts in appendices, that might make it a lot more readable. And for mine, for instance, I'm going to have quite a few tables. I'm going to discuss in words the various in vivo studies but then have tables that summarize those. Would those be best in with the text or would those be better in the appendices? That's the kind of decisions that I think as a committee we need to make. Since we were talking about sections of the report, one of the sections which is not listed really is the conclusions and recommendations in the process for putting that together. I know it will be a group project and decision, but most likely one or two people will take the lead in writing that. So not that we need to decide that now, but I think we should during this meeting, in terms of the process for the overall conclusions and recommendations, how we'll put that on paper. Yeah, I think if I remember correctly, let me just look at my outline here. Yeah, I think, no it's not, but I think that's something that Byrne and I said we would take a stab at and then get input from the committee. So what do you think would be the timing of that? I mean you probably can't write or put thoughts down until the November meeting or possibly after this meeting between now and then, I'm not sure because that really will be a critical part of the report for us to have time to... Well, I would hope, for example, I've read the update from Chris and Holger and the one part that's not there is their summary. And had that been there, I would feel much better about beginning to write that conclusion. So if in our discussions today, those kinds of things begin to emerge, then I would feel better about beginning a draft. Likewise, if what Paul tells us today begins to give me some framework in terms of exposure scenarios and how that's going to play into our conclusions and our recommendations, then I think Byrne and I could begin to start writing that. But right now I don't have any of that, so... As far as the exposure goes, in May we developed a bunch of scenarios that we thought would be reasonable to consider. And since then, Versar has been gathering data, putting it into tables, categorizing it. This summer we expect to see the results of their analysis for exposure, which would be kids toys versus food versus anything else we can find that would be reasonable. Once that's done, and I think what will happen is that the material will trickle in over time this summer. It should be then taken over and given it to Holger and Chris to see how they want to compare it to the daily intakes they've been estimating with the biomarker data, because that will be a real crucial set of analyses before anyone can write a summary. Otherwise the summary will just be based upon biomarker data and I don't think that that's sufficient when you're trying to consider the children's toy issue or those issues that may in fact impact pregnancy in a woman. So I think that's where the whole... It's not a hold up, but that's where the process is going. It takes a lot of work. The final report I think will be a lot of tables in the appendix with crucial data and analyses that have been done to select the scenarios that were most appropriate being used in this part of the text. That's where I think it's going to fall. And when do you envision that the text part, the summary, the nuggets are going to be available? Any idea? Once the analyses are finished, we're going to get a summary this afternoon of where they are and then once we get a summary of where they are based upon the charge we gave them we'll be able to give you a better, clearer picture. I would say the fall would be more logical. Their annual... Their final report is due in January on the whole analysis but the results should be available within October. I'm not sure that's sufficient time for the current status of the November meeting but that's the way the contract was written by CPSC. Is that fair way of describing it? Yeah, and it's really what they can... They understand that we need this as fast as possible and they're working hard on it. It's also a case where there's a lot of sort of groundwork and once they get that done then the rest is easier. Chris and Hope, where do you see your analyses standing at this point? I mean, how much... What are you waiting for, if anything, to push your analysis further or do you see that you've pretty much done as much as you can and where are you? Good question. I think there's still the opportunity... Holger has some additional values that we should consider. I think I have tried to implement some of the new reference doses that tables... Maybe there was an update that needs... So there's some... Yeah, that will change as I develop the developmental talks ones, I think, because that's the ones I think we want to use rather than the re-produced, re-pro-talks, no-ells. So that's something we need to discuss at some point. But once you have those, what else would you... I think with the structure we are pretty confident that we have really reached the final structure of our chapter and we will do some minor modifications in terms of conversion factors and points of departure. Some recent literature came out, but that's very easy to modify. And as you said, we haven't yet written the final conclusion, but I think in a way, although our chapter is independent from what is Paul working on, I would prefer at least to have a first impression of what result he comes to, because although our chapters are separate from each other, we cannot pinpoint exposure to the source. So in my conclusion of our chapter, I would prefer to have a first impression on the findings of Paul and the verse approach. And I think your chapter, Paul, will also be quite an extensive one. Mostly in the appendix. But I would like to see if the approach is fixed. I would like to see the... That will be part of it, yes. That needs to be. They have to write a full report and then include approaches used, approaches to be defined, how we got to the decision. Absolutely, it has to be part of it. An interesting little rub here could end up being if your estimates for daily intakes are much different than ours, then I think we're going to really have to address that. Yes, that's the main issue is where they fall. Because biomarker data, we're going to be doing exposure construction, we're going to try to estimate children's toy emissions, and then exposure issues based upon how kids touch, use, suck, and track material. Those all will determine the kind of percentage of the pain. You will actually come up with estimates of daily intake, right? Yeah. Per chemical, per valley. For ones, there's data. And then we'll try to use analogs for others. But the rub is where the data is. Now, there is a question there. We're going to be grappling this for a while. It's a pregnant mother issue. How that, you know, which her intake is, how it affects the fetus. And there are a couple of things we'll be talking about this afternoon, a little bit dicey, but we may have some ramifications also that we haven't really in a group thought about before, but we thought about May. So we'll leave that one for this afternoon. I wonder how, sorry, it's a bit loud. What you come up with are estimates for NOELS if you want to use that term or points of departure will also impact on what Chris and Harga do. But from what I've seen, one result of that analysis is that the impact is not very big because the numerator of these terms, exposure, varies so much. And that I will certainly have that information available well before the next meeting that I can get to you. I should have had that done already, but I don't. In terms of timing, I think it sure would be great if we could have the Astrid Index process taken as far as we can get it by November. And if we can have the opportunity to not wait for a report from Versar until January to have these two pieces be the substance of our time in November so that we see how these two fit together, because if we wait until January, we may have the same problem. If we wait for the report and it comes out towards the end of January, we're not even going to have a meeting in January, February. And we're looking at extending this process quite a few months, even without doing a peer review. So if we could get the exposure data and the Hazard Index information to work together for a November meeting, that would allow us to begin to talk about recommendations than in that November meeting. So is that possible, Paul? I'm handicapped by Versar. Yes, sure, but my handicapped is the, does these calculations, which should be relatively easy once they finish developing the scenarios around which we described it, which we set up in April and May? Well, depending upon that. But at today's meeting, Paul, you could go over what the objectives were of the exposure assessment and what some of the input pieces are and expected outputs without anything quantitative, but at least letting us know. That's what the presentation this afternoon should be covering. Okay. All right? Correct. By the way, I haven't seen the presentation. I'm a little distraught. I'm supposed to be done beginning of July. And so therefore I'm worried about Versar's ability to make a deadline if they weren't able to get something to me by the beginning of July. So right now I'm a bit frustrated because I haven't seen the darn thing. You'll see it. Yes. Well, I will learn what you'll learn at the same time. Okay? So it's not without pressing these people if they haven't done it. It's a little disconcerting because if we have to meet a deadline, they have to ramp up their activities. This needs