REPORT FROM THE GENETICS SUBCOMMITTEE: TISSUE SAMPLES REPORT DR. MURRAY: Thank you very much. It is, indeed, a shock when one finds one's chairman in the morning on CNN discussing the previous work of the commission and discussing it very articulately. It is not true that the commission is going to fund its work from here on in by establishing "Clones-R-Us," is it, Harold? DR. SHAPIRO: That is not true. As you know, as everyone knows, I am sure, that cloning is again in the news. It is interesting how this sort of gets itself into the news again and you have seen it in this morning's paper. As far as I can tell, nothing has changed since we discussed this last time so we will not take that issue up again at this moment but thank you for noting that, Tom. DR. MURRAY: The Genetics Subcommittee is trying to finish up its work on the issue of the use of human tissue samples. Our interest in the work was, as many of you know, was motivated by a variety of factors. One is we know that there were a lot of tissue samples out there. We had no idea quite how many. We have a much clearer idea. We have a very conservative estimate but a much clearer idea now of how many. It is well over a quarter of a billion and it may, in fact, be as many as half a billion tissues, identifiable tissue samples, in various tissue banks and collections in the United States. A second motivating factor was that these samples, which have been collected for as long as a century and were thought to be of some scientific interest, now we understand if they are properly stored can have analyzable DNA, which can offer an increasingly vast amount of information about the individual from whom the DNA was taken. Thirdly, a series of commentaries, some by official groups, some by ad hoc groups, some by individuals suggesting what ought to be done about research on such tissue samples, both samples collected previously and samples to be collected in the future, came to some very strong conclusions and in many cases contradictory conclusions. So those factors together led us to take on the tissue sample issue. What we are going to do this morning: We will hear from a couple of guests. As you see at 9:30 and 9:30 Susan Old from the National Heart, Lung and Blood Institute and then Patricia Barr with the National Action Plan on Breast Cancer but we will have till about 9:30 to begin our conversation. We will add those two folks to the conversation. We will resume it again and we will be able to continue on till around 11:30, at which point there will be comments by Harold and statements from the public. So we have around two, two-and-a-half hours of actual deliberation, and we look forward to having the members of the other subcommittee join us in these conversations. Our goal for this morning is to talk about four, perhaps five, things. Some of them will take more time than others. I am going to do the first thing and I hope it will not take long at all. I am going to describe and try to defend our decision to talk about research conducted in an anonymous -- research conducted on tissue to be used in an anonymous manner and explain how that differs from some of the ways people had conceived of these tissue samples prior to our work. Secondly, Zeke Emanuel will describe the matrix that we have been using to look at possible cases. That, I think, will take the longest time. Bernie is going to talk about the concept of community consultation. People have described it other ways, community involvement, et cetera. We have been using the phrase "Community Consultation" as an ingredient for some of the situations that we intend to encounter and we expect to encounter. Fourth, Trish and Larry, and I think they have enlisted some other folks to help them, are going to take on the issue and we actually want your help, the other commissioners' help as well, in discussing the issue of whether to have this wall through which information is sent with the identifying information stripped off. Whether to have this wall be perfectly inpenetratable, or as perfect as human wile can make it, or whether, in fact, to allow under certain, perhaps rare circumstances, people to go back and to try to rediscover the identity of people used. This was a contentious issue and we will talk about that. Trish and Larry will lead that off. The fifth thing, and I think if we do not get to that, probably I think it is a little bit less controversial but I suppose I would lead the discussion of that, is some mechanics of the consent process that we have in mind. That is our agenda. It is a very ambitious one for two to two-and-a-half hours but we will do our best. Zeke, did you want to put that up? DR. EMANUEL: Yes. DR. MURRAY: Yes. Why don't we put that up. Thank you. While Zeke is -- right. (Slide.) This will be helpful all through our conversations this morning. By research conducted in an anonymous manner what we mean is this: One problem we had was many of these tissue collections for a variety of reasons would be inappropriate to strip identifiers from the tissue held say by the pathology lab at the teaching medical institution so, therefore, did we have to think about those tissue collections as identifiable. Well, clearly in the form in which they are held by the pathology laboratory they are identifiable and need to be identifiable. What will that mean in terms of any research that might be done with some tissue collections? Well, it struck us after a considerable amount of conversation that the most important -- that it made more sense to think about anonymity in the context of the particular use of the tissue. If, for example, I have -- in an extreme case I have a series of fully identifiable tissue samples and I send them on to another scientist with every piece of identifying information stripped, there is nothing except the raw tissue, and that is what the scientist works with. It seemed odd to think about that as "identifiable" tissue. That scientist would have absolutely no way of going back and finding out from whom that tissue came so we came to a model of thinking of the tissue as used. It struck us that we accomplished the goals of protecting individual privacy and protecting individuals against potential discriminatory uses of genetic information about them if we, in fact, endorse a process by which a scientist could make a request for tissue, have the tissue sent with perhaps some information but not enough -- not sufficient information to identify the individual from whom the tissue came and to say that in that sense the tissue was being used in an anonymous manner. There is, what Zeke has labeled, an encryption barrier that is better than the metaphor of a fire wall. It is an encryption barrier which involves the stripping of a considerable amount of information. It might contain such things still as a medical history or at least the relevant points for the disease in question. It might contain information about sex, about background, about geography and some other matters but not sufficient information to walk back, as we use the metaphor, walking back across the barrier and figuring out from whom the tissue came. So if you see the phrase or a variance of the phrase "research conducted in an anonymous matter" that is what we mean by it. It is the tissue samples and whatever information is bundled with the samples has gone through this encryption barrier and it would be impossible or unreasonably -- reasonably -- I am not sure quite what the phrase is there but it would be reasonable to think that the researcher could not walk back and get the individual's identity from the information the researcher has even with the aid of other kinds of publicly available databases. That is really all I need to say. I would like to ask the other members of the Genetics Subcommittee to elaborate or correct what I have said and I would like to invite the other members of the commission to join in. David? DR. COX: So I think you have very fairly stated the attractive aspects of this approach to dealing with stored tissue samples but it is particularly relevant for the large numbers of pathology samples that presently exist and then the clinical samples that are going to come in. It does not very well take into account potential future types of research that are going to require closer and closer interaction between the researcher and the subjects. Although it does take into account subject's viewpoints from the point of view of their privacy it does not take into account the subject's involvement in the design of the research studies at all. The other aspect of it is that in many ways for researchers it perpetuates a distancing of researchers from the research subjects at a time when the whole direction of much of the science is an increased involvement in relationship between researchers and their subjects so that I, for one, have sort of a difficult time balancing these different issues and I just wanted to mention what some of the down sides of this approach are. The final point that I would like to make is that any kind of such barrier that is put up to protect people is only as valuable and only as effective as its ability to limit loopholes for people to go through it. I must say that it is easy for commissions like this to talk about how encryption completely limits loopholes and on the other hand in reality have it simply be on a piece of paper and not how it works. If the commission sets up things that work on paper but does not work in reality I am not sure that we are helping. So this is not -- I just wanted to talk about what some of the negative sides of this approach are and perhaps by considering them and having some flexibility with this we may make it more practical for research in the future. DR. MURRAY: I certainly agree with the latter two points you made as I understand them. I mean, if we recommend something