DR. SHAPIRO: Okay. We’re going to have to really end this part of our discussion. I know there is a number of others who want to speak, Alex, and so on. But I really — if you could hold your questions, we’ll be coming back to this. This is not the end of our discussion today. But we do have some guests here, and they have been waiting, and I would like to turn to them, so that they can present their material to us. Lisa Brooks and Mark Guyer are both here, and I want to welcome them. Please come forward. Lisa Brooks is a Program Officer, the Genetic Variation Program, the Genomic Informatics Program; and Mark Guyer is currently Associate Director of the Scientific Coordination of the National Genome Research Institution. We are very grateful for them both coming here today. They are going to talk to us about plans to develop a resource for discovering human genetic variations, with special emphasis on — and I’m going to test which one of the Commissioners understands this—single nucleotide polymorphisms. You can all write me about those later. Welcome, it’s very nice to have you here. Thank you very much for taking the time to come. We appreciate it.

DR. BROOKS: You should have two handouts. One is the research —

DR. SHAPIRO: Is there no chair over there for you?

DR. BROOKS: I’m going to be standing by the overheads. One is a resources for finding human genetic variation, which is what my brief talk is an outline of, and the other one is the consent form that we’re using. I’d also like to introduce Diane Wagner from CDC. We’ve been working with them about the use of this resource.

DR. SHAPIRO: Just hold on a second. Let’s be sure that everyone should have a copy in their place. It was passed out this morning. If someone doesn’t have a copy, maybe you—Trish, you don’t have a copy? Let me see if I can find another copy. Anyone else? All right. We’re all set. Thank you very much.

DR. BROOKS: Okay. We were asked to present a case study. So, we are going to call on the box, on the top left, the nonidentifiable case. And this is arising—it’s coming from the context of trying to map the genetic basis for various diseases—that the methods for finding the genetic basis for diseases that are caused by single genes are fairly straightforward. So, that’s something that’s happening now, has happened in the past. The methods are fairly clear for how to do that, and current genetic maps are sort of more or less adequate. But it’s turning out that many diseases of interest, many common diseases—heart disease, cancers, asthma, psychiatric diseases—are called complex traits. They are effected by multiple genes, and they are also effected by environmental influences. And to find the genetic basis for these diseases is really a much more complicated process. In order to do this you really need much denser map, in order to find those things which do contribute to these traits. So, it’s turning out that there is now new technology. What’s driving this now is there is new technology, in order to find what are called single nucleotide polymorphism (SNPs). You have may have heard SNPs on chips. Well, that’s what we’re getting towards here. These are single places in the DNA sequence where there is a variance. For most individuals in the population, for example, they have an A, where some individuals may have a C. So, there are single point changes in the DNA sequence. And new technologies are coming out that are making these much easier to find. It’s important to recognize that the genetic differences we’re talking about are—most individuals, if you compare two chromosomes — or most individuals are about 99.9 percent the same. We are the same species. We do things pretty much the same, but about in one in a thousand cases there are differences. For the single nucleotide polymorphisms, the SNPs we’re talking about, there is a minimum of 6 million in the human genome; maybe as many as 30 million, or even more. There are lots of these differences, so that we’re aiming to have a map with at least 100,000 SNPs, which will enable the mapping of these complex traits much better. Now, of these single nucleotide polymorphisms, there are sort of two types to keep in mind — those that actually cause a functional difference; those that are actually the variants that cause or contribute to some disease. And, of course, there is a lot of interest in this in finding the actual genetic basis for the disease; drug companies, who are, of course, interest as well as researchers. But even those variances—and most of the variants are not going to have a direct functional effect, but the rest of those variants are very useful for mapping the ones that actually do have a functional difference. NIH now has a Request for Applications out. This is joined by 18 institutes. There is very widespread interest at NIH in finding these—these points of variations, these SNPs. Under two points to this RFA: one is to actually find SNPs, get as many SNPs as possible, as quickly as possible; and the other is to approve the technology for being able to detect SNPs, both to find them overall, and to be able to score them in large numbers of individuals. In order to allow this—in order to allow researchers to be able to find SNPs, we’re putting together a resource. So, this is that box stuff in the top left-hand side there. We are talking about 450 U.S. residents. They’re anonymous, in the sense that all information—there is going to be absolutely no information identifying these individuals; no medical information; no ethnicity information associated with these individuals when it is in the repositories (the repository is going to be Coriell). We want to have this be an ethnically diverse resource. It’s going to be from U.S. residents. But by ethnically diverse we’re talking about individuals with ancestors from various continents: from North and South America, from the Americas, from Africa, from Asia, and from Europe. And individuals whose ancestry is from mixed continents are fine. And, of course, a lot of individuals in the U.S. have mixed ancestry. So, that’s fine. As I said, no information is going to be associated with these samples. We’re holding a very high, a very strict standard here about the anonymousness. The cell lines that we’re using are coming in two sorts of ways. Cell lines that already exist. This is where we’re cooperating with NHANES, that they have been very nice enough to share with us samples that they have. So, we’re going to be getting European Americans, African Americans, and Mexican Americans, from the NHANES samples from around the U.S. There is also we’re getting pre-existing collections of Native Americans. And for both the NHANES, and for the Native Americans, we’re going back to those individuals and asking them for their consent to be part of this study. And you have the consent form in draft. It’s not the final version yet, but it has been looked at by a lot of individuals. So, you have the draft consent form in front of you for that, which has the three pages of the consent form. And also, there are about two-and-a-half pages of additional information explaining what genetic variation is. For the new samples, we’re getting consent as those individuals are agreeing to give their samples. Now, the goals of this RFA, and of the resource in support of it, is really—the driving goal is to find the single nucleotide polymorphisms, and to place them into public domain. We’re very concerned, and you have probably heard about this, that companies are getting very interested in this, and we’re very interested in making these single nucleotide polymorphism available to all researchers. So, there is a real urgency issue here about wanting to get those SNPs out, and in the public domain, in public databases. So, the point of this resource, and of this RFA, is not to relate these variations, these single nucleotide polymorphisms, not to relate them to disease, or to ethnicity, or all of those sort of other interesting things that you will want to do, but really simply to find, to discover, these variants. So, that’s the goal of this resource, and of this RFA. Eventually, of course, the point of doing these—getting this resource, and of having this RFA, is so that later studies will use these variants that are found. These variants will be found. Technology will hopefully have improve a lot to make it very easy for individuals, so in later studies we’ll expect that people will look at patient group, say, for particular diseases, and use the SNPs that have been found in this resource, and with this—the RFA, and use them in order to find the genetic basis for all sorts of traits. But those are going to be additional studies. Are there any questions?