to be a really deep discussion between Chris and Holger and Paul and Versar this afternoon. I think that's the key component that's going to play into what progress we make in November and beyond. I think the other components, component burns, mine, Andreas, will have those essentially drafts by November. And then we can just tweak those with any data that comes in subsequently. But your two components are in flux until we get Paul's paper. I'd like to revisit the question of peer review, external peer review for another minute because I think if we are going to request an additional extension, the sooner we can do it, the better. And if we commit to external peer review today, I think we need to ask CPSC for at least one more meeting beyond what we have scheduled to do that. And that's about a three month period. We're talking about being done by July, August, rather than April. And if we don't do external peer review, we may still need that meeting, but we don't know that now. We may know more by tomorrow. But I think if we are, if we want to commit to an external, if we want to request CPSC to have an external peer review done, I think it would be helpful if we made that decision and that recommendation today. Well, either today or before we convene or before we adjourn tomorrow. Yeah, that's what I meant by today. What's the ramifications if we request now and we don't need it? Well, I think I would probably not put forward the request until we get closer to the deadline. So basically what you need is an affirmation that if we need it, we want to request it. That's basically what it comes down to. If you're not going to put it in now, we're going to wait. Yeah, I mean, I would another around it. You know, do you guys agree that we should have an extension? We say yes in principle. And if in fact we need it, you pick the date. Yeah. And you're going to submit that request. Yeah, I'll find out how much lead time need. I suspect there are people in CPSC who are asking Mike, are you going to be done in April or not? These are the budget people. They need to know. Well, I mean, it's not just budget people, but people want to know. But we're talking about a couple of them. It's not unusual for something like this to be a couple of months late. But just to come back to the exposure part, in tab six, there's a copy of the work plan gives the chemicals and the kinds of things that we're looking at. That's at the end. It's the, I think the end of tab six. So we decided on, there's I think eight valates, which reflects its combination of what is the ones that we think are important and the ones for which there are data. You know, the funny thing is what's the chemicals that you find in the biomonitoring studies is not entirely the same set that you can find exposure data for and so on. Good situation. Which is not unusual because the purposes of the biomonitoring is X and purposes of exposure assessment are usually be acquired by Y and they never been done simultaneously because they weren't part of a consistent study design or regulatory process. So the fact that there are differences in terms of the types and the varieties of data is not to be surprising to anyone. But are there examples in which both were done biomonitoring and exposure assessments? And children? In any scenario. I'm just getting at, you know, what kind of agreement or disagreement are there between the two methods? I mean, yeah, the only one I know is a Wormuth's paper. Yeah, but that's the only example. Anyway, I'm told that because this is a congressional Congressional mandated deadline that we might not have the flexibility. We're not sure what the process would be so I'll have to find out more about that. So that would have an impact on whether we get peer review or not? Well, if in fact there's no flexibility then I think we need to think in terms of meetings at shorter interval than three months to get this done. But somehow we have to focus our attention to get it done by April if that's the case. So it may require instead of a meeting to talk to a meeting in November and another one in January, February, we may need to have one in between there. Mike, would you be able to determine while we're at this meeting what the situation is? I can ask. Any other input? Well, this guy's the limit. Okay, well, the rest of the meeting was to be spent going over the various sections. And I think we've finished the agenda for this morning. We might as well launch into the afternoon's agenda and Paul will have to wait on yours because Versar is not here. Well, Versar and Matt are going to be here after the first thing after lunch. So we'll have to wait. We'll do that when the first thing after lunch. And Andreas has agreed to begin a discussion of his section on mixtures and cumulative exposures. Okay, this is... Shall I briefly talk to it, talk you through it? And this is... This material was essentially solely reviewed already for the NRC report 2008. So what I've done is essentially an update of that because that was necessary because since publication of the NRC report, quite a few other experimental studies have appeared in the literature. But just to summarize it briefly for you, the situation is very similar to that with the study of individual phthalates. The rat was selected as the model for study. And so it is not surprising that almost all mixture studies have been conducted with the rat. I've structured this in the following way. First of all, I've summarized experimental empirical evidence to show combination effects between various phthalates. And there's a very recent study that has just come out, Olga has distributed it by our grace group. And then secondly, evidence to support the notion that there's also combination effects between phthalates and other anti-anderounds. So we acknowledge that a lot of the efforts in that area of study were devoted to asking or answering the question which of the two competing models, independent action or dose addition, is best for approximating experimentally observed mixture effects. And the answer seems to be it is dose addition, which is good and has relevance in relation to the hazard index approach exercised by Chris and Olga. So there's strong evidence in the experimental literature to support the notion that dose addition is the concept that approximates best the experimentally observed effects, both of phthalate mixtures and of mixtures of phthalates with other anti-anderounds. So really that in the nutshell is a summary of this. It is of note that a couple of quite important studies appeared after 2008 after the NRC report. There's one paper by Kristiansen et al. which also carries my name on it, where some synergisms were observed, but also additivity, and then further work from Olga's group. Andreas? Yeah. So if it's dose addition, is it dose addition without any qualifier as to... or how does one do that? It is dose addition, but what happens is, you know, these are concepts. Mixture toxicology has one goal. I always had one, and that is how can we or can we indeed anticipate the effects of mixtures from the toxicity of its individual components? Can we make predictions when the mixture effects are not known? So you always start off with the effects you observe with the individual chemicals, and you select one endpoint for study. And in this case, in this experimental literature, for example for phthalates, the most sensitive endpoint in the rat has always been selected, and that is suppression of fetal andron synthesis. So you measure this for each individual component, you construct those response relationships, which you then utilize to plug them into the dose addition concept to make a prediction of the joint effect of a mixture with defined composition. And then you ask the question, do the experimentally observed values agree with this prediction? Often it is amazingly accurate, sometimes not so. But in this literature, side by side, the predictions generated by using the concept of independent action were also compared, and there's one strong thread coming through, and that is that independent action, or as it's here sometimes called response addition, consistently underestimates the experimentally observed mixture effect. So dose addition is really a good way of approximating the experimentally observed values. Yes, there's one point to add. What requires consideration is the potential for effects to occur that deviate from your auditivity expectation, for example synergisms or antagonisms. Antagonisms haven't really been observed much in this area, but there are indications of synergistic interactions when you choose penile malformations as the endpoint for analysis. That came up in our paper, Christian Sinetal, and we've been puzzled by this. We cannot offer a mechanistic explanation for it because all the other endpoints in that same study typical of hallmarks of sexual differentiation in the male were all conforming with dose addition. Only hypospadia's penile malformations did not, and there are echoes of that and some hints in various papers from Earl Grey's group and Cynthia Ryder, Cynthia Ryder being the first author. The problem with these data by Cynthia Ryder at ALIS is that it's difficult to conclusively analyse it because there are a couple of provisors. The data they use to construct predictions come from historical databases where the exposure regimen adopted for the mixture isn't exactly the same as for the single chemical, so that complicates the analysis a little. But there are hints, and this would require attention. Actually this analysis of the literature would, if you like, provide the foundation for