which either is impossible to implement in practice or would be widely abused then I think we have not done our job well. I do not think that will be the case but we need to be cautious about those options. In the former you are concerned about the distancing of researchers and subjects, et cetera, but this is not the only model as you know for research with tissue. There is -- you can do research with tissue where the tissue is used in an identifiable manner but a requirement is then upon the researcher to get express informed consent for each -- for the particular use of the tissue envisioned is very powerful. So I think there is not a distancing -- we are looking at the -- we are looking at cases in which the researcher for a variety of reasons may not need that kind of close contact and may want to use either large amounts of tissue or tissue that is with relatively small amounts of additional information. So there are other ways of doing it. There need be no necessary link between this proposal and a further distancing between researchers and subjects in those kinds of cases where distancing would be inappropriate. DR. COX: Just one quick follow-up point. I quite agree with that except I am basing my comments particularly with my experience as a geneticist because genetics research -- in fact, we are the Genetics Subcommittee -- I am not saying that genetics, you know, is inherently different from other types of medical information but genetics is only as good as the definition of the phenotype hooked up with the genotype. So, I think, at one sort of fairly extreme end of researchers that need to have close relationship with the phenotype I think the geneticists are very much on that one end. So that is sort of why I am making my comments. DR. MURRAY: Right. Thanks, David. Alex? MR. CAPRON: I think that I would really understand this only as I begin to see what consequences you think grow from the description that has been given here but I would like to begin that process by asking you whether the category here is the one which is described on the charts that we were given as samples that are to be used in an individually anonymous manner. Is that correct? DR. MURRAY: I have not seen this morning's version of the chart. That is my understand, yes. MR. CAPRON: If that is the case I want to suggest very strongly that we consider separating into two categories what you seemed to have lumped into one. As I understand it, there are any number of -- DR. EMANUEL: Alex, maybe I could explain the chart before we already divide it and we break it apart. MR. CAPRON: Well, let me make my comment first because I think if -- you can respond to it. As I understand it, there are any number of situations in which researchers are interested in samples which have no identifiers at all on them, that is to say if they are looking back at the PKU samples and they have 10,000 of them and they just want to ask is another -- what is the prevalence of another gene in the population of babies born in Denver in 1996 or something. They do not have to know anything about it and that is truly an anonymous sample. It seems to me what you have described here is something that would be more correctly described as the use in an encrypted manner and to lump together something which is anonymous with something which is encrypted is to me a basic mistake and I would be very disappointed to see us move in that direction. I look here and I see -- this is a problem with graphics rather than having text -- what I see here is something which says it has a sample and on one side it has the name on the sample and in another one it has a number, and then it has something called "medical record," which on one side has a name on it and another has a number on it. Now you have made certain comments, Tom, and I think this is relevant to David's comments a moment ago, about what that information would be. But if we were to literally publish this chart as our explanation of what it meant to encrypt something I would say that simply underlines to me the problem with calling this anonymous research. I mean, if we have a medical record from which my name has been removed and been replaced by a number we have a lot of information and I cannot believe that someone looking at that -- a clerk working on the project who knows that I was in for a removal of a cancerous growth and you are now looking to see some other genetic factor would not be able to look at that and say, "Oh, that is Professor Capron." DR. MURRAY: That would not be anonymous, Alex. You misunderstand what we are saying. Zeke will have some comments. DR. EMANUEL: I think -- MR. CAPRON: And then, Zeke, as you do this could you explain what the results "Name-A," results "Name-number-A." What that X means and what those are? DR. EMANUEL: Alex, this is why I suggested before criticizing the boxes I thought we would explain why we got there. Maybe I can explain why we got to where we have and how this graphic fits in with the thinking because, in fact, we began exactly where you and most of the recommended statements begin exactly where you are, which is making more than a few distinctions. So everyone has -- all the commissioners have this but I am going to use some overheads so we are all on the same page as it were. (Slide.) And I want to, in part, talk a little bit about the evolution of the thinking because this is not where we are today. All right? But I think by trying to explain a little bit of the evolution of the subcommittee's thinking it will become clear why we have gotten rid of some of the distinctions. So this is sort of transitory intermediate framework that we use and then we will talk about it. Is that in focus? COMMISSIONERS: No. DR. EMANUEL: This will not do this. I apologize. DR. SHAPIRO: Help is coming. DR. EMANUEL: All right. You have it on a sheet of paper. Okay. If we walk -- just if we walk down from the top you see we have made one division here which is previously collected samples and samples collected in the future. By that we mean -- and please my fellow subcommittee members jump in if I have made a mistake or inserted my opinion over the agreement since I was not here all of yesterday. Previously collected samples are those samples that are collected before we publish the report and before our recommendations have a chance to get in to effect. Samples collected in the future would be samples collected under recommendations that would modify the procedures currently used. We thought that there were important reasons to distinguish those two. In part, one does not want to throw away what we have -- the 200 plus million samples. At one point in our deliberations we had divided samples collected for the purpose of clinical care, that is you go in for a biopsy for your care, from samples collected as part of a research study, part of N-HANES, the Physicians Health Study, the Nurses Health Study, whatever. In our deliberations we began to see, I think, that those were not tenable, those distinctions, and that we should, in fact, collapse them and treat them the same, that whether the consent procedures were different, and in many cases they are different, the requirements that we would want to put into place, in fact, are the same or similar. Then we made this distinction between those samples that are going to be used in an individually anonymous manner from those to be used so individual identification is possible. Initially we had the following three-part distinction, actually four-part distinction: There was samples which are anonymous, Guthrie Cards; samples which could be made anonymized or anonymizable samples; samples which are potentially identifiable; and samples which are going to be used in an identified manner. We have not found that distinction helpful because when one thinks through or when the subcommittee thought through the kinds of recommendations we would make under those categories, in fact, they collapsed into these. We thought one of the problems of the current debate was the fact that everyone was focused in on how the samples are stored rather than how they are going to be used because the key issue is not whether your sample is in a research study but linking the result with the name. That turned out to be the key potential where harm can occur. So the key issue is are you using the sample in an anonymous manner in this research study? Then we made some distinctions here, which we have subsequently collapsed, and I am going to talk about that in the next frame, which is there are samples which we have collected. In the past this has been true where you are looking at individual samples, there is no community link, there is no link even in an anonymous fashion, you are just say looking randomly for colon cancer genes not in an ethnic or racial or some other geographic group. Because of genetics but not simply limited to genetics as we see in some of the kind of research studies we have looked at we could imagine that there could be circumstances where even if you collected the sample in a manner or the sample was individually anonymous there might be relevant items because of the kind of sample you use or because of the kind of sociodemographic information that might have implications for a community so we began to make distinctions between that kind of research which might have implications for a community but might not in our imagings be harmful and those which might be harmful. As a result of yesterday's discussion these were -- these two were collapsed. I do not know where I have the overhead. (Slide.) So I think we are at the stage, and since I was not fully part of that discussion, this is the current -- Sally, I apologize. (Slide.) This is the current operative model. DR. GREIDER: Zeke, aren't we missing some boxes on the right-hand side? DR. EMANUEL: No. I thought at our last meeting, not