DR. SHAPIRO: Dr. Lo?

DR. LO: Can I ask you two questions? First — well, first of all, I’ll just make a comment. I think in the schema we’re talking about, you’re not in the upper left-hand box, but you’re in the next row on the left-hand side. Because the original NHANES does have identifying information—that’s the materials of the repository. You’re taking a sample of that and stripping identifiers.

DR. BROOKS: Right. We’re taking the samples from NHANES. We’re only taking—we’re not including any — they have, of course, a large amount of phenotypic data. We’re not taking any of that. They are sending the samples to Coriell, and codes to us with sex and ethnicity; and then we’re going to make sure we have a balanced sample, tell Coriell which ones to use; and then everybody is going to destroy any sort of coding there. But that’s NHANES. There are also samples that are not coming from NHANES, where we are not going to know anything about the individuals, except their ethnicity and sex, and now we’re going to get rid of that information.

DR. LO: First question: Technically, as I understand it, you could do this study with at least the NHANES data, without going back and getting consent under current CFR regulations. I applaud your going back and getting further consent, but if you could share with us the reasoning why you decided to do that, that would be important. Secondly, when you say later studies will relate the SNPs to medical data, are you intending that these later studies will be done on different samples than this group of 450? And what happens if one of those 450 has a very, very interesting S.P., or collection of SNPs, where some researcher says, "Boy, if we really could identify that person and get more clinical information, and maybe talk them into giving us a family tree and additional samples, we would be ten jumps ahead of the game." You’re basically saying, "No, we’re not going to do that." Even though you may have some high-powered technology that would allow you to say, "We know the ancestry. We know — we are guessing it came from the NHANES study. We could sort of rummage around it and find it. You’re saying, "No, we’re not going to do that." And, again, I personally agree with that. But was there a lot of pressure from researchers saying, "You’re giving up an awful lot here. Are you totally going to shut the door?" At least I’ve heard that there may be other situations, not in this State, in other States, where researchers at the onset may say, "Okay. We don’t need to know the people and they find something really dramatic, unexpected, say, "Well, we didn’t think we wanted to go back, but now we want to go back. This code you gave us, which was supposed to be unbreakable, you really can’t break it, if you just really devote a lot of energy and resources to it. It’s really important now to do that," for whatever scientific reason. So, if you could sort of help us understand a reasoning that led you to these sorts of policies, I think that would be helpful.

DR. BROOKS: Right. To take your questions in their order, the first one was —

DR. SHAPIRO: Why are you consenting?

DR. BROOKS: Why are we being so strict? Partly, this is — the Human Genome Project—you have all heard of that—and that’s going for the sequencing. And that’s having very strict similar safeguards about getting individuals who are going to be sequenced as part of these libraries, and having consent for those individuals—even though they are also anonymous. So, it’s following that same sort of very strict standard — partly, because this is a very visible project, and we want to really be absolutely ethical, absolutely unassailable — you know, "gold standard" sort of things—without—we are aware, and we are trying not to make this become the standard. We’re simply saying, we want to be very, very strict. Without saying that we expect that this should be the standard that everyone else should use for every other study.

DR. LO: My question is: Why did you decide that? Was it just being cautious? Do you think the fact that this is all going to be in the public domain means that you really do need to get specific consent from the individual, even though it’s anonymized? I mean, again, I’m just trying to sort of think through, is it just that this is the first one out of the block, so you want to sort of create goodwill, or are there more other than prudential reasons for wanting to be especially cautious?

DR. BROOKS: Well, we simply want to be cautious. It is going to be very visible. I don’t know if Mark wants to add more.

DR. GUYER: I think the answers are, yes, to all of the suggestions you made. In addition, we’ve been concerned about the question that came up earlier about the distinction between identified and identifiable; that in the long term, although it’s not possible now, ultimately, there may—these samples, because they are encoded by the unique DNA of each individual, they ultimately potentially could be identifiable, so that in this very early stage of the process, as we’re starting to enter this era, in which, ultimately, anonymity will not be possible, we wanted to go very cautiously and make sure that were doing everything—that participants who were—whose material was being used in this were doing it with consent.

DR. SHAPIRO: Your consent form contains a paragraph—I only had a chance to read it quickly.  I wish I had read it carefully, but I read it quickly. But going to the paragraph which says, "Yes, you might be identified in the future, if you provide additional information somewhere else. These things may come around the back door and be matched." Is that the sense of when you mean potentially identifiable?

DR. GUYER: Yes, a correlation in the future.

DR. SHAPIRO: Okay. We have got a number of people who want to speak.

DR. BROOKS: Can I answer the second question?

DR. SHAPIRO: Yes, please, go ahead. I’m sorry.

DR. BROOKS: There is essentially a very common question. Everybody who hears about this resource is horrified that we’re stripping off medical information, ethnicity information, all of this really great stuff that scientists and doctors would love to know about. We understand that. In order to do a resource, where you actually kept that sort of information, especially the ethnicity information, that would take two to three years to have the sort of meetings, the sort of consensus, have groups—because if you’re identifying individuals as part of ethnic groups, you have to decide what ethnic groups are; talk about group consent; understand group stigmatization issues. That’s a very complicated process. Why did we decide which group who wants to be involved, who doesn’t want to be involved? That’s a very complicated process. It would be very interesting. But, as I am saying, given the interest by pharmaceutical companies in finding SNPs, we’re very concerned that these SNPs get out and be in the public domain, rather than spending a couple two or three years getting a wonderful resource, while the companies are rushing ahead and getting all sorts of SNPs tied up in private databases. So, speed is one issue there. But the thing is, by finding these SNPs, as a large group of them, then researchers will be able to take these SNPs and look at them in their groups of individuals. Because in terms of applying it to certain diseases in 450 individuals, you can’t begin to touch on all sorts of diseases. So, this is a first-step resource. By finding these variants, you’ll then be able to take these variants and look at some group of patients and find whatever variants associated with diseases of interest. You couldn’t do that. You couldn’t get the medical information in such a small sample. It would have to be zillions and zillions of individuals.