the hazard index analysis conducted by Chris and Holger. It actually exposes one gap or one difficulty, and that is the experimental literature suggests there's clearly the potential for phthalates to interact with other anti-androgens. But this knowledge cannot easily be taken forward into the hazard index analysis because often we lack the exposure information for the other anti-androgens. That is a problem. It has an impact ultimately, in my opinion, on this hazard index analysis in the sense that you probably have to be a little flexible with the interpretation. For example, if the hazard index approaches one and this is based only by considering phthalates, you have to take into account the background exposure to other agents and bear this in mind and probably not get too hung up about number one being exceeded or if it ends up around one. You want to take into account that exposure and reality is to a lot of other anti-androgens as well. And a special issue, that's more recent evidence from a GRACE lab that is the interaction between phthalates and TCDD with these endpoints. TCDD itself doesn't induce the typical spectrum of the phthalate syndrome, but indications are that it can exacerbate the effects of phthalates. So that's quite important as well. Andres, I guess go ahead. No, I rarely speculate. Andres, so if you're saying that something approaches one, you have to use a degree of caution, especially if we're doing these added doses and the fact that there are other backgrounds. So when do we get concerned? When does the hazard index begin to rise to the level that we think that it's real for the phthalates we're under consideration? Because that way heavily on where we go in terms of our conclusions and recommendations. You're asking me in essence how long a piece of string is, but all I can say is, for example, if you end up with a hazard index of around 0.8, just below 1, then you have to exercise caution if you then want to blow the trumpet, oh, that's nothing, no case to answer. But anything else is very difficult to quantify simply because we're lacking the exposure information. So what if it goes to 1.2? Well, the same problem. With phthalates alone to 1.2, then according to the premises of deterministic risk assessment, you would have to say two things. Either you want to stop and then say, okay, we have to do something in terms of exposure reduction, or you want to refine your analysis if this is possible. But often that's not possible. At least you're giving me some boundary conditions to think about it, which I think is important for the whole committee. The last points you make, Andreas, remind me that in each of the sections that we write, it would be helpful for Berne and for myself if you would identify issues like this that need to be included in a summary statement, for example, in our document. I haven't done that yet, but I will. And that's for all of us to do that. Say this in the past, and I hate to sound like a broken record, but when we began to talk about other anti-androgens, it seems to me it was a box to address how big that box actually could be without actually having a number. But I mean, that is a very big box, right? There are a lot of other anti-androgens that potentially could have an increase, a combination effect with the valleys. I don't know how to go on about this. Yes, we're venturing in the territory of Rumsfeldia known and unknown here. That's the problem. That's the problem. This is quite frankly an area that requires more research. The scientific, if you look at the scientific literature, what scientists like to do is to use certain chemicals as tools to clarify mechanisms to clarify phenomena. And the selection of tested chemicals there is therefore biased. It doesn't necessarily reflect what is really out there and what perhaps should be studied. That is a problem, but we have to deal with it pragmatically because we can't reinvent the wheel here. So, Chris, in answer to your question, well, you remember in the NRC report there was a list of substances potentially deep to be considered in this. This included a couple of pesticides, but also other more persistent chemicals. I would like to mention the polybrominated diphenyl ethers, the dioxin-type contaminants, et cetera, et cetera. That's a whole host. But again, as we have found in our Hazard Index paper, which I sometime ago presented here, what this analysis shows more than anything else is in fact the data gaps. But there we are. We need to therefore exercise caution in interpreting a Hazard Index based solely on joint exposures to certain phthalates. Any other comments? I suggest that we take a short break. Reconvene at 4 to 11. Okay, I'd like to reconvene the chat committee, please. And what I'd like to do now is to have the committee go to tab 8 in their workbook. I wanted to go through my section on developmental toxicity and not going to read it word for word, but just to go through the sections and get your input. What I decided to do is to start off the discussion with an introduction, and in the first section have a few paragraphs on male sexual differentiation in mammals. And essentially what I did was to take this verbatim from the NRC publication that we're given a copy of. I don't know whether that's something we want to do, but I felt just as I was writing it, why reinvent the wheel? They had a nice concise summary of it and why rewrite it using words that probably were less succinct. So that's what I did. I thought that was important to have that somewhere in the document since we're going to be referring to that quite frequently. So any comments about doing that and is that extensive enough? I think that's a great summary. We don't have to reinvent the wheel again. And the NASA report is perfect this way and the way you exerted it, that's just perfect. I agree. The second section is where I described the rat phthalate syndrome. I thought that obviously needed to be described in some detail and I think we probably need to have that in there once. And then we can all refer to it. So does anybody have any comments about what I've put there? I've tried to be quite concise. If you don't have comments at the moment, please submit them to me after the fact. Not only did I put in the obvious outward signs of the rat phthalate syndrome, but also linked those mechanistically to what was going on at the biochemical levels, gene expression levels. And also what I've done all through my document is given references right at the end of each section and we have to decide at some point how we're going to deal with references. Can I make one comment about the page 5, the top where you end summarizing the study by McKinnell et al. of the marmoset. Probably we have to build on the entire discussion about that, but could you perhaps add or stress that the measurements of testosterone were conducted at birth? Yes, after your comments earlier, yes. And if you did the same with the rat, you wouldn't see anything either. And that section 3 then is the phthalate syndrome and other species excluding humans. I think I've captured the literature there, but if anyone knows of any other references that I should cite, I'd like to know that. Chris? Maybe I'm reading quickly, but just the importance of the fact that it is a syndrome and on a particular, on a given study, if you look at a single endpoint part of that syndrome, you may actually miss the more strength of it as a syndrome. Like the phenotypic response on a given animal may not show all of them. Yes. You need to make that clear up front, I think. Not all features are seen in anyone's study. But by taking it as a group, one or these, I'm not sure people analyze the data that way. And I'd be happy if you e-mail me verbiage that you particularly like. I'd be happy to consider that. Okay, and the mechanism of burn? The fact that in your title of number 3 on this page 4 brings attention to the fact that you're excluding humans suggests that it's someplace else. There must be a parallel discussion about what the ethylate syndrome looks like in humans. Yes, that's a good point. That, I would assume, would come in Russ' section. It's not currently there. Well, I guess when we get to the human data, we can discuss the statistical genesis syndrome, which may be a counterpart, but the level of human data is very limited in relation to that endpoint, or that syndrome, TDS. But I would think at some point you'd want to discuss specifically what the testicular just genesis syndrome is in humans, what the features of it are, and then you relate that to the rat phallic syndrome. Yeah, there's a, when we go through it, there's very little in there now, but I could definitely expand on that. That'd be a nice connection to make. One thing that I am not sure we had in our outline to begin with was, we need this section on the human syndrome. It isn't the ethylate syndrome, but it is as close as we come to it with what little information there is. But someplace in here there has to be a visible discussion about based on what we know in humans and based on what we know about this one species response in the rat that we call the thalite syndrome, what does that lead us to conclude about risk to humans? We have to have that discussion in here someplace. Yeah, and I think that's going to come at the end, I would think. Yeah. But that's another thing that we need to put to make sure that we don't forget later on when we're writing the discussion. Okay. Because if we fail to do that, the rat data stand on their own and readers will wonder, why do they have all that rat data when they don't talk about it in terms of the human