yesterday's but the previous meeting, we had suggested that there was no distinction between the clinically relevant and the research but I would stand corrected. We have all the permutations here. DR. GREIDER: My recollection was that was true for the previously collected samples but not for the future samples and maybe other people can let me know if -- other subcommittee members. DR. EMANUEL: Your recollection is this. (Slide.) DR. GREIDER: Correct. DR. EMANUEL: My recollection is at the very end of the previous meeting was that Steve Holtzman -- we had suggested -- well, we can go through it because the suggestion is that the distinctions here, the recommendations we are going to make are going to be no different and, therefore, should be collapsed but this is a work in progress. DR. COX: Zeke, I would say at least -- although I was not at the meeting but having read things, the logic, the exact arguments that you make for collapsing them in the prospective or in the retrospective for me fit for the prospective too because if you can collapse them for the retrospective then why can't you collapse them for the prospective so it does not make any sense to me not to collapse them. DR. EMANUEL: Well, I think the way to understand that is to work through each of the boxes as the subcommittee did and the rationale for them. Let me emphasize what I think are the -- and I would hope that my fellow commissioners would again -- the three path breaking, I think, distinctions we have made. One is between the previously collected samples and the samples collected after the report's recommendations. The second is that the evaluation, the ethical evaluation, should be based on the use of the tissue, not on the manner of collection or storage of the tissue, because what we are interested in, and the reasons we have worries is the harms that result and that depends upon being able to identify a specific result with a specific person, and that recognition that some research conducted on individually anonymous -- in an individually anonymous manner may nevertheless have sufficient sociodemographic information to adversely affect communities. MR. CAPRON: Could you pause now because -- DR. EMANUEL: Well, let -- MR. CAPRON: -- because you think you have responded -- DR. EMANUEL: I want to -- can I -- (Slide.) To put this a little bit in a framework the current system, the Common Rule, recognizes only two categories. All right. It has nothing to say about the rest of this. This is really the reason we are here and looking at it because it is solid on all these other boxes. Now, I do not want to -- I do not know if the commission wants me to potentially jump ahead and suggest what the recommendations were or should we just leave that? DR. MURRAY: I think we should go ahead. DR. EMANUEL: Does that sound -- DR. MURRAY: Does anybody want to comment at this stage? MR. CAPRON: Yes. DR. MURRAY: To respond to Alex -- well, but I think -- okay. MR. CAPRON: May I -- DR. MURRAY: Go ahead, Alex, have your say. MR. CAPRON: Well, thank you. I agree entirely with the notion of the focus being not on the way samples are stored but on how they are used. It seems to me that there is a self-evident distinction between a sample which has no identifiers and those which have some information and are encrypted. There are two distinctions. One, the information, although it may seem to the person who is making the decision at the time is sufficient to make it anonymous may not make it anonymous. Secondly, after the fact a researcher with findings, which he or she regards as important enough, will have information which could be unencrypted. That is a fundamental distinction it seems to me and the whole notion that in certain research you need fire walls or you need one way barriers and the like because you have information which has an encrypted number on it, which if unencrypted, goes directly to an individual suggests that there is a distinction. I do not think that what I have heard thus far explains to me why you want to lump those two together. DR. EMANUEL: I think, Alex, the answer to that question is let's get through the protections we would like and see if, in fact, they collapse or they do not collapse. Right? That, I think, is the rationale that led us to collapsing them because, in fact, the kind of protections you would want, the kind of consent or IRB review that you would want for those two different categories, in fact, collapses them. They would be the same. MR. CAPRON: I read your charts before this meeting. I came to a different conclusion. It does not seem to me that I have the same sense about information being used where a person could have results of great importance to me which they could unencrypt and where there may be a moral obligation to do so in order to give me a warning or conversely where their scientific interest in unencrypting it is very different. To have a sample used in advance seems to me does not fit under your -- the conclusion that you have given about no IRB review, no individual consent, no community consent in the same way as it would with a sample about which there is no individually identified linkage at all possible. Therefore, that is one of the reasons why it seems to me that different policies must be in place. Certainly the policies having to do with whether you could under any circumstances go back through that wall only applies to information for which the identifiers are there. DR. MURRAY: That is not true. At least not in the hands of the researchers. The identifiers might be perhaps in the hands of a trustee of the tissue or even in an additional party, a third or fourth party. MR. CAPRON: If there are no such identifiers you have no basis for going back. You do not need a policy about not going back, right? DR. MURRAY: Well, I think you are not hearing what we are trying to say here. I also want to make a conceptual point, Alex, and that is if you talk to privacy experts, particularly for an issue like tissues, DNA samples, tissues containing DNA, there is not a bright line distinction between those samples that are wholly anonymous and those samples that are not. I mean, if I had access to DNA fingerprint databases then I might be able to link this particular sample even though every piece of otherwise identifying information has been stripped from it simply because I can do a DNA fingerprint from this tissue. It is really a matter of how difficult it becomes to go back from what I have in my hands, from tissue sample with or without additional information, to a specific individual's identity. It is a continuum rather than a clear and bright line. I think that helps -- that helps me, at least, to think of it in that way. So the question becomes what protections can we put in place that would reasonably assure that a person whose sample with or without other information has gone forward to a researcher and can count on that not being then subsequently identified. MR. CAPRON: Well, obviously if a person has a sample, an anonymous or encrypted sample, from you and finds out information about it and then later gets another sample from you, a genetic fingerprint can indicate that you were the source of the first sample. I totally agree and that is something that raises a different issue about genetics. I totally agree but that is dependent upon that person getting another unencrypted sample. DR. MURRAY: Get access to a state DNA fingerprint database. I mean, they are -- and privacy -- I mean, you probably know more about this literature than I do but privacy experts assure me that it is really a matter of how tightly you wish to protect it. MR. CAPRON: Are you in the state DNA fingerprint bank at the moment, Tom? DR. MURRAY: Not at the moment not that I am aware of. MR. CAPRON: All right. DR. MURRAY: But -- MR. CAPRON: So, in other words -- DR. MURRAY: I know your point but things will become more widely available in the future. I mean, we need to think not just where things are today but in the future. Bernie wanted to say something. DR. LO: I think one of the problems we are having is that we are trying to have a full debate in miniature and I think all these issues need to be -- all these issues need to be discussed and I think I just want to make two points. One, where we end up in our matrix may not be where we want to start. So at conception I think most people do come with the intuition that there are many, many more rows and columns than we may end up with. I think as Zeke was suggesting it is only if we go through the arguments and find that a lot of the rows and columns are identical after deliberation. Do we then say the recommendations will collapse? But maybe as we present this we should start with the fuller matrix and argue through why it collapses down and obviously we cannot do that in an hour-and-fifteen minutes. The other point is that, Alex, what you were saying about the importance in some situations of being able to deencrypt that information either for the purpose of reporting back to an individual patient, close but not there yet, to report back to an individual patient the findings that may be of clinical import to that patient. Or the other situation where that is likely -- that may come up is where the scientists wants to get back to that patient because they have such an interest in genetic constellation they want it to be studied more. Now whether we build that in to the model at the onset or have a simplified model which adds these in sort of as variations on policy I think we need to argue out but certainly we do not want to lose track of the points you were making, Alex, about how the fact that it is encrypted or presumably at least the possibility of unencrypting and there may be valid moral reasons for wanting to do that or compelling more reasons to do that in some situations. Obviously you cannot do that if it was