DR. SHAPIRO: Zillions and zillions, all right. Tom—okay, Carol.

DR. GREIDER: Could I just put the chart back up that we were discussing as a—So, just to bring this back to what we were talking about. From what I understood, you’re using material that is in the upper-left, nonidentified, and also materials that is the next one down, coded, but not traceable, because you’re throwing away the code.

DR. BROOKS: Because what we’re saying with the NHANES —

DR. GREIDER: The NHANES is the coded, not traceable. And some of it is the nonidentified. And so, you just answered one of the questions, as to what is different between coded, not traceable, and encrypted, feed-forward and why. The answer that you just gave about it would be much quicker to do this study if you throw that away just for issues of expediency to use that, coded, not traceable, as opposed to the encrypted, feed-forward, or some sort of encrypted method. And I think that this is sort of one of the areas of debate that we’re having, and it’s going to come down to issues of how the experiments are actually done, and a lot of discussion with people about the differences between those two areas. So, I appreciate you telling us exactly why it was that you chose to be in the different —

MR. HOLTZMAN: And also, it’s important to make clear that the goals of the S.P. project—and the nature of the kind of resource you’re trying to establish—that the phenotypic information is irrelevant. You’re just looking to map pattern, define common variants. So, it would make no sense to go through trying to collect for the reasons that have been cited. You are not going to get any kind of correlation of interest, unless you get up to 10,000, or when you start, and until you have 100,000 SNPs that can give you the mapping density. So, don’t take this as a paradigm — this is the exceptional study where a piece of meat is good enough. Right? And the same way in which when you’re going to — if you’re going to look for a DNA sequence with the genome, just give me some DNA from a few people.

DR. SHAPIRO: That could be a subtitle of our report. Alex?

PROFESSOR CAPRON: While we have that up here, it strikes me that what you were doing, because you were creating a resource, which is another word for a repository, reminds us that we may have to add another level, or recognize that it can’t exist. There is one repository, which is NHANES, the CDC’s repository. You’re going to be moving from that, in which we have identified samples, to create a new repository; and then once it exists, individual researchers will then be asking for material, or these SNPs will be established. This is going to be a reference material, right?

DR. GUYER: No, it’s a database that simply says at this position in the genome, that there is a common variance of one in a thousand of the G versus the T. All right. It’s not—there will not be a repository of samples. It’s just like mapping information.

PROFESSOR CAPRON: Now, how large is NHANES? How many samples?

DR. GUYER: They have 8,000 cell lines. The actual sample is about twice that.

PROFESSOR CAPRON: And are you going to go to all 8,000 of those people? Using some random method?

DR. GUYER: It’s complicated. There are two subsets within the NHANES. The first one is relatively small. And, actually, even for their purposes, not usable for anything else. So, we’re going to go to all of those people, and then we have — in order to get the numbers up, we have to go to a few more from the larger set, which they are doing analysis on to identify some people in that larger subset, which, if removed, won’t disturb their ability to use the larger sample for the purposes they want, so that taking the small number of samples out of the NHANES collection won’t affect the ability to use the larger set for the purposes that the NHANES study wants to use them for. The cell lines that are removed, that are used, transferred into this repository, will not be used for any further studies, NHANES studies. They are going to be removed from qualification for the NHANES. So there won’t be any additional data gathering on those.

PROFESSOR CAPRON: And out of your 450, how many do you expect to come from NHANES, as opposed to the Native American, or the Asian American?

DR. BROOKS: 310. So, the NHANES people will be approaching about 600 of their individuals asking for consent, expecting roughly 60, to 70, to 80 percent consent; and then, of those, we’re expecting to get 450, which is a different 450 than the total in this sample. So, for 450 individuals from NHANES, we’ll get their consent from; of those, we’re only actually going to use about 310.

DR. MIIKE: Why are you removing them? What is the point of removing them?

DR. BROOKS: The point is so that individuals who give their consent are not going to know whether they are actually in the sample or not.

DR. MIIKE: No, no, but you said you’re going to remove them from the NHANES sample. What’s the point of this?

DR. GUYER: Again, for the same purpose, that if — that, ultimately, in the NHANES study, they’re going to want to do genetic analysis. And there will be a database for our resource saying, "individual sample 363 has a certain genotype," because those data are going to accumulate. If those samples were used in the NHANES study, that correlation could be made, and the link made.

PROFESSOR CAPRON: Now, what I think we should take from this, besides the detailed information you have provided, is a reminder in your decision to go through this process, that with a great many of the kinds of research issues that we have been presented—and I hope that we will get a chance to talk about these when we look, for example, at the Buchanan paper. We are talking about not a tradeoff between science and ethics, we’re talking about what efforts and what resources we’re willing to expend to achieve certain kinds of protections. Protections is even the wrong word — certain kinds of respect for the fact that a person is being — or their material is being used for a particular project. I take it that sort of the subtext of what you’re saying is you have a sense that the Human Genome Project is still a little bit of a hot-button issue, and that the notion that I went in for a National Health and Nutrition Examination Survey, and then that I ended up being used to develop the Human Genome Project might surprise; and then 10, 20, 30 percent of people you expect to say, no, bother those people, to the extent, that they wouldn’t want to be there. And so, you’re going to the trouble, even though it’s a lot of trouble to go back to these people, and some of them will have moved I’m sure, and some of them will be hard to find, and so forth, you’re saying you’re willing to expend that extra effort. And I hope we can keep that in mind as we talk about this; that we’re talking about whole different sets of tradeoffs. It’s not between doing something or not doing it. It’s the effort that goes into doing it.

DR. EMANUEL: I think they’re following these people anyway, so this is much easier. They already have them, that following them is part of the NHANES.