risk? Yeah. Mike, are you taking notes? Could you put that down in our parking lot of, yes, ideas that we want to include in the... My section, which I'll highlight, but literally it's just a few lines. It's trying to do what Bernie said, is basically take the human findings and put it in the context of what we see in rats, the thalite syndrome. And just to step back for a second out of context of this to pick up something else and put it in the parking lot that we talked about earlier this morning, and that was, we said there was a need to have additional information on anti-androgens in combination beyond thalites. And how big is that box? And how important might it be? Or how unimportant might it be? I think that's another parking lot issue that we need to have in the discussion someplace. And, Andreas, I didn't have a chance to really look carefully at your document. How much discussion, if any, do you have on anti-androgens in Europe? It's a separate chapter. Thalite's in combination with other anti-androgens. There's quite a few studies published in the peer-reviewed literature now, which I've summarized. So it's definitely an issue to be taken forward. And it can be easily done. Rusty, you have a comment? Yeah, were you going to go on, I guess, to section four? But I had a question about whether, because I haven't looked through everything you've written either. But in terms of differences in metabolism across species, where would that come into the document? Good point. Cover that a little bit in mind, but not very extensively. Well, the next section is mechanism of action that really isn't metabolism. Because it's relevant, I think, to the more of a set. And I think that discussion about species differences in metabolism or similarities between thalites and metabolism is important in the context of which of those metabolites might account for the biological effect in that species. So all of that needs to be brought together. Yep, I'm going to have to add a section. Would have counterparts where you'd have metabolism in rats, other species, and then humans. So, would you cover the metabolism in humans, or would that more easily be done all together with rats and mice and whatever? I haven't currently included anything on metabolism. I wasn't sure if it was going to be there's exposure assessment section, or it's something that with Holger, I think we can put into whether it goes in the epidemiologic section or in exposure section or standalone section, I'm not sure. I'm yet not sure how deep we should get into that. What would it lead us to? Do we know really anything about differences in metabolism, species differences, and the data we are comparing the human biomonitoring data, it doesn't have to be the active metabolites, the ones that are excreted nearing. So it doesn't really tell you something about the metabolite levels relevant in target tissue. Well, I was thinking of it mostly in respect to the marmoset data, but I agree with you. It could potentially be part of a human discussion, whether it's in the epi section or another section, but it would be very brief. But I was thinking of it more as potential explanation for species differences or dosing differences across animal species apart from humans. Let me take a stab at it and then we can see where it best fits. Holger, that sounds reasonable in terms of... for relevance to the animal data, the metabolism. I would really keep it brief. I support Holger. I think I'd keep it really brief. I don't see where it's going to add that much. I'm fine with keeping it brief. I just think it needs to be there because it could potentially account for species differences in helping interpret the data. It's a lot of rudimentary data there, like metabolite levels and amniotic fluids. And we don't have much human data on that compared to some animal data. And again, I think the levels of metabolites in urine doesn't say anything about activity if it is the active metabolites or what is the active metabolite and in which target organ at what time. Yeah, but still, if Russ wants to discuss that in relation to possible species differences, that can be very focused. Fear not. I don't fear. It's perfect. Fair enough. And then mechanism of action. I've just really briefly touched on that. That needs to be completed. There's a fair bit of literature there in the rat. I don't know much. I don't think there's anything in the mouse or rabbit or marmoset that I'm aware of. And so I plan on getting into, you know, testosterone production, gene expression. That'll be summarized. And then on the next page, six, I started putting together a table for the different phthalates under consideration. And then you can see that the table is not at all well-spaced on the page. But what different phthalates do in terms of these different endpoints? So testosterone production, star mRNA, insulin-like factor 3, SIP 11A, et cetera, where it's known. So give us some idea of that in a very nicely focused. And again, we'll have to decide whether such a table is best kept right in with the text or put as a appendix. And then... I would really like to have this table in the text. That's an important table. And, of course, it can be updated with recent literature. But I think this is very important and nice table. It doesn't take up a whole lot of space and easy to update. In section five, just the subject of cumulative exposures to phthalates. And again, that... I'm not completed. Can we not slot in there what I've written? Pardon me? Can we not slot into this what I have written? Like a Lego brick, put it there? I guess we could. I guess there's no reason why we couldn't do that. Otherwise it would be partially redundant. Okay. Can you think of any other topics that need to be included? The next section I go into is the developmental toxicity of phthalates in rats. If you think of anything subsequent to the meeting, let me know. So where's the whole issue of anti-androgen is going to be prior to us talking about it, that it's a combination of phthalates? Are we going to have some review of the anti-androgen? Yeah, that's going to be in Andreas's contribution, which is going to go in. I deal with them only in us and when they turn up in mixture studies. But we could, yeah, I think Chris, that's a good suggestion. It's easy to have somewhere a section about what other anti-androgens do. It's a parking lot thing. You have to remember it. I can write that. It's no problem. That could be another section right in here if we wanted it. Yeah. But very brief, Olga. Androgens. Can I make a comment about page 8? Developmental toxicity, phthalates and rats. You very helpfully list a set of criteria for what you regard as an adequate study for deriving a noel. And I agree with most of these points. But the last one, where you refer explicitly to OECD guideline 414. I'm fine with that as well. Only to note that this guideline as far as I can see would miss out the most important end points for the study of developmental effects of phthalates. They should be noted somehow. In what sense? It doesn't stipulate measuring AGD or nipple retention. You're right. Does OECD have a new protocol that includes that? It's under debate. It's the new extended one-generation study, but it's not sanctioned yet. It's under debate. Going back to the beginning of that section, the goal is to systematically review the published peer-reviewed literature reporting the in-utero exposure of phthalates and pregnant rats. Then after careful consideration by the committee, this review is limited to currently banned phthalates. The three phthalates currently on interim ban and four other phthalates, DMP, DDPP, P and DIBP. I need to add a justification for the four other phthalates. We've talked about why we thought we needed to add these, but I haven't added that justification yet. I also add that because the first six phthalates were extensively reviewed by a phthalates expert panel in a series of reports from the NTP Center for the Evaluation of Risks to Human Reproduction in 2002, our review of these phthalates begins with a brief summary of these NTP reports which are then followed by a review of the literature since that report. So the decision was made not to review all the individual reports prior to 2002 because that had been done beautifully by that, by those NTP committees and we felt it wasn't necessary to redo that. So all that I have done is review the literature since that report was published. For the four other phthalates the following review covers all the relevant studies available to the committee. And then from the available literature for these 10 phthalates we then identified the most sensitive developmentally toxic endpoint in the particular study as well as the lowest dose that elicited that endpoint, the NOEL. Finally we hope that should be NOEL. Finally we evaluated the adequacy of a particular study's in terms of our confidence in using a specific NOEL to calculate a reference dose and that's what when I completed this will there be a table that I will then pass on to Kristen Hoger that they will then use that in their hazard index analysis. And then that completes I complete that section with our criteria for determining an adequate study on phthalates is really mechanistic in nature was not done to determine a NOEL yet I think it's important to evaluate that literature to acknowledge that it's there what it determined but not to determine a NOEL so just make that clear. I really like this entire structure so that you divide it between the 2002 watershed I like that very much it's very clear. I have a couple of questions in these tables some