collected anonymously as opposed to collected with identifiers which are somehow stripped or coded but that is something we started to talk about yesterday and I think it is fair to say that we have not quite resolved that one. DR. EMANUEL: Can I say something? If you think through these boxes there are three and only three protections, I think, that you can have. In each of the boxes you can ask the question has an IRB reviewed the protocol. I have never had a complaint about the volume of my voice. You can ask the question has an IRB reviewed the protocol, has the individual consented, and has the community in some way offered its consultation. Those are the three possibilities. If you just do the math you have got nine permutations. We have got more than nine boxes, which means that some of the boxes are going to overlap. That is just on a simple level. I think by going through each of those boxes you are going to see that the -- I mean, your view of what the kind of protections you want may be different from my view but, in fact, there is going to be some collapse there. There has to be some collapse. We do not have other kinds of protections or we have not proposed a lot of other kinds of protections. Now your moral intuition that these, in fact, initially look different is exactly why the commission, and I firmly believe why many of the other groups have come with making lots more divisions there, collected in an anonymous manner, you know, made anonymizable, et cetera. But, in fact, I mean again to reiterate I think we have come to the view because we have actually tried to work through the boxes and said, "Well, you know, the protection we think is appropriate here recognizing that there is going to be some trade offs, in fact, look the same in these two boxes and that they are not conceptually distinct." MR. CAPRON: Well, I will wait until we get to the point of looking at what the protections are. DR. MURRAY: David? DR. COX: Zeke, I think that you said that very nicely and in terms of what the motivations for the boxes were but perhaps the debate can be -- and, in fact, it is perhaps the reason why the other groups had more boxes is that they did not start with the premise that the only things that were available were those three things, those three types of protections. Now certainly in practice those are the three types of protections that are existent today. The question, I think, a third question to ask, is should we start with that premise and say that those are the three types of protections because there is not practically an option for other ones right now or should we say -- should we back up and say because things do not fit into these nine categories very clearly that we should have other types of protections as an initial starting point? I think that is a very important thing. Certainly the subcommittee by signing on to the matrix did the former but if other members of the commission do not start with that assumption then it is going to be confusing about why the matrix makes any sense. DR. EMANUEL: Hold on. That I do not think is fair to the history of what happened, David. I mean, that suggests that somehow this is Zeke Emanuel foisting this and the rest are signing on. This was a long debate of us trying to reason through what the protections are and those three, I should say, are not the three we have today. Let's be clear. We have two today. Community consultation exists no where in the Common Rule. We have individual consent and we have IRB review. We actually added permutations on those in terms of IRB administrative review, possibly a general consent as opposed to a specific consent, so we have been trying -- I think we have been trying to be innovative in the kind of requirements we are suggesting. I think this has been a long process of deliberation, you know, and one of the problems of the subcommittee framework is the months of trying to think through and argue through by using examples, you know, the Physicians Health Study or the Angiogenesis Factor of Breast Cancer Women, or some of the other studies, the kind of reasoning that we have collectively come to is hard to, you know, recapture in a short succinct manner. I mean, it may be, you know, if we want to think of some research and test out in those boxes that may be the most effective way to get everyone at the same place. DR. MURRAY: Bernie? DR. LO: Since I am congenitally optimistic I would like to suggest I think this is actually a fruitful discussion. I mean, first of all, I think as we were talking yesterday about planning the outline and the drafts of the report, I think here we have clearly seen that we need to separate out our recommendations in terms of our final matrix from the intuitive matrix most people bring to this and to sort of lay out in an earlier chapter all the considerations that lead to different rows and columns which, I think, we intended to collapse down in the draft that we saw yesterday. The second issue is one of maybe we should readdress the issue of are there other types of protections other than just IRB review, consent and individual consent and community consultation. I think there are other things out there that we should think about. One is sort of a national review body beyond IRB review, sort of a RAC model if you like, with all the pros and cons of that. Secondly, we have played around with variations on IRB review and I think in addition to administration review and full IRB review there may be categories that segregate out as exempt from IRB review because people have gone through enough studies to realize that these do not really require anything more than, you know, what now I think are termed exemptions under the Common Rule. But I think this discussion to me is valuable in that it makes me realize we need to articulate better the rationale for collapsing down the matrix in final recommendations and also forcing us to rethink are there other kinds of protections that would give us even more permutations for the different boxes. DR. MURRAY: Harold and Larry? DR. SHAPIRO: I have just a simple -- I think it is a pedagogical suggestion. It does not enter into the substance of this argument but I found it helpful and just pass it on. I found it helpful in looking at these various possibilities and matrixes to organize it somewhat differently, which gave me more flexibility in my thinking, namely I would put along the top "possible protections," and they define all the rows. And then -- excuse me, they define the columns. Excuse me. They define the columns. And then down -- but to define the rows are just differences you would want to make, whether you want to use the differences you have or additional ones, or add additional ones. And that all will enable you to keep in front of you easily protections on one side type and type of experiment or something on this side. You may or may not find that useful in dealing with this. I have found it useful in my own work now. DR. MURRAY: Thanks. Larry and Steve? DR. MIIKE: I think the purpose of a body such as our's is to get down to the elemental considerations and then it is for others to put permutations on them. So, I mean, I think that is a fundamental reason why I would say that we want a simple model and then you argue about the distinctions between them. So if we start with a matrix that is so complex that nobody can understand what the underlying basic rationale is we will never get anywhere but if you start -- but if you end up where we, as a subcommittee, currently are and then you can argue the permutations around that like Trish and I were doing I think it is clearer to others. Then, finally, I think if I remember my math, the magic number is seven plus or minus two and most people cannot remember anything beyond that. So we are in that magic circle right now. DR. MURRAY: Steve? DR. HOLTZMAN: I guess I have a very simple view of the point Alex is raising and thinking about our deliberations, and that is what we care about is the nature of the protections, the nature of the processes that will go along with the research being done or not being done. So, Alex is simply pointing out something we started with as well, that there is a distinction between samples where it is logically impossible to connect them to an individual, samples where they are connected to the individual in the research and in between ones where it is physically difficult but not logically impossible. The question -- where the rubber hits the road the question is are your protections different, are your processes different? We concluded that with respect to the logically impossible and physically very, very difficult the protections would be the same, the processes would be no different. So, I guess, what I am saying, Larry, I would start with the more complex conceptual scheme because it is out there in the literature and explain why we have reduced. I think there is a reasonable discussion to be had, and I think Alex wants to lead that, that says he feels either that there are three different processes, three different levels of protection, or he wants to collapse the physically difficult into the same as being identified. MR. CAPRON: Tom? DR. MURRAY: Alex? MR. CAPRON: I think that is fair and I like Harold's way of going about it. It seems to me that the matrix we are talking about is a three-dimensional matrix and the dimension that has not been mentioned so far is what risks is a person exposed to in any particular situation. What types of risks? For example, the risk that someone knows something about me that I do not know and the risk that, therefore, I will come to harm, that was preventable if only I knew, or the risk that I will have a knock on the door with someone saying we would like now to get more information about your current health status because we have found something about you genetically