PROFESSOR CAPRON: Okay. I mean that—what you’re saying is that money is already being expended to stay in touch with them, so it’s easier to contact them. With any particular project, it is going to be, in other words, a balancing of how much effort. And you’re saying that the effort involved here is less than it would be if you had a bunch of five-year-old tissue, in which you hadn’t been contacted. I understand that. I’m not trying to say that because this is being referred to as a gold standard, that we would then say that it’s applicable in all of the cases. But we’re balancing the kinds of interest that Buchanan talks about in his paper against—well, among them, scientific interest, and the personal interest. That balancing isn’t struck in a yes/no way, or it’s impossible to do it. Well, it’s possible, but it’s more effort. Here you’re saying it’s relatively not that great an effort. I still applaud them for doing it.

DR. SHAPIRO: Thank you. I have Trish and Carol, and then I also — excuse me. I’m sorry. Diane Wagner from the CDC is here. She may wish to say something, but is not required to. Let me turn to Trish and—you’re okay? Carol.

DR. GREIDER: In this discussion that we’ve had, it brings up to me an issue that it really depends on the background assumption about the protections in those different boxes that one assumes are there. So, if I interpret it correctly, the reason that you’ve chose to put — to go for the nonidentified and the code, but not traceable, was because the assumption was there would be a lot more background protection; that is, there would be more protection for those individuals, than if they had been encrypted, traceable and encrypted walk-back and feed-forward. If we had a mechanism whereby one felt that there were very good protections for people that were in those lower-down boxes, one might be able to then have better protections for the individuals, as well as more information for the researcher. For instance, I notice in the consent form one of the questions is, "Can I withdraw from the study?" And the answer is, no, you cannot withdraw from the study, because we’re stripping—we don’t know who you are. But if we felt that there were very good protection in the encrypted feed-forward and walk-back, then one could withdraw from the study. You could take that person out of the study if they didn’t want to be in the study. And if there was a request for more information, so that it would be more robust within the research side, and more robust in terms of the protection for the individual. But the reason that you’re not going that route is that we feel that the protections aren’t strong enough in those categories.

PROFESSOR CAPRON: That’s not what she said. She said it would be a complicated process to get there. She said that there are issues, for example, about defining what the impact on groups would be that requires you to first figure out what the group is. What does it mean to be an African American, or a Mexican American, or an Asian American, or whatever. And that’s work that they don’t want to spend time doing while commercial outfits lock up the genome.

DR. EMANUEL: But also, Alex, beyond that, that once they do, what process and safeguard—I mean, as I understand it, they didn’t know what the proper thing to do was, because we had — I mean that we, NBAC, hadn’t defined the rules, as it were. So, they made the rules up by the seat of their pants.

DR. BROOKS: And I should add that as part of the planning process, the LC Group at Genome Institute is making dealing with genetic variation one of their priority issues, so that we actually have your good minds lurking about how to deal with ethnicity in an ethically acceptable and useful way.

DR. GUYER: I think it’s fair—the last point is very true, that we didn’t quite know what to do. And if there were better ways to assure protection to the individual, we’d be more than happy. We hope that the kind of lengths that we have had to go through in this case are not going to be the kind of lengths that people will have to go through to do studies in the future.

PROFESSOR CAPRON: I had just one question—is there anything in your resource that will link SNPs with each other? That is to say, will there be any correlations run on —

DR. GUYER: There will be a database in which the genotype—the S.P. data—will be linked to the sample itself, so that we’ll be able to say sample 153 has the following SNPs, and that will be cumulative. And that was one of the reasons for doing this as a central resource, so that all of the work done around the country and around the world on these samples could be cumulative.

PROFESSOR CAPRON: And can you imagine in the future someone coming back to the resource and saying, "I didn’t see this particular issue addressed?" Could you look at your samples again and tell me whether this S.P. is associated with the following other sequence, which isn’t a S.P. sequence? But this XYZ gene that we know, is that something?

DR. GUYER: That is possible.

PROFESSOR CAPRON: Well, to me, then, the answer I got from Steve is not quite accurate. This isn’t just a database which has data in it. It will be something where questions can be asked of that database on a biological level.

DR. GUYER: To a limited extent.

DR. BROOKS: Right. But researchers wouldn’t go back to the database or say, "Would you do this?" They will have access to these resources.

DR. GUYER: The public will have access to the cell lines themselves, and to DNA samples.

DR. BROOKS: As well as the database. So, if you want to correlate some sort of gene with other SNPs in that database, you can get the DNA and —

PROFESSOR CAPRON: This is then a second level repository. The first level is NHANES, and this S.P. database becomes a new repository with data, which at that point, is not identifiable to any person.

DR. GUYER: That’s correct.

PROFESSOR CAPRON: And on which subsequent researchers can do biological research, or the molecular research with the cell line itself.

DR. GUYER: That’s right. And I think the concern that was expressed initially is that the level or degree of research that can be done is probably relatively limited.

PROFESSOR CAPRON: Yes, I understand that. But it would involve another researcher saying, "I’d like to get the cell line to look at it in my own way now." Thank you.

DR. EMANUEL: Why isn’t that a research sample? I guess I’m — as a researcher, that sounds to me like a research sample, not a repository. Okay?

DR. GUYER: Once you get the sample for your research purposes, we were conceiving of —

DR. EMANUEL: Well, let me just give you — the support study, for example, has put all of its information up on the Web, and you can go noodle around and do correlations you want on their data.

PROFESSOR CAPRON: They’re going to passing out the sample themselves.

DR. EMANUEL: For the support study, there aren’t physical samples.

PROFESSOR CAPRON: That’s right.

DR. EMANUEL: That’s all I’m saying. It’s not a matter of critiquing it. It’s just a matter that we were thinking of repositories being the kind of things that pathologists have, or that the people who run the blood spots for PKU have a bunch of stored material, and a researcher says give me some material. Here they have NHANES, which has that material, and they’re going to be creating a new subset of material that’s been removed from NHANES, which they now hold. And it’s different. It’s now unidentifiable — and add it to non-NHANES sources. That’s right. This is going to be a new repository. It was just a subtlety that we had.