studies are highlighted in bold I guess this is because you lend particular importance to those for deriving a NOEL but it's nowhere stated what the bold means yes I think in terms of that first table your D.B.P yes yes that's exactly why and I didn't indicate that anywhere that's the reason and these tables are still in process and I need to modify them because what I've done on some of the tables and you don't have copies of it when I have a NOEL what I'm doing is putting down the NOEL based on what end point it's not clear from this table for example on the Mill Creek at all 2000s study 50 milligrams per kilogram per day it doesn't indicate based on what end point that will be added so this at a glance gives you some of the important information and that from those criteria that I indicated in terms of an adequate study you should be able to look across and glean from that why we either did or didn't choose a particular study for consideration just a quick question we spoke at this I think at the last meeting in terms of finding the literature how did you go about making sure that there aren't recent papers since 2002 that may have been missed I'm asking to because for the epidemiologic literature I've been looking and potentially could miss a paper so have you used the systematic process well my systematic process is to do PubMed searches in multiple ways and I think I've captured most of the literature but I Holger and Andreas are going to do a backup for me and Mike as well just to make sure I haven't yes and that's where potentially someone like Earl Grey or Paul Foster may be helpful yes in making sure there aren't any papers that were missed with the degree of variability in some of these studies when do we add to it it seems like there's you've done a really good job here if there's a one or two papers missing will that influence how we look at the data well I want to know where you end because you've done a really good job here I suspect that I missed what I would call an adequate study may have missed a minor study that wouldn't qualify as one that I would derive a no well from so but I just have to be careful where in the asymptote you want to stop I just want to make sure I haven't missed any obvious study and unless you know what you missed then you can't determine it's important can I just ask a higher level question and that is so it seems that you're focusing on developmental yes and Burns going to focus on reproductive correct are we going to put the two together in terms of maybe the most sensitive of the two combination are we going to do a review for developmental and reproductive and then the combination or what what are you all thinking about well Burns done a review of the reproductive on the developmental talks and Burns developed a list of no wells for reproductive endpoints I will develop one for developmental talks I think at that point we need to discuss which no wells we're going to use this is a point of confusion and maybe a point of controversy as some people try to establish turf because generally there's a there's an agreed upon difference in the design of developmental versus reproductive studies it's uncommon for reproductive studies for example to be done by gavage developmental studies are typically done by gavage not by dietary administration but there are some reproductive type studies that have been done by gavage and they go on for several months not for several critical days in pregnancy but they go on maybe for a couple generations that animals have been dozed by gavage well as a result it gets kind of fuzzy sometimes whether you call something a developmental study versus a reproductive study I have focused more so on whether or not I think in general the reproductive studies are looking at endpoints that are different in the endpoints that define the phthalate syndrome and that it would be more difficult to identify pups with some of those phthalate syndrome effects in a reproductive study for example if a pup doesn't look right a mother will eat it but you don't have accounting of 100% of the pups in a reproductive study in a teratology or an experiment in a developmental talk study where you're dosing for three months you're dosing for four or five days that are critical and you're looking directly at the newborns or the pups that are less sensitive to identifying the effects of the phthalate syndrome in the developmental talk studies where you're dosing for three months you're dosing for four or five days that are critical and you're looking at newborns or the fetuses in a way that you're more likely to detect these structural abnormalities than you would in a reproductive study if people think they're going to identify phthalate syndrome effects through a multi-generation reading protocol that may not be the right conclusion because it decreases the sensitivity of being able to detect those effects so I messed up as I created the first version of this table on no whales because I included a study that was more developmental in nature than it was reproductive and fortunately Phil pointed it out to me and I took it out and replaced it with a different study that in fact had about a tenfold higher no effect level but it was the best no whale that we had for that phthalate, for that design but it's not 100% clear that that one's developmental, this one's reproductive but we need to look through all of it and that's why in the table for mine I identified that one column as being, as whether or not the no whale was a phthalate syndrome effect or some other effect that was detectable in a fetal death in a reproductive study that's one of the sensitive measures of an effect of reproduction is litter size well that doesn't give you much information about the phthalate syndrome but it identifies this chemical as a reproductive toxicant so that kind of discussion needs to go on between Phil and me all the time just in the sense of process I mean are we meant to do an analysis with the reproductive different than the developmental or should we come to sort of a blending of the two maybe perhaps the most conservative well if it's been my sense that what we're most concerned about are those studies where there is evidence of a phthalate effect in a unique one of the phthalate syndrome and that would drive the hazard index evaluation as opposed to the fact that we have decreased fetal weight or we have fetal deaths or we have some other endpoint that is related to that phthalate but is not part of the phthalate syndrome so I guess that's something that we should discuss but my assumption has been that at least the phthalate syndrome effects have a higher priority and it's a neater part of what we're trying to evaluate and bringing in all kinds of other toxicology responses that may make it more difficult to understand whether this is really a phthalate effect but is it reasonable to assume that the phthalate syndrome reference doses or noels would be lower, more sensitive than not necessarily that's something we should be able to answer that's another question for the parking lot I mean if there were for example a study done a reproductive study where the exposures were done from adults to phthalates and they found an effect on sperm quality or fertility or some such and the noel was lower than any developmental tox noel then we'd have to we'd have to discuss and say well maybe that's a noel we need to use I suspect that won't happen but we'll have to see well you know for it depends on the phthalate for the ones associated with the phthalate syndrome that may be the most sensitive endpoint but for the others it probably isn't I think I look at it as for the cumulative risk you might want to look at phthalate syndrome endpoints but if you're looking at individual chemicals I would look at whatever the most sensitive endpoint is that might not be the phthalate syndrome and that's gonna I think be an item of discussion if that ends up being the case just a comment from my experience of having done many developmental tox studies in rats, mice and rabbits and reproductive studies on the same chemicals in rats generally I think there's enough of a difference in the sensitivity of those kinds of studies between development and reproduction development and reproductive studies I can't remember any examples where the reproductive studies had a lower noel than the developmental tox studies and I wouldn't say that it's automatically because development is more affected I think it has to do with the sensitivity of the design of the studies you know when you're counting pups in a cage that's a pretty low level of sensitivity when you're looking for very very subtle changes in the ossification of bones in the sternum of a large number of fetuses from a large number of litter that's a different level of scrutiny than what you get with just counting the number of pups in a cage that are left from yesterday that's what Paul Bernard said brings up another question for me do you plan to add a chapter on comparing the potencies of the different delights? like something on the relative potencies because we have seen that looking at the non-observed effect levels we have a lot of different studies made in different points in time and looking at slightly different endpoints and in the cumulative dosing studies like that some excerpts have been published by Earl Gray you can see that they had problems using this published non-observed adverse effect levels in generating data for their mixture studies so they looked at the ED50 values to prepare mixtures on the relative potencies and this is as you can see in the approach case 2 so that