that you did not know and we did not know until we did this study. Those are different situations. As you say, there is a situation in which it is impossible and another situation in which it will happen, and another in between. So it seems to me that it is not just talking about what the protections are but the reason to lead towards one protection or another is going to vary depending on what you see the risk is. Let me follow this through in another way. I think I am persuaded by your decision as to previously collected samples to collapse those which were collected for clinical reasons and those which were collected for other research purposes, obviously not for the current research because otherwise it would be a prospective study. That is not the intuition I started with and it is not as though every division and every distinction I think of I follow through to suggest we have to show it. The reason being is if you take that risk approach it seems to me it is very likely that the things that would concern people would be the same whether or not their tissue had been taken out as a result of a diagnostic or therapeutic procedure or some unconnected research and that is not the intuition I started with. I should note, however, I think we need to address that with some care because in the first chapter in giving the overview you note quite correctly that most people whose samples are among these hundreds of millions that are being stored do not know that those samples are stored because it was not an explicit part of the consent process and it was never focused on. That is not true for people whose samples were taken for research. They at least know samples were taken for research, they do not know about this research, they may not know how long it is stored but at least they know that someone took it to study it. Now that is a distinction. But if we look to the future and say not what I was originally concerned about, sort of the dignitary difference between having something done to a sample you do not even know anyone has versus the other, but what risks you are going to be exposed to. I can understand why you ended up collapsing those. Just to show that I am not totally pigheaded, Tom, I can understand why. But I do immediately when I think about the risks see differences so I will wait and see whether the collapsing that you have done, which apparently has been done as to future samples in different people's view. I mean, Zeke thought you collapsed research studies and other samples of the future. Carol thought you had not and you were still keeping them separate. So maybe even the committee is not quite clear where its matrix goes. Certainly in looking at the materials that were distributed in advance I understand why Zeke came to that conclusion because it seemed to me that the 1b and 1c looked very much like 1e and 1f. So I understand why that would have happened, Zeke. But again there may in the end be there some difference in how we think about people knowingly encountering a risk. So I want us to -- when you lead us through this -- address this question of what different risks you thought were at issue and why you think that treating different types of study the same way is right and why these three levels of protection -- the third level by the way, of course, is in the CIOMS documents and so forth. It is not as though we thought up community consent but it is there. There is that recent article that was in Nature Genetics that you have probably seen by Foster, et al., which addresses that process. So, I mean, I think it is a worthwhile concept to bring in but I do not think that the fact that there are only "three types of protection" means that the level of risks that are involved are the same for all and, therefore, we would invoke the protections with the same expectations of need for using them. DR. EMANUEL: I think, Alex, your suggestion about the three-dimensionality of the framework is absolutely right and that is why the boxes are, you know, in some sense -- while risk is an important consideration, the way you take care of risk, what you do about it, how you operationalize it in terms of protections, that is what we have put in. So your thinking and my thinking are exactly parallel and I think what we are seeing here is the question of, in fact, when we think about the kinds of research that are going to fit into these different boxes what are the levels of risk that might be involved and part of the problem is at least at the moment we do not have actually concrete research studies. One of the things the commission did do is to look at some of the studies that have existed that have worked with these kinds of samples and talked about what we thought some of the risks might be. One of the problems we have is there is a big long future out there and it is very difficult for mere mortals, especially some of us who are very distant from the lab, to imagine everything that is out there and, also, imagine what might come about but we have to do our best. Again, I do think that -- I mean, I would just mention that at the end here the idea of collapsing the -- in the future the clinical research and the research studies, I was the last hold out. Carol was the leader of that as I recall. You know, this is a work in progress. I do not know what you want. Do you want to go through the recommendations or do you want to go through some of the other issues? DR. MURRAY: I think the most -- in my view but I would invite the other subcommittee members, in my view the most important thing now is to sort of go through quickly the recommendations for the various conditions, the boxes. How do the rest of you feel? If we can do that -- that is the most important thing we can do. I want to also have some time to talk about the other issues, community consultation and whether -- under what circumstances you would ever walk back through this fire wall. I want to do that but I think we can do that after the two visitors join us and give their talks. Carol? DR. GREIDER: I was just going to ask a question about which version are we going to go through. This discussion that we just had it sounds like we need to go through a more full matrix version rather than the mini-matrix version based on the discussion we just had. DR. HOLTZMAN: I think we could go with the mini-matrix. I think we all know what the full matrix looks like and if we go to the mini-matrix as we articulate the recommendation we can say why we collapsed. DR. MURRAY: I agree with Steve. Zeke? (Slide.) DR. EMANUEL: I think this might be the most helpful matrix to look at for a framework. DR. MURRAY: Right. MR. CAPRON: Can you tell us where we find this in this so-called hard copy? DR. EMANUEL: In your place. MR. CAPRON: In today's -- DR. EMANUEL: Handed out today. MR. CAPRON: -- handout as opposed to the stuff that was in our book? DR. EMANUEL: Yes. DR. MURRAY: It should be this one. MR. CAPRON: These pages are not numbered. DR. EMANUEL: This one. Proposed -- it says proposed policy just like it says up there. And it has got -- because there are two things labeled and you will see that -- one -- what happened is that they have got a row collapsed. MR. CAPRON: A row which I should point out is based on this risk differentiation. DR. EMANUEL: Yes, of course, that is why we did it. MR. CAPRON: Yes, exactly, right. DR. EMANUEL: But that is exactly the way it should be. There was an assessment that these risks should be -- I do not want to speak because I was not there at the collapsing but as I understand it that the risks, in fact, were something that ought to be determined by the IRB and not prejudged but whatever. I mean, someone else could speak to it more intelligently. DR. SHAPIRO: I think in community consultation. DR. HOLTZMAN: Well, to Alex's point about a third dimension about risk, it could have one of two components. The first has to do with the identifiability and I think what you have said, Zeke, correctly is if that is all you mean by risk the third dimension collapses entirely into what are the nature of your protections. On the other hand, if you want to start making risk distinctions based on the nature of the research then you do have a third dimension where you might then start to make differences in the kinds of protections. Why we collapsed the community from nonstigmatizing to stigmatizing is we made the determination that if a community is implicated that one ought not, other than by going to the community for consultation, prejudge whether or not it would be stigmatizing. (Slide.) DR. EMANUEL: Look at the box labeled "to be used in an individually anonymous manner" and "individual, no community linkage" for a second and let's -- my paradigm, and it is only my paradigm of the kind of study that this involves is to think about a paper that I passed out on tumor angiogenesis study where people at the Brigham hospital went to -- got samples of women who had breast cancer lumps removed five and ten years previously and were then looking at a new -- not actually genetic test but a new kind of test to make predictions about who would have recurrence or who would die from their disease. They took 104 samples completely anonymous -- I mean, to the researcher anonymous but obviously to the pathologist who drew them out and wanted to correlate clinical information with the tissue sample. So that is a paradogmatic case, I would think, of that box. There was no interest in identifying ethnic groups or racial groups or other groups. Okay. So the question is these women did not consent to this research when they came in for their biopsy. They may have signed a general consent that their samples because it is a teaching hospital would be used for education and research purposes. So what kind of risk do they face and what kind of protections do we want to put into place was the question. In