MR. HOLTZMAN: But I think it was a subtlety, which we’re putting a lot of pressure on, which I don’t think is important here, because it’s going to go — those cell lines will now sit in Coriell, along with 100,000 other cell lines, where you can call up Coriell and order those cell lines, if you want to do studies of interest. The reason they have a specific — jump in here—sample is that they want everyone in the world working on the same sample, in order to define the SNPs. Once you’ve got the SNPs defined, there is nothing particular interesting about those cell lines. You are going to then take the SNPs, the definition of the SNPs, and interrogate samples of Alzheimer’s patients, and diabetes patients from other cell lines. Because of the relevant clinical information that would have made those interesting is gone. The only thing that is useful for is now going into identifying more SNPs. Is that fair, Mark?

DR. GUYER: Yes, it is.

MR. HOLTZMAN: But the relationship between SNPs here is just where they lie in the genome, and with the frequency that one is associated with another.

DR. SHAPIRO: Last comment before I ask Ms. Wagner if she would like to address this. Bernie?

DR. LO: I have another question about the consent form and the considerations that led you to include, or not include certain things on the consent form. From what I gather, using it as a sort of tool for commercialization would be a pertinent consideration, whether they choose to be in the study or not. And, yet, that’s not in the consent form. I was wondering if you could sort of think through — help me understand a reasoning for not putting that in, because commercialization is one of the things that’s alleged as an interest some people might have in either joining, or a reason for joining or not joining this type of study.

DR. BROOKS: That’s something actually we’re discussing with the CDC people and with the IRB—exactly how strong to say this—because the previous draft did say about commercialization then — I don’t know whether it’s the IRB, or CDC who said, well, it’s okay simply to talk about financial benefits. And so, we’re actually discussing now does financial benefit include commercial? Because we agree that it’s actually important to make that point. So, that’s under active discussion right now, that exact point.

DR. SHAPIRO: We thank you. Ms. Wagner — there’s nothing left to say? You’re not required to say anything, but I just wanted to invite you to.

MS. WAGNER: I am not familiar with the protocol, so I don’t think I want to say anything representing CDC. But let me just make a distinction that’s come out in the conversation, as you think about that one box called "coded, nontraceable." And, as Mark pointed out, that is one of the considerations, in terms of going back and giving informed consent, if it was going to be put into another repository that you have mentioned, and if there was a conceivably a chance of making it identifiable. And it wouldn’t be identified, but it could be identifiable. There would be other studies I would see from that box where you might do sera, and just test one item. Those would never be identifiable. So, when you consider that box, remember that there are two ways you can get information from this. And, potentially, there are two ways that information about the person might happen.

DR. SHAPIRO: Thank you. I propose—first of all. Let me thank our guest. Thank you very much for coming today, and taking your time. We very much appreciate it. It was very helpful to us. We’ll take a break in just a moment or two. And when we reassemble, I think it might be helpful if we looked at some other cases a little while. Steven provided Kathi with some, and David has some cases. And let’s look at that and see, just to give us some experience with particular cases, then we’ll return to the more general issues. So, let’s take a break. Let’s try to reassemble at 10:30.

[BREAK]

FURTHER DISCUSSION OF HUMAN BIOLOGICAL MATERIALS REPORT - Dr. Harold Shapiro, Dr. Thomas Murray and Commissioners

DR. SHAPIRO: We still may have some Commissioners trying to check out of their rooms or something. So, why don’t we just wait one or two minutes more, because I’d like as many here as possible for our discussion. I’ll proceed with our agenda. As I indicated before, I have asked a few Commissioners to think of some cases that might be useful. And I’m going to turn to them in a moment to describe briefly some cases that we may—we help ourselves understand the kind of evolving structure we’re using to see where they fit in that kind of a model, whether exactly that one or not is not critical, that kind of a model; and even considering these particular cases, what protections might seem appropriate in those cases, without trying to reach any final conclusion. Once we have done that, we will turn our attention really to the discussion of protections, and then begin really a lot of more serious consideration of what protections, what the catalogue of protections are, which may be appropriate in some cases, and so on, and see where that discussion takes us. So, let’s begin right now with the cases. And let me turn to David, first. I think you have one or two cases you had thought about. It probably might be useful for us to think about.