we on the one side of course look at the interval studies with the non-observed adverse effect levels but also on the other hand side at studies that really enable us to compare the relative potencies of the different delights so I would like I can help you of course I would prefer to also see a chapter on comparing the relative potencies of the different delights because sooner or later this will have to be a decision we have to come down to does it have to be a separate chapter can this be included on a separate section it could be included in these tables in some way that when we look at it we look at the relative potency as a right along side of what we have here not a chapter on its own of course but just some kind of conclusion or a final paragraph but if you put the information in a table either relative potency xxx versus minus minus minus and you know all down the list that helps I think drag that conclusion simply into this section rather than making it a little bit I agree that we need to have this information in there in terms of how what's put in the report is important because while we think of writing this and reading this from front to back there are very few people who are going to do that and there will be people who will open up this report and want to know where is the discussion about metabolism and potency so it has to be identifiable in a table of contents or in an index so that people can find it as opposed to having it scattered among ten different chemicals and in that context I would think I'm not the expert in pharmacology of the phthalates but it seems like their absorption is extensive there are rates of their half-life of elimination is probably pretty consistent from one to another so it's not misleading metabolism and distribution and elimination-wise to do that comparison and the potency probably reflects the biological effect not a consequence of the metabolism I think it would be a good place a good thing to have in there someplace this is simply and we have to be aware that there's not really much literature out there looking at the or comparing the different potencies of the phthalates naming which of the phthalates are really the potent ones which are the lesser potent ones which are not potent so I think it's important that we kind of sum it up based on the most recent literature you would do that based on noels or lowels that's exactly the question I raised because it might be difficult we might have to look at some of the mixture studies that are right now going on because they have the problems using the noels to combine to prepare their mixtures and they looked at the effective dose level at the ED50 values to prepare their mixtures potency means at least two different things to different people one of them to some people means the nature of the effect that it causes at whatever dose level so if you have something that's carcinogenic it's more potent than something that's not carcinogenic according to that end point the other group of people who are trained more as pharmacologists and toxicologists think of potency as causing an effect at a certain level of exposure so in that sense I think we're interested in comparing different levels for the various ones and are they less than 10 milligrams per kilogram per day or are they 10 to 50 or are they over a thousand and that's another definition of potency but we have to be aware that there are at least two different audiences looking for different things in the question of potency to add to that of course you can compare the potency of phthalites by looking at media and effect doses by looking at some other measures of effects smaller effects I don't... I would hesitate us having to get into this discussion but I think potency comparisons cannot easily be made on the basis of no-else but I'm not sure we should go there the simple reason being that if you understand how no-else what they are they are not associated with the same magnitude of effect you can't easily compare them that's right and I'm not clear since I don't do these kinds of studies I'm not clear how you would derive potencies based on these in vivo studies that I'm citing Andreas, can you... you seem to have you needed a rationale behind creating the mixtures as Earl had it or has it in his studies and this rationale is not solely based as you said on the no-else or on... it has to be based on something else it's based on those response relationships of all the individual chemicals so you have that but the issue is the way for example we have done it and the way our great lab does it for that we characterize the responses around medium effects it's a totally different kettle of fish if you want to design what we call a low dose mixture study that forces you into going really finely into doses associated with very small effects these are two separate issues two totally separate demands in terms of experimental study assessment resources etc etc so yes for those studies where mixtures were conducted say and then mixture ratios constructed in proportion to median effect doses you could compare the potency but yes it is possible but Andreas would you agree that if you had benchmark doses instead of no-else but it would be absolutely yes but do you want to calculate benchmark doses for all well I don't know when you're doing your reviews do you actually identify are you looking at a no-el versus a benchmark dose are there benchmark doses that you're seeing rarely if ever I don't know that I've seen one study that had a benchmark dose I mean the trick is getting consistent end points and if the data are there we can do them easily enough actually Kent has been for the 10 tox reviews that Versar's doing Kent has been calculating benchmark doses but I don't know if they're all the same end points I mean we could look into perhaps when I complete my table of the studies that I think are the best studies for a no-el we can then look at those and go back to the papers and the data and see if we can from their data calculate benchmark dose the calculation isn't hard it's finding the right dose response data but I would still support to add some data as you said on the 50 levels the potency estimations you use for the mixture studies because mixture and cumulative effects will be one major topic of our task Yes Holger but it doesn't for designing and interpreting these mixture studies it's not that instrumental it will be instrumental and important for other parts of the report and yes you can lift the data out of the mixture studies but I wouldn't put it there together with the mixtures but we can glean a lot and I think it's a very sensible proposal Phil made look at the high quality studies and see whether in addition to a no-el we can also calculate the benchmark dose There I say it's lower confidence in it Well some people would say for comparisons we would probably look at the MLE we're getting very technical Okay so that's my section and it just goes through one after another and as you can see as you get to the phthalate substitutes there's very little and sometimes no information at least that I was able to find and that's going to present a problem in terms of what we say when there's no information actually it's easier there's no information so we have to make a decision as to whether we want to request follow-up I think that's the issue pardon me I missed that if there's no information there's no information but the problem becomes what we do is refer quest for follow-up if there's no information request for follow-up yeah research or well yes that's a different question when there's no information I mean how I'm thinking of in terms of how do we factor into our charge to make recommendations when there's an absence of data there's an absence of data yeah you can't make a judgment unless you have definite analogs that you can compare it to you know we're exactly identical except for one or two components but we're really getting into arena speculation if we don't just say well there's no data we need research to quantify whether there is a danger here or not so as we proceed we're going to refine these documents more and more and I think it would be it's really going to help us to proceed if we get these documents emailed out well before meetings and receive back comments so that people on the committee have really read these and made comments that's really going to facilitate our meetings and I will I think as soon as we leave here begin to put together different sections so we have what looks like a document and did you want to add anything burn I think this back and forth emailing is going to help us reach our deadline yep just back looking at your chapter so when you have a bolded table in a table a bolded study that maybe reflects the ones that you think are the best and if there's a discrepancy in the estimate the no well have a procedure in your mind about are you going to how are you going to select a no well from that do you choose the lowest one? I wouldn't put anything into the fact that some are highlighted and others are not at this point because quite frankly I'm not sure why some of them are highlighted and others aren't I will sort that out but just in terms of your thinking when you're going to look at ten different values how are you going to choose the one well many of them can be eliminated simply because they don't meet the criteria and those that do and again this is a discussion that Bernie and I will have but I think what we'll do is probably choose the one study that has the lowest no well be the most conservative when you have to justify just by choosing the lowest no