the boxes you can see the recommendations that we are suggesting, that the IRB -- I mean, it should be said that under current proposal, under current Common Rule guidelines no IRB review for this necessary and no consent necessary. At least that is our interpretation of the guidelines. MR. CAPRON: Could you explain one aspect of the research? The researcher here is the geneticist, is that right? DR. EMANUEL: This actually turns out not to be a genetics study, which is I think relevant. Not all of the studies that should be -- I mean, we have not emphasized that but this does not only apply to the genetics. MR. CAPRON: Right. But where genetics -- our task starting off with was to look at this from a genetics point of view. Let's try genetics for a second. DR. GREIDER: We have debated that a lot. MR. CAPRON: Who is doing the study? I mean, it is not the person who holds the samples. DR. EMANUEL: That is the pathologist. No, it is a researcher, a separate researcher. MR. CAPRON: A researcher. He or she is looking at the tissue sample for some reason? DR. EMANUEL: Right. MR. CAPRON: Right. And what he or she has done has gone to the colleague in pathology and said, "Can you send me 100 samples of women who came in and had biopsies taken and who had X, Y, Z disease," is that correct? DR. EMANUEL: Yes, that is right. MR. CAPRON: And he gets the 100 samples and they are labeled one to 100 and -- DR. EMANUEL: Right. MR. CAPRON: -- and the pathologist does not keep a record of which people those came from. DR. EMANUEL: Well, even if he does I mean we can play it through. But say he does keep a record for the most extreme case he keeps a personal record. I mean, one of the reasons for talking about the encryption barrier is to say that there is not -- you cannot walk backwards. MR. CAPRON: Well, encryption -- with barriers you can walk backwards but if there is anonymous samples with just one to 100 and he does not keep it you cannot. I mean, if he later -- now what we are -- it does seem to me that the genetics aspect comes in here. Suppose that what the researcher is doing is not asking for the medical record to find out about the sexual history or the gestational history of these women but is instead asking is there a gene here and he looks through these and he says, "In this group I get 88 of these women have a gene," and he goes out and he says to the pathologist, "Send me samples from 100 women who did not have this cancer." The pathologist sends them and he does not find the gene in any of them. Now at that point if the sample is totally anonymous and he says, "I have got to tell these women something," the pathologist will say, "Sorry, there is no way I can. I just sent those out to you. I put numbers on them. There is nothing you can do." If he says to the pathologist, "I have got information that may be of relevance to those women and their sisters, and their daughters," and the pathologist says, "Oh, well, if that is really that is important I can -- we can tell those women to come in and see you because we have now found out which of them has this gene and they can then make contact or give us the names of people we should contact." Now to me those are different situations. Facially obviously different. It is a whole different set of considerations that should come in. DR. GREIDER: Can I make just one point, which is what you are also making -- not making and distinguishing -- is research and clinical care. Just because a researcher finds a particular mutation in the gene does not necessarily mean that becomes the norm in clinical care and that those people need to be told something because of one particular study. MR. CAPRON: I realize that in the lab he has it is not a CLIO approved lab and the results may be inaccurate for that reason. I mean, I know the difference between -- DR. GREIDER: So the question is would you want to walk backwards under those circumstances? MR. CAPRON: Exactly. In other words, I am not saying that the response should be to walk backwards. I am just saying that the possibility of having results which would cause the researcher either to say I want to know more about these women -- I mean, is it, in fact, this gene that I have found or is that the gene that causes them all to be great pianists. What I am really looking at is the coincidence that they have that gene and they were all living in an area that got irradiated in the 1950's and no one realized it or they were all drinking the milk or, you know, whatever the reason or some other factor. I am looking at a totally spurious unconnected thing and I do not know it. I need to know more about those women. So whether it is a therapeutic impulse on the researcher's part or a, gee, I need to know more about these women now to know whether this finding has any significance. The answer in each case may be we have encrypted it and we have encrypted it for the reason that you should not have access and we are not going to give you access. That is a possibility. But we all know human nature and we know that there is at this point the potential to say there is a good enough reason to do that. The situation is different in these two situations on the face of it between the encrypted and the totally anonymous cannot be linked, you know, you did not get anything other than the fact you got a sample that started off with a diagnosis in the category that you wanted to research. DR. EMANUEL: But, Alex, I would go back -- I agree with you 100 percent. Facially they are different, right. One you have the potential if, in fact, you kept that sheet of what number one really means of going back or -- I mean, even if you actually ripped up the sheet if you are in a pathology department with enough work you could actually go back. It is not like you cannot go back just because you have ripped that sheet up. So now the question is what kind of protections, Alex, would you like in place, how high do you think that risk is and what kind of protections do you think should be in place before you -- to do that study? Now traditionally in this country we have said you do not need consent for that. It is an existing tissue. You do not have to have informed consent. That is what the Common Rule says. If you get consent to go back to that 104 you may face lots of problems to do that kind of research and the longer back you want your samples,the more clinical follow-up, the more difficult it is going to be. People will die. People move. America is -- MR. CAPRON: Right. DR. EMANUEL: It is a very difficult place as opposed to other countries. MR. CAPRON: Right. DR. EMANUEL: And the question you have, I think, is how high are the risks to these people, what kind of protections do you want to put in place, and while there is this temptation to go back and forth can you create a system, devise a system as we have been thinking about of encryption without going backwards or going backwards under certain procedures that satisfies you that, you know, you have lowered the risk to a reasonable level. You are not going to lower them to zero. As Tom says, even with Guthrie Cards to the future you may not lower anything to zero because we are all going to have our DNA sample on a micro chip. MR. CAPRON: Well, then the question then is we are now talking about a protection that is not one of your three. It would be a protection that would say it shall be unlawful, it shall never be done, or it will say it shall never be done except when the following extreme circumstances are made out. I mean, I assume your committee has talked about what that except when will be and I am eager to hear it. So, I mean, maybe we should at some point get to that point. But then we have a need -- in the category of something that is encrypted, we have a need for that policy. Because of that we are placing a person at risk that they will get information which they may not want or other people will have information which is potentially accessible to other third parties. I mean, let's talk realistically about nothing is totally confidential here or we talking about the need in all of these things to talk about what problems will come to a researcher who does not keep the information which he has managed to make un-anonymous from others. I mean, it is an enormous difference for a person then to learn that they are at a genetic risk which they did not know they were at when they go to fill out their next life insurance policy. Have you ever been tested for X, Y, Z gene? DR. MURRAY: Excuse me, Alex. I am just concerned about time. We actually -- the issues you are talking about right now are exactly the ones that we, ourselves, are wrestling with and want to talk about the circumstances under which, if ever, we would want to say there would be even a very rigorous procedure by which you would ask permission to go backwards and find out the identities of persons. MR. CAPRON: You have not come up with -- DR. MURRAY: We have not made a firm and full decision. We are fully aware of the kinds of risks you have talked about but we -- I mean, we want to have you involved and everybody here involved in that conversation which Trish and Larry and others are going to lead but I want to do that a little after 10:00. So if we could just allow Zeke to go through the rest of the boxes and the rationale for them. MR. CAPRON: Fine. DR. MURRAY: Thank you. DR. EMANUEL: I mean, I think that you have a box here that sets a standard policy recognizing that there may be exceptions. The issue is what is the standard operating policy without extraordinary circumstances? Let me distinguish going down here in that first column from a situation where you might want to implicate a community or your research might want to implicate a community. So, for example, you go through a Tay-Sachs bank where the samples were collected, you know, might now be completely anonymous, maybe