DR. COX: As in the briefing book at the last meeting, some papers; and so, the details of that—we didn’t get a chance to talk about last meeting. And so, if you want to go back for the actual papers, it’s in your previous briefing books. And the specific case I want to refer to is a case that basically deals with determining the frequency of the BRCA1 mutation in the Ashkenazi Jewish population, and how that was accomplished. I’ll at least give a minimal of background on this. The illustrative point, why I’m bringing up this case, deals with the issues of community involvement and community consent. And the lesson to be learned from it is, even in the very short time, over the past couple of years, is that the views of community involvement and consent have changed. So what’s the example? The example has to do with there was some preliminary work suggesting that in one particular community, the Ashkenazi Jewish community, that there was increased frequency of mutations in this one particular gene, the BRCA1 gene. So, how could one graphically and efficiently confirm that to see if it was in fact true? Well, this was a situation where one didn’t need to know the identity of the individuals at all. It’s one of those cases that Steve Holtzman talks about all of the time, where that you didn’t really need to know the individual people. You just needed some material from them, and you needed only one other piece of information, whether they were Ashkenazi Jewish or not. You didn’t need to know anything else. You didn’t need to know, in fact, whether they had breast cancer; or whether they didn’t have breast cancer; whether they had a family history of breast cancer or not. That could have been interesting information, but the question that was being addressed was simply what was the frequency DNA changes in this population, and was it different from other populations. So, in order to do that, and to do it efficiently—are there samples available already existing, that could be used for this purpose? And it turns out that there were such samples available. And they were samples that had been collected for another purpose. They had been collected for two reasons. In one case, it’s collected in families that have a history of Tay-Sachs disease, a separate genetic disease, in which that was really used—people could say whether it was used for research, but it was really used for clinical information, in terms of deciding whether people were carriers or not of that separate mutation. And those samples, those DNA samples, were available. The individuals, when they had that material used for Tay-Sachs, had had informed consents done with them, for Tay-Sachs, but no one had ever mentioned anything about breast cancer. There was a separate group of samples from another genetic disease, that individuals were known to be Ashkenazi Jewish, in terms of where the material was collected, and those were individuals that had family histories of cystic fibrosis. Again, that was being done in the context of that disease. There was informed consent for that disease. But at the time that the samples were available, no one had said anything about breast cancer. So, those samples—both of those types of samples—were not just from one place. Some of them were available from Israel, some were available from different places in the United States, and the samples were available, the DNA samples in the researcher’s labs, where they were—had the identifiers stripped off of them. So, in this case, they were nontraceable, they had been identifiable, but as the samples existed in the researcher’s laboratory, and, as far as I know, with the original repositories, or the original owners, is that no one could go back and figure out who was who. So, this fell into the category of "not traceable." So, under the present regulations, these were exempt from any human subjects considerations, or any special protocols. And so it’s stated, if you read in these papers, is that these were samples that were used that were fully in compliance, because they were consistent with the previous regulations; that is, that they could not lead to identifying individuals. What wasn’t stated in the papers at that time, because it hadn’t really been brought up so much as an issue, is that neither the regs nor people made the point that there are in fact group identifiers associated with this, that all of these people were Ashkenazi Jewish. So, what happened then with these samples is they were used without going back to these people, and to get a major of the frequency of DNA changes in BRCA1 gene. And there was no way of identifying individuals, but I think I don’t need to tell anybody, or remind anybody on the Commission how that came out. It was in newspapers all over the place, that basically Ashkenazi Jewish people have an increased frequency of DNA mutations in the BRCA1 gene. Now, what was the impact of that? Well, it’s hard to say. But it’s clear that there were group concerns that were involved with that. So, for me what’s the take-home message from this? This is a situation of samples where there was a complete firewall, because you could not get back to the unique identifiers at all. But—and it completely was consistent with the present Federal regulations—but, on the other hand, I don’t believe they took into view a sensitivity of going back and getting group consultation, with respect to doing this. Now, that’s because I’m not saying it was done wrong at the time, but that times have changed. So, then what would be the suggestion of how one could get group consultation? Which rabbi do you talk to, or who should you talk to? And so I have a suggestion, in terms of this, which comes out of exactly this study, is that this wasn’t one group. It wasn’t just like a big pot of Ashkenazi Jews. It was different local groups, which were different. Some of them had a family history of cystic fibrosis; some of them had a family history of Tay-Sachs; some of them were in the United States; some of them were in Israel. But each of those different subsamples had different representatives and different people contributing. And I would be willing to bet that each group would have a different view, in terms of what the group risks were. And so, for each individual group, although you may not have been able to identify and go back to the individuals, per se, what the group was known; and that it was possible to go back to that group and ask, not to figure out what all Ashkenazi Jews think, but what each one of the contributors of those samples thought; and that this case will come to Alan Buchanan’s paper. But I think he’s right on the mark, which is that individual rights don’t trump group rights; but, even more importantly, group rights don’t trump individual rights. And so, what you can do is do the compromise of going into the subsets of the groups that are contributing the samples, and ask each subset what their feeling is. And would this have been onerous? I think in this case it wouldn’t have been so onerous. I speak as a researcher, because these were four specific groups, which you could have gone in and got the answer directly. What would they have said? I can’t tell you the answer, because nobody asked them. So, that’s my case.

DR. SHAPIRO: Thank you. Alta and Tom.

PROFESSOR CHARO: You know, speaking as a member of this pot of Ashkenazi Jews, I got to tell you that no matter how it is that you slice me up, as an Ashkenazi Jew, or an Ashkenazi Jew with x number of generations in the United States, or particular medical background, or whatever, I’d be shocked if you could find a single person I found to be a good representative of me. This is a perennial problem.

DR. SHAPIRO: I would also be shocked.

PROFESSOR CHARO: I don’t reject the notion that there is a phenomenon called collective interest. I just have seen, as a lawyer and a law professor, so many other settings where the collective interests are implicated, whether it’s in affirmative action, or it’s in litigation strategy, where one group—I saw a presentation once about gay rights, in which one group of people in the gay community decided to pursue a litigation strategy that others viewed as being detrimental to collective interests. I have seen many discussions about the ways in which collective interest could be vindicated through some kind of process that allows for collective decisionmaking. And in every case I’ve heard so far, the cure is worse than the disease. The efforts to find a procedure to vindicate collective rights winds up having so many problems in the determination of who speaks as a representative, and the degree to which their choices can trump individual choices, that in the end one keeps returning over and over to the individualistic model, not because we think it most perfectly represents the interest at hand, but because its like Mark Twain’s definition of democracy, as the worst system of all, except for all of the others. And I’m extremely nervous about moving toward anything that smacks of—even a community consultation model that sets up a presumption against an area of research, because of the expression of disfavor by some people, who have been fortunate enough to be designated the community representative. I am not unsympathetic to the issues about stigmatization. I am just extremely nervous about any solution that depends on a semi-official statement with any force and effect.

DR. COX: And in these cases, both sides—because it’s not someone from one of these subgroups, that basically is speaking for all of the other people, but it’s informing members of the subgroup. They know who they were. And any individual who doesn’t want their samples to be used can say so; and anybody that does want their samples to be used. It’s more informing and having people make the decision, and who makes the decision. I’m not saying that some community—appoint a community leader for each of these. But it’s making known the choice to the individual.

DR. SHAPIRO: Are you saying, David, you would just think that it should have been reconsented?

DR. COX: Yes. Well, it’s complicated to go and identify each individual. There are different ways to reconsent. So I think, yes, to be reconsented, but at least to the point of views that the community, whose samples that was coming from in each local area, and the people knew about it.

DR. SHAPIRO: Okay. Tom, Bette, Alex.

DR. MURRAY: This is probably an utterly, hopeless, wish on my part, but I’m going to—I wish that we could publicly announce my wish that we could impose a kind of discipline on the discussion of the cases. The discipline is essentially this. What I want to know in each of the cases is that—David has additional ones, Steve has ones — can we use this conceptual schema that is still up on the screen, to say, okay, we know where it fits, number one? And I think the answer to that is, yes, it fits in the coded, not traceable box. Question number two is, does it adequately describe the—can we identify the interests, which are raised in this particular case. I think we can identify the interests. Now, David, that’s layered an additional kind of interest, which we haven’t, so much talked about today. We have talked about it extensively in the past, and, that is, that group identity seems to be relevant here in some way, even if we can’t identify the individual. And so, number one is, do we know what—where it fits in the schema? Number two, can we identify the interests at issue here? And, number three, what protections strike us as appropriate? Those are the three things I’d like to get out of each of the case presentations. I think we’ve got one; two, I think we didn’t exhaustively list the interests; and, three, we looked at one kind of protection. And we’ve had Alta expressing some reservations about you know at least one way of one sort of protection that has been proposed.