well being most conservative with X number of studies you have to justify that selection whether nine studies with a value of X and a study with a value Y is lower than X how do you justify selecting Y that's going to be part of what I write at the end of these the summary section where I lay out the rationale for the table I'm going to have to justify that as long as we have it justified but you have to be really careful when you do that so that was going to be my suggestion is it might be that you want to choose a conservative one and then come back and you know something that's not so conservative whether it's the higher one or whatever and then we can do analyses and say they didn't really matter we don't need to split hairs 50 versus 100 it's not going to matter but I guess I would like to say there needs to be a limit of how many times I'm going to run analysis some sensitivity analysis I think would be helpful because then people don't get hung up on we can certainly do that and in response to your comment I can almost assure you that there's not going to be a situation where there's going to be one study that gives me a noel of one and then there are going to be five that give me a noel of 100 it's not going to happen it's just a matter of discompletion using a sensible approach to summarizing what I think are really good tables there's no question about it it makes things a lot easier to understand you did end up in that situation where you had one that appeared to be an outlier and you had a cluster of three or four other well designed studies that all have a similar noel I think there's a reasonable basis for taking that cluster you might even take the average of the noel if they say they range between five and ten milligrams per kilogram all of them and then you've got one way down there where it's a tenth of a milligram per kilogram I think you have a basis for saying that there was a cluster of noels at seven and then there was what almost appeared to be an outlier at X fold below that I think it's reasonable to point that out in your evaluation in our interpretation you're right though it doesn't happen very often but I have seen that happen and you deal with it as an outlier but is it an outlier just because the others didn't design a study that has is it a dose of design no you have to have a basis it could be that the level of examination of the data was poorly defined and it has a very low noel compared to the standard design where it's very clear how many pups they evaluated per litter and how many did they do for soft tissue versus for skeleton or did they do both all of them for both so if that's poorly defined and the one that has a very low or the dose preparation is not well defined they just gave this but they have no chemistry data to support what was actually those are things that you would take into account by sorting that one that outlier off and handling it differently than the cluster of those where everything was well defined as it should be they were equally well done but I'd also include equally well done in a similar design if you're saying an order of magnitude lower that the other studies didn't even consider I don't understand that because just because somebody has a much lower number they're right and the other five or six are wrong I don't understand that you have to be really careful on what you're aiming toward you're aiming toward consistency and coherence which is what I do in understanding risk model you have to take into account whether or not there's a coherence among the data because you have one that's an outlier doesn't necessarily mean it's the right order it could be an outlier that's low it could use the same criteria on a high there's a high one here really high and the rest seven or eight are really, really low does that mean that the one really high is worse or maybe better really careful on how you're going to do that I think that's the complaint about the NOEL versus the BMD and so the design plays a major factor and so if different studies have different you know, largely different designs then you could be seeing an effect that's the point I was trying to make the thing else being the same it's a transition into my section I'd like to talk a little bit more about uncertainties about NOELs because there are a lot of them and they often don't get talked about despite the very large number of studies that we've had in front of us to review on the Thalates there are very few of them that are well designed for identifying a NOEL they frequently involve small numbers of animals they frequently involve a selection of very high dose levels where somebody in a lab is trying to confirm that their mouse of choice either does or doesn't respond to this effect or their rat of choice whatever it is they're using or their method of analysis would detect an effect of a Thalate that's not risk assessment type of studies so despite this huge literature there aren't a lot of studies that were truly designed to establish a NOEL so we're picking the best from suboptimal studies in quite a few cases the NOEL is highly dependent on the power and the background response mice for example in particular have very high background of cleft palate some strains of mice have 50, 60, 70% cleft palate and you obviously don't use those for developmental tox studies you'd never detect anything well the rat is resistant to developmental toxic effects compared to some strains of mice and rabbits so that's another factor to consider that when you establish an effect in the rat and you have a NOEL that's dependable it's pretty meaningful because it's a harder animal to get an effect in so the power of the design makes a big difference in what you would conclude is a NOEL and if it's a study that is claiming to have a NOEL and they only had 4 or 5 liters per dose level nonsense if it is a study that had a significant effect in almost all animals and they only used 4 or 5 liters per dose level and they only used 2 dose levels it's hard to say that there wasn't an effect in that study but it wasn't designed according to the minimum standards of FDA, REPA or OECD so it's a little bit harder to know what to do with positive studies when they're sub-optimal in design and it is a negative study when it's sub-optimal in design it truly is dependent on the selection of dose levels as you've already noted and because some studies are poorly designed to identify a NOEL the dose levels are 10, 20 fold different from each other so there's a huge place there for a NOEL and a LOEL but you'll never define it well with a study that has intervals of 10 to 20 fold between dose levels and especially when it's a study on a phthalate and they started at 750 milligrams per kilogram and went up well there are a huge number of those studies in this literature the distinction between reproductive and developmental effects we've already talked about a little bit NOELs are much more useful within a laboratory or within an organization than they are across laboratories because if a laboratory has been doing a lot of developmental talk studies and they use the same criteria for selecting the maximum tolerated dose and then spreading the dose levels below that and they have other pilot data to determine how wide that spread of dose levels should be that's the standard way of doing studies in many laboratories from those labs the NOELs have meaning and within the National Toxicology Program we had standard criteria for the selection of dose levels so that NOELs from one study compared to another done by the NTP in the same laboratory because we use the same criteria all the time for how we design studies but when you've got ten studies done in ten different laboratories and many of them have never done a talk study before a developmental talk study before and they just kind of guess at how to do it those NOELs are pretty shaky now it's very difficult for readers who aren't developmental toxicologists to know which ones are firm and which are shaky because they pretty much look alike in a peer reviewed manuscript often but those are the kinds of things that Bill and I need to take into account so that we when we have a table of NOELs it's really not gems mixed in with garbage which is not too hard to come by but is it with that thinking is it reasonable then to actually identify labs or authors that would fall into your description of being of high quality and then let the others be more sensitivity comparisons to the you can do that let me use the EPA labs as an example versus the NTP conducted studies Earl Gray and his colleagues conducted a huge amount of knowledge but not all of their studies are well designed for establishing risk many of them are for understanding the mechanism or for establishing the sensitivity of Wistar versus the Spragdolly rat for example so they don't their goal in the EPA labs is not to design these magnanimous big studies to have a lot of confidence in a NOEL in contrast the NTP studies are less well known for mechanistic contributions but the slope of the dose response curve is pretty firm from the NTP studies because of a selection of a fairly large number of animals number of dose levels and the importance and the uncertainty the more dose levels we would add and we would add more animals to the dose levels or we would duplicate studies but there are two government labs ten miles apart doing studies with similar objectives in mind but doing them a little bit differently because it comes down to the mission of the organization I know that that isn't all transparent in publications that's a fair way of commenting they're two different