the samples had no information but the results could have implications for a community. And in that situation we suggested -- again that is something actually the current regs do not recognize. We have made some suggestions here. MR. CAPRON: You have gone through the word "consultation." DR. MURRAY: That is the working term right now. I am sorry that "consent" even appears there but consultation is the current. MR. CAPRON: That is good. DR. MURRAY: Yes. DR. EMANUEL: But what appears here is individual consent and community consultation. MR. CAPRON: In the previous box. DR. MURRAY: Yes. DR. EMANUEL: Oh, I am sorry. I did not correct all of those. Sorry. DR. HOLTZMAN: It is supposed to be community consultation. DR. EMANUEL: Consultation, yes. MR. CAPRON: Right. DR. EMANUEL: I apologize. The reason I did not do a search/find replace is because it also appears in the individual context. DR. MURRAY: Yes. DR. EMANUEL: We have distinguished these from cases where in the nature of the research you need individual identifiable -- so, for example, family pedigrees is the paradogmatic case in this situation or you have particular samples where you keep going back to a person and get either more sample or different kinds of sample, or do additional tests. So it is just an individual basis. In those two cases the main -- the main difference here is full IRB review and full informed consent because it is potentially individually -- I mean, the researcher knows who that individual is. The researcher knows. It is done with a specific identification. Even in the family pedigree where all you know is daughter number five, you know, daughter five-years-old, you have the potential to clearly identify them. That is the previously collected samples. The ones in storage now either from research or from clinical care. Now the future offers us opportunities -- Sorry, Steve. DR. HOLTZMAN: Can I just take one step back? DR. EMANUEL: Yes, please. DR. HOLTZMAN: Just to explicate some of our reasoning. We did start by saying with respect to extant samples that maybe there was an in principle distinction between those collected and the research context of those versus those collected in the clinical context. So we then asked ourselves so how would that play out and why would it be different, and I do not have to rehearse the arguments for even why they are different. But what we concluded was that the collection in the clinical context there was essentially no consent for future research. In the context of collection in a research context, even though they had agreed to engage in a research enterprise, they had not engaged in the consent of future research enterprises which were not envisaged and so that was morally no different than having not consented. So, therefore, we collapsed those together and said, "What are the levels of inappropriate consent that would be important?" We said, "With respect to if it is anonymized that no consent was necessary because there was not the potential for harm." I am not going to get into about the wronging aspect here. I am just stating the conclusions. And that with respect to if it was going to be research in which they would be identified consent was necessary because even if there had been a general consent to future research it was not logically possible to have been an informed consent because they could not have made an assessment of the risks, harms, benefits, et cetera, to research that had not been envisaged. DR. EMANUEL: Thank you. That was excellent. In the future the main difference is that we can change the consent process for clinical collection as well as research collection and these are not settled categories as you heard. But if you talk about situations where people are coming in for clinical care and there is no plan or known research to be associated, again we could divide these into two different categories. I think generally one should identify that any time you are going to -- any situation where the individual is going to be identified we have agreed there should be full informed consent or potentially identified, even if in the results they are not going to be identified, but if someone could walk back either because it is a rare disease or the way the pedigree is laid out. Two, samples that are to be used in an individually anonymous manner, the issue here is what kind of individual consent should there be. And out there in the debate there are some people who want the current system, no consent or the one line that is in the sort of general consent when you come into a hospital. There are some people who want a full informed consent down to, you know, I give permission to this specific researcher to do this study but to no one else. Contrary to what is written here I think our general view is there should be a general consent for research or a general consent to have their stuff not for research. We have tried to work through some general -- what those consent forms would look like and I think, in general, they turn out to be very difficult. The one we have from the Breast Cancer Coalition is specific to breast cancer. The problem is if you try to make it more general for anyone coming into the hospital or something like that you find some difficulty. DR. MURRAY: Zeke, can I make just one point that would go for all of the -- particularly the ones collected in terms of clinical care? DR. EMANUEL: Right. DR. MURRAY: Whether previously or now. If there is on the record that a person did not want their tissue used for research that preempts any possible use. We do not -- we did not specifically note that in this table but that should be noted. DR. EMANUEL: The other thing to note is that we heard from Bartha Knopers that in Europe or at least in the Netherlands they were going to a presumed consent with an opt out. For our reasons we had thought and discussed why that might not be good and it might encourage sloppy record keeping if you could not identify a record and other reasons. So I think, in general, we are moving to having the general consent process and we had thought through some of the problems and difficulties because we had heard from some of the people in our mini-hearings about the fact that they do not remember even signing a consent form and they felt coerced, et cetera. Now without going through each of the boxes, I mean we can again try to discuss -- MR. CAPRON: How does the general differ from the present situation? That is to say -- you know, the assumption is when I went into the hospital last year I signed a consent form that allowed general use of tissues. I mean, there was some language that was not brought to my attention but it was there. Is that what you are thinking about? DR. EMANUEL: No. MR. CAPRON: You are thinking about something that goes well beyond that. In practical effect -- DR. EMANUEL: Let's say -- let's be clear. One of the reasons we are not going to full informed consent is because in many circumstances when you collect your tissue you -- MR. CAPRON: You do not know what the study will be. DR. EMANUEL: We have no idea what the study -- and we do not want to tie hands today for studies that might -- we might want to do fifty years from now or whatever. So the issue is what kind of consent can you have? Is signing a piece of paper where there is one obscure line -- and actually some of us have looked at some of those lines. They are not nearly as good as you would like them to be currently. So one issue is to make that much more explicit to people to bring it to their attention to think about ways in which they might be alerted. The other question is what kind of check offs or limitations can people provide? Wanting to be recontacted for future involvement in studies. Wanting to limit it to certain diseases. DR. MURRAY: Right. Steve and David? DR. HOLTZMAN: Just to walk through a little bit of the thinking and here it may be more my thinking than the subcommittee's, I think. I start in the research studies box of tissue to be used in the future in an anonymized manner. There is the case where we are not saying you just get a -- it is -- since you are in a context of a researcher describing the specific research the question is whether or not you can get an open ended consent to future at this point unenvisaged studies. DR. EMANUEL: Yes. DR. HOLTZMAN: Some have argued in the literature that that cannot be informed consent so, therefore, you should not be able to do it. You can only use the sample for a certain study and then you have to go throw it out, et cetera. I think what we concluded was that it would elicit to obtain such an open ended general consent to research in conjunction with the specific consent to the specific research provided that that open ended referred to future studies conducted in an anonymized manner. Okay. So that is how that box comes about so even though it says general what we mean is the specific study consent plus an open ended general consent with the opportunity there to say but not research of this nature or not research of that, that it cannot be given to a commercial firm, it cannot be used for whatever because you are in a research context. So the question about collapsing them is now when you look at that which is collected in the clinical context clearly there is no research protocol you are describing. All right. The principle of if there can be general consent to open ended studies is on the table. All right. We think it can be. All right. We think provided again that it is conducted in an anonymized fashion. The question then becomes to some extent a pragmatic question about what level of detailed and consent one can engage in and should engage in, in the clinical context where we have heard much discussion about being sensitive to the patient. All right. That is the last thing on their mind when they are going into the biopsy is about research. And