DR. SHAPIRO: Thank you. Bette.

MS. KRAMER: David, two questions. If you had gone ahead to get community consultation, number one, what would you have done by way of getting it? And how could you engage in reconsent? Because if I heard you correctly, you said that the samples were untraceable. That’s number one. And the second question is: In the last analysis, the research has been done, the discoveries have been made. Do you, do we, does the public think that the Ashkenazi Jewish community is worse off, or better off having gained that knowledge?

DR. COX: I’ll deal with the second one first. I don’t know the answer to that. I don’t know. Some people will tell you that they are better off, and some people will tell you, "Hey, the community is worse off." So, there is no consensus on that point.

MS. KRAMER: Just a follow up question to that. Does not now every individual Ashkenazi Jewish woman have the opportunity to choose for herself whether she wants to pursue this further, and, that is, be tested, or not be tested, and gain the individual knowledge?

DR. COX: Well, actually, the answer is she doesn’t, and not because someone is preventing her from doing it, but there is a whole variety of access and economic issues.

DR. COX: And that’s actually what the potential harms are. That’s what people would argue about from one side, that people would argue from the other side. So, I think there is no consensus on that. Whether it’s been good or bad is the point. The question, which is, how do you go back to do the reconsent, is that these were individual groups that were known. Because, as is described in the research paper, the groups are very clearly identified. Not by individual name, but in terms of where these samples came from, the clinic where they would have come from. It doesn’t state that it’s the exact clinic, but it says the location. So, these are identified groups. And so, what would the reconsent be? And this is just my own personal view. I don’t think it’s possible to go out, and send a researcher out, and do a detailed form of consent to each of these people. But I do think what would be possible — let’s say that I was the medical geneticist in the clinic that was doing the Tay-Sachs work. Then what I would do in that clinic, because I may not know anymore, exactly which of the samples. If I don’t know exactly which of the samples were involved, then this is a problem. As I put it—that it’s unlinked, so I don’t know which one. So, then I could go back to all of the people who I thought were in the biggest group and say, "If you don’t want your samples used, let me know." But in this case, you’re quite right that it wouldn’t work because I don’t know who is who. So, as soon as these are unlinked, where no one knows anymore who it is, you can’t give people the option of not having their samples be used. So, what you can do though, in this case, is go back to the broader group and get a feeling of what fraction of people would want their stuff used or not used. This will not be sufficient, though, for Alta, because the individual no longer has a choice anymore.

MS. KRAMER: Let’s jump forward. Okay? The work has been done. Why are we worse off for that work having been done, and that knowledge having been gained? We are worse off because we don’t know what to do with it, or the options of what to do with it at this point don’t offer really good choices. But the problem doesn’t lie with the work having been done; the problem doesn’t really lie with the group about whom the information has been discovered any more than it’s going to be with any other of these genetic investigations.

DR. COX: And I completely concur with you. This falls—the kind of reconsenting we’re talking about here is basically the—I can’t remember the exact words that Buchanan uses in his paper, but it’s basically one of respect. So, it’s not of being informed, but of being respectful for people to have their choice heard about whether the samples are used or not used. If you don’t know, though, whose samples they were, then individuals can’t have a choice. So, I’ve raised this as a case study, without sort of saying—I have just raised it as a case study.

PROFESSOR CAPRON: I think David’s case study is interesting, but I think two issues have been—at least two issues are being confounded here. One of them is the question, which you were getting at, Bette, at the end, which is the implications of a finding. And those implications can be adverse for members of the group who are identified at the present moment. I would say that, in terms of the design beyond the consent issue—I don’t want to rest everything on consent. One can imagine situations in which you would have to say it is wrong for a researcher to go about a process, given the situation, that if, right now, insurance companies were to announce that an XY degree of risk, we are not going to write insurance for people, life insurance, unless they will undergo certain tests. And if found through the test to have the gene in question, they are not going to get life insurance. We say, "Was that true of breast cancer?" They say, "No, the rate is too low and the population is not worthwhile our screening every woman who applies." If I told you it was five times as high in a particular group, yes, it would be worthwhile. So, if you develop the information, and say that’s true of this group, however defined, then you input them at that risk. And I think that question has to be asked by the research community is that, given that circumstance, not the making of the investigator, that’s an existing social circumstance, it’s a question that would have to be asked. But that would be true if we were not talking about stored data, if we were going out and getting the research information now; that is to say, if the researcher was not in a situation where the IRB would say, you may not do this research, because you were imposing upon a group of people who are not able to consent—that is to say, all of the women identified as Ashkenazi Jews, and who were going to have this high rate, an unacceptable risk. And they said, no, this is—we don’t have social circumstances that make this prohibited. It’s a matter of choice. Then what would seem important to me would be when you approach an individual, asking them either whether their stored sample, or a newly collected sample may be used — you say, "Our objective in this research is to make an epidemiological sort of statement. That is to say, the prevalence of this gene, is it higher or lower in people of a certain descent? Is it, for some reason, found at a greater rate here? And that is a question which we haven’t fully figured out—which is, how do we expect people to take into account the interest of others who are affected by their decision? That also arises in family situations. If I get screened, it has implications for my relatives, etc., etc. We recognize that as an issue, and that’s one group where we tend to say the consent will handle it. I’m with Tom, however, in wanting to look at our chart here and to realize that what is—the Chairman has occasionally spoken as though we get down to a bottom line here. But I think we actually—what the chart will eventually show us is all sorts of questions that get asked along the way; and then to loop back through questions. For example, here the question would—and so, you are under the nonidentified, or something, because we don’t know any particular sample. But the question is, will the research results have implications for an identifiable group? Yes/no. And, perhaps, if yes, then we were saying then somehow consult with that group, or at least find out what the sensitivities of that group are, etc., etc. Now, or we may say, "No, you can’t do that," because for most groups there are no legitimate representatives, and we just leave it to each individual. But those are the kinds of choice points on which we should be focusing, and try to move our discussion forward, beyond these sort of rough boxes, to the questions that get asked with yes and no answers.