philosophies and I think it's important to consider but that I think it's reasonable to say that these tables would be helpful like you're saying some of these you as by knowing that kind of information you would discount some of these more than others one of the value of the NTP reports that they'll refer to because there are studies that would otherwise appear to meet the criteria but the panel of experts disagreed with how they selected the NOEL or LOEL and it's always noted in there that the panel disagreed with the author so they might give a different number but they don't just give a different number out of the dark they explain that this is different than what the author said the NOEL was we have an obligation to do the same thing again it depends on the quality of the study and the experience of the investigator and I would not want to second guess the panel of experts who were on a CERHR panel because they collectively had much more experience than I do so I wouldn't be inclined to second guess them and I wasn't there for the discussion but when you have the paper in your hand then it comes from you don't know where and you have never seen a paper from this lab before and it looks a little bit you're uncomfortable with what thought might have gone into the design of this and I wouldn't mind interpreting it myself just based on what I see and comment that the reviewer's interpretation is such and such so we can disagree on a NOEL but we have to be very clear about documenting why and if Bill and I and Russ can be helpful to identify what are the studies in here that really have the best quality to them that probably gives the best number and that's regulators do that all the time what you're saying then is that at the wrap up of each of these chapters then there may be a final table that takes sort of that synthesis up with a single number from this my I think Phil has done the same my tables reflect that process having going on already so all of those sub-marginal studies I just don't put in I don't have a table yet but there will be one and I think if you look at the end of yeah on page 12 so I have a short paragraph derivation of NOEL for DVP one of those and that's not complete there will be one of those for each valiate the table that I have in the front of my section is section 7 is that right is more of a distillation than what Phil's tables are his have really reproduced a lot of the experiment a lot of what's in the report and I haven't attempted to do that but what I have I've boiled it down to the minimum beyond the name of knowing what the NOEL is and what endpoint drove it primarily trying to distinguish whether it was something from the valiate syndrome cluster of effects versus some other reproductive endpoint and then the reference so mine is a pretty minimalist approach I guess it could be a table of the this could be a summary of the tables of yours so I would ask those of you who are looking at using NOELs whether this table has enough to be helpful and I think if any of the NOELs from your table are going to end up in the hazard index analysis I think we need to have more robust discussion of that compound that follows back in the text it hasn't been a table that's between the text and this table so I can create more tables like what you have if that would be helpful so I mean right now I would say don't do anything but we'll see what happens when I complete my analysis yes and Mike a question for you are there more tox reviews coming yet on more chemicals but on the CDs that you have there are I think eight finals and there's two more coming you I just emailed Friday one to you that came in late and that was dimethyl and then diethyl is the last one so there's one more coming so we've got somewhere somewhere in my notes we've got let's see a set of 20 30 some tox reviews of phthalates this last group of 10 are relatively high priority ones first we had the six that are covered by the CPSAA you got those a year ago since then we've given you maybe 19 or so that were sort of lower priority ones and then right now you've got eight of the 10 what we did after the six we took all the other phthalates and prioritized them and you got the low priority ones first and now you've got on that CD the 10 top priorities after the first six if that makes any sense and so you will and I can actually explain that a little bit better when I get to my part but I'm going to go through all of what we have tox reviews for I think I'm number five so Mike if you compare will you be comparing your data with the summary table well the plan initially is for the panel to have these to use to have all that information there to use I mean obviously you've done a lot of writing in reviewing these things in reviewing the different sections I'll make I'll do the best that I can to make sure that you have all the relevant papers and that this point out what's not consistent or something like that what I hope to do is put together some sort of table you know I prepared that overview of phthalates actually burn if you look in section five there's a after the text there's a giant table we're up to 51 phthalates now that includes that includes the six phthalates in the regulation it includes anything where the production is high enough to report to EPA that's more than a quarter million pounds a year or 25,000 pounds a year whatever the threshold is also what's registered and then there's a couple like dye that we added because more so because of academic interest because they cause those effects it's not commercially significant so we've got a total of 51 chemicals and to the right of that well table just says why it's on that list the second page actually if you go to the last page it tells you what we have talks reviews on and there's a few that don't have talks reviews and I'm not sure we need them most of them there's no talk about it anyway but you can see under CPSC it shows how many we've so we have done a total of well 25 because I'm double counting the cast numbers for DINP and DIDP so we've done 25 six of those were peer reviewed and HRSAR is doing 10 and they're almost done and then we've got reviews from others the NICNAS and the CERHR reviews and what I hope to do is make a table at some point I've started to showing what for each of these phthalates what data they have what kinds of studies what endpoints they saw it would be a crew summary but at least we'll have a sense of what causes the phthalates and Romans so on but it's a lot of data that making sense out of it will be a little bit of a challenge in terms of the branch versus linear or however you want to we as a committee need to make sure we agree on what's in our domain now among these that have been provided with CPSC reviews or HRSAR reviews we have to make sure that we're all looking at the same ones and that we're not inconsistent in the I highlighted in the green cells are the ones that are of particular interest in terms of the risk assessment that's either because of bio monitoring data or because of animal data or because they're covered by the CPSIA I didn't think it was that high I see I counted 16 that are highlighted but two of them are the you know because of duplicate cast numbers so I I well it almost doesn't matter any one of the tables if we look at the last page of that the light green the ones that I highlighted in light green those three on the last page not this thing here that was my first mistake the ones that are in light green yeah right yeah yeah 16 well it's 16 but DINP and DIDP have two cast numbers so it's really 14 14 valites the numbers in the multi that one column are it's the number of tox tox line hits which gives just a crude estimate of how much the database how big the database is and you can see some of them have five or less the dark green ones have minimal data some of them are high production volume chemicals and in there tables two and three give production data in this case it's domestic production which of course isn't the whole picture the interesting thing is there's a high we have production information from EPA and from a market report the years aren't exactly the same they match up pretty well but there's a few exceptions where they don't match up well and you can also see some trends I mean obviously DPHP has gone from nowhere to one of the top ones the dye to propyl heptyl it's number 33 on that list it's on the next page are so many there aren't even abbreviations to go around a lot of them sound alike I mean there's C6 to C8 and C6 to C9 and so on I'm sure there's a lot of overlap in the content of some of these mixtures Mike, but we should still be aware that DPHP is one of the deletes that is on the worldwide base is one of the deletes produced in the highest volumes oh absolutely and we do have some data on production in other countries it's a little less detail but I haven't seen anyone do a I haven't seen a summary of total production worldwide but for the most part I think no matter whether you look at Europe or Asia I think the top ones are still the top ones and the ethyl will seem to be declining a little bit definitely methyl and of course the DPHP and there's probably a couple others that are growing so we're anyway we're still filling in some of the blanks on these tables I mean these are the ones that these tables are complete the others we still have a lot of data enter the PCAM properties and in the health endpoints but it's not yet I think it's time to adjourn for lunch and we're supposed to convene again at one but I think we'll be a little bit later than that well do you want to call it one thirty or whatever