weighing the -- let me call it autonomy rights of the individual which we give very robustly and fully when they are in the calm atmosphere of a research context versus when they come into the clinic and weighing that against the potential for those samples just not being available whatsoever. I think that is what we struggled with here about whether or not to collapse. DR. MURRAY: Thank you, Steve. That was very helpful. DR. COX: So to go one step further that was the reasoning. So in my simple minded way what is the punch line? Zeke has filled in the boxes so what is the punch line, big picture punch line if we look at those boxes. First we have a ton of samples that were previously collected. Can we use them or not? Can we say -- okay, even though those were not -- and that is both from the point of view of research and from the point of view of the clinic, the -- if we are going to use them with identifiers we have got to go back and get consent, full consent. If we are not going to use them with identifiers, okay, it is okay to use them even though we do not have the consent right now. That is what we are saying. Plus or minus the community interest. What about in the future? What is different about the future? There is only one thing different because using it with identifiers is no different whether we are doing it in the future or we are doing it later. We have not addressed using things with identifiers. Okay. Because we have not -- we have addressed it but we have not made any distinction. The only distinction is in the future if we are going to use things anonymously, okay, we get a general consent from people. We get that general consent whether they are in research or whether they are clinical. So what we have done is said the thing that is different in the future right now is that the Genetics Subcommittee is coming down as saying that we agree that there should be some general consent even if things are being used anonymously. What we have not done, okay, is changed anything about the status quo from the status quo of things that are being used with identifiers. That is the way I read the way the boxes are filled in and what the summary is. DR. MURRAY: Zeke, do you want to respond to that? It is 9:30 and we are supposed to go to the next item but I want to give you a word here. DR. EMANUEL: I think that the -- I mean, there is a balance here and what -- I think David's general picture is right but we recognize that there are going to be some exceptions and tough cases. The one of you find something either serendipitously about someone and you might want to walk backwards. How -- there is also the policy issue of how detailed that encryption or how that rigorous that encryption barrier is. And I think those are important issues. We are trying to create a workable policy again which can be implemented by IRB's throughout the country and -- because I think realistically we are not going to have -- these are not the kinds of studies that you are going to have a RAC-like -- because there are going to be hundreds of them throughout -- if not thousands of them throughout the country. We also -- I think my final comment is we need to -- while genetics is here everywhere, I think my own reading is there are going to be just as many studies that are not genetic and we need to be very clear about that. There are a lot of immunology studies. There are a lot of studies of new factors that are not at all genetic. So we need to be concerned about -- in some sense the genetic ones raise some issues because you can have a genetic fingerprint but the policy has to be broad to cover all the items. Now, I mean it may be a worthwhile intellectual exercise to say let's look at the Guthrie Cards and let's look at the pathological samples. What are the kind of different protections you would like, Alex, or you think might be in place there? Would you -- and here is where -- would you in the pathological -- in the case of the pathology samples want to have full consent because -- on existing pathology samples because that is what I think would be required? I would just say that in my reading of the literature no one has suggested that. MR. CAPRON: Why isn't that a matter of the choice of the subject? In other words, looking at the breast cancer documents and the -- DR. EMANUEL: In existing samples? MR. CAPRON: No, not existing samples. Excuse me. Future samples. DR. EMANUEL: But that is what we were talking about. In existing samples -- MR. CAPRON: The existing -- you had moved forward to the -- DR. EMANUEL: Well, you raised the objection in the existing samples first. So let's just talk about the existing samples. I am going -- I have a pathology lab. It has got hundreds of thousands of samples. MR. CAPRON: I would say those can only be provided on a truly anonymous basis. DR. EMANUEL: I ask you what does truly anonymous mean to you separate from the -- MR. CAPRON: It is not encrypted. Anonymous. It means that there is -- that you are getting samples that -- DR. MURRAY: Could sex go forward? MR. CAPRON: Excuse me. DR. MURRAY: Male or female? DR. EMANUEL: Could any clinical information be attached to the sample? DR. MURRAY: Age? Or could go nothing when you said nothing? MR. CAPRON: I would say that if there is a clinical category -- DR. MURRAY: Disease. MR. CAPRON: -- and you are asking for a group of samples of males or females or people within a certain age range you can get a group of those samples but among the samples individually there is no encryption. DR. MURRAY: We are going to have to come back to this because I think I disagree pretty strenuously. DR. EMANUEL: I am not sure what you mean. I am not sure what you mean by -- MR. CAPRON: Your example of the 108 breast cancer or whatever samples. DR. EMANUEL: Yes. MR. CAPRON: All that you wanted was send me your breast cancer -- DR. EMANUEL: No, no. With attached clinical information but not identifiers. That is what we are distinguishing. No social security, no birth date, but age, clinical course -- whether the -- I mean the essential information being whether the cancer recurred or not. DR. MURRAY: I think we just need to think about this one because we will come back to this but I do not -- we have guests here and I do want to -- we have to let our guests speak and I have a feeling the issues will come up as they speak so it is not like we are completely suspending this conversation. I know Rhetaugh had her hand up and I want to give her the last word now and then we are going to turn to Dr. Old. DR. DUMAS: I think I might be missing something but my lack of understanding might be useful in this case. It seems to me our basic principle is informed consent and if we have existing samples and it is possible to obtain informed consent isn't the question how to obtain that consent and if there is no way to obtain informed consent then there needs to be some statement about exceptions. Am I missing the point here? DR. MURRAY: Well -- MR. CAPRON: The exception is that they want to be able to use the samples without any consent. DR. DUMAS: Well, if it is not possible to get the consent but I am not hearing that the basic over arching principle is informed consent. And either you are able to get it or you are not. Now if you are not able to get it then you have to talk about the conditions under which you would be able to use the sample. DR. EMANUEL: Well, let me say I think informed consent for some research studies -- one of the reasons we made the distinction between using it in an anonymous manner and used in a potentially identifiable manner is that in an anonymous manner you are not linking a result with the person's sample. The second thing is the issue here is it is impossible to identify the person or is it possible. This again is going to be a big spectrum. The question is how much effort is necessary to do that? Remember in this -- many of these people will already be dead. Many of these people -- you know, 20 percent of Americans move every year. Outside of a research setting where you are tracking them for some reason it is enormously difficult and you are not interested in the particular person. I would also say, Alex, on your remark it is not us. I mean, the current policy is no consent. Let's be -- DR. DUMAS: No consent? DR. EMANUEL: That is the current policy. DR. DUMAS: All right. Well, then -- DR. EMANUEL: Because this is existing data. DR. DUMAS: All right. And I will hold my comments because we are going to talk about this again but I really would urge that we put as our primary focal point informed consent and how to obtain it. DR. MURRAY: Steve? This is going to be the last comment from a commissioner before we move to Dr. Old. Steve? DR. HOLTZMAN: Just to maybe lay out a little of the thinking here. This is not necessarily the subcommittee's thinking. I think it is general thinking about informed consent. Clearly if you can have it reasonably happen you want to get it. Then the question is why is informed consent important? There may be two elements to that. Simplistically the autonomy right of the individual as well as the protection and the potential harms to the individual and so then when you look at the extant samples, all right, you then ask the question pragmatically the value to society of doing research versus the cost and difficulty of going back and getting the consent and that if you protect them against harm by anonymization or conducting the study in an anonymous manner, all right, that that protection against harm plus the value to society outweighs the autonomy interest. I mean, bottom line I think that is the argument. DR. DUMAS: I think you are getting to the whole issue of who makes that decision. DR. HOLTZMAN: All right. DR. MURRAY: Dr. Susan Old has joined us. Thank you for your patience and thank you very much for taking some time this morning to come speak with us.