DR. COX: That’s exactly why I put this case study forward, because of the discussion that’s come out, this is about as straightforward a one as you could in the box of not coded — or coded, not traceable. And where there are groups having said coded, not traceable, there are limited definable number of substantive people that those samples came from.

PROFESSOR CAPRON: Yes, I understand that. But it seems to me that there is a small confusion here in talking about the people from whom they came. If we were saying that you were now going to make statements about the congregants of Temple Beth El, then it would make sense that you would go to that temple and say, "We have sample which we collected when we did the Tay-Sachs screening 20 years ago. Yes, we would like to use them now for this other purpose." When we’re done, people are going to be able, either because we’re going to say it’s Temple Beth El, or because the people will be able to know we were in Baltimore, and we were asking this group, and we know they know that Hopkins has those samples. Somebody will figure it out from the Baltimore Sun or something, and publish it that way. Then you are actually saying, "This group should be able to protect itself." But if you’re using those samples, but you’re not going to make any statement about that to a particular group of congregants at that temple; but, instead, you are going to be making statements about an ethnic population spread around the world, of whom this is just what you believe to be a representative subset, then you’re in a different ballpark, it seems to me. Because then, while it makes sense to ask these people are they willing to have their samples used for this research, if they answer, yes, their yes has an influence on people who are not being asked. And you’re, in effect, saying—and I’m not saying this is wrong. I’m just trying to spin out what to seems to be the implication. You’re in effect saying that if you make it explicit to them, you expect them to weigh the harm that could be done to others, with whom they share a characteristic by these findings, or the good that could be done for those others, are we better off alerting this particular subpopulation that they should be tested more frequently for breast cancer than the average woman, because they are at higher risk; or are we telling them they had better buy their insurance today, because they won’t be able to buy it tomorrow? I mean what are we telling them? Those are the kinds of considerations that we have. So, we have two different meanings of going back to the group. One is where you’re going to make statements about this very group, and you have to ask them before you do it. And the other is, well, you’re asking them as surrogates for a larger population, and your statement is going to apply to that larger population. And are they an adequate representative, or do we just say there is no way of getting that, and that’s one of the problems that we have here? Alta quoted Winston Churchill, to whom she gave the name, Mark Twain, about democracy—yeah, Winston Churchill, I think. But, in any case, I want to quote Bernard Dickens yesterday, though, in saying that—reminding us that, even if there is not the very best way of doing something, there may be still a better—there may be — it may still be worth doing it. And, in this case, I’m not referring to the research, but the looking for the consent. I would still gather from this that it is better to have asked people who at least identify with the interest of the whole population, whether they think it’s appropriate to go forward with the research, presupposing that it is. I mean to give a different example that Buchanan gives here. What if we were looking at the correlation between a group and violence, or "criminal behavior," or something. Now, the notion of that kind of thing, which has such racist overtones, could be done without asking people whose stored examples you have to have, as the next-best surrogates for the group interest, would strike me as wrong. Because if I knew my sample had been taken to make statements about Caucasian males’ potential for violence—as opposed to Caucasian females’ potential for violence—I would think that even though it’s going to show that our potential is much lower, I would want to be able to know that I was in some sense contributing to that. And if I thought it was pernicious research to say, "You’re going to have to find some other way of doing it. You’re not going to do it with my sample, thank you very much. I think that still has, and still — I differ with Buchanan on this. I think there is a question where respect means the right to nonconsent, and that’s — to me that’s—I’d love to get—when we talk about his paper, because I do think there’s a conclusion that he doesn’t seem to reach from this, where it’s worthwhile going to the person and asking, "Can we use your sample?" Because it would be research that I wouldn’t want to have done. I wouldn’t want to contribute to, even in an indirect way.

DR. SHAPIRO: Why don’t we hold off. A lot of people want to speak. If we go in order, we can accumulate your thoughts and responses. Bernie.

DR. LO: David, I think your study is very illuminating, because it suggests that the model up there needs another dimension. So, I agree with Alex. There is a question — there is an algorithm or flow chart begins, or is it not identified community of collective, you might want to say, implicated. If, yes, then there is a whole new set of questions we haven’t really sorted through. But I wonder if, again, using the analogy of community consultation and HIV research, suggest different meanings for that, than I think people are talking about. I think a lot of us are working with the assumption that you assemble representatives, or somehow will have the authority to speak on behalf of this community, and either approve or disapprove of the research, so that it’s sort of parallel to the individual consent. There is community consent or approval of the research. I think Alta’s concerns about who do you get, and are they represented, in what sense are they legitimate decisionmakers, are all very important and are solvable problem. I would suggest that the true value of community consultation is not that you could walk away saying they approve, it’s an okay study, full steam ahead. But that if you don’t ask, then you’ll never — first of all, otherwise, the issues will get submerged. If you don’t ask, you’ll never find out what the concerns and problems are. And what we have gotten away from in the AIDS world is not—we want to do it; no, you can’t; yes, we will; no, you can’t. It’s, look, we have a problem here. We’d like to solve it. There may be some concerns. If we understand your concerns, we may be able to design a better project. And what typically happens is not that the community representatives say you can’t do it, they say, "Well, why are you doing it this way? We’re concerned that this may cause that." And what happens is that you end up designing a very different project than you started with. So the consultation is not for approval, but it’s to sort of strengthen the protocol. I would say it’s not just to address ethical concerns and community interest, but also it makes a tighter study. It always is an implicit veto, that if we really don’t like it, we will cause enough stink that you will be — you, Mr. and Ms. Researcher, will be so unhappy that you will withdraw on your own. I mean that’s, in a sense, a threat, and it was used early on. But now I think it’s less of a confrontation. The second way in which community consultation is really helpful, is not redesigning the protocol, but correcting the researcher’s idea. David alluded to this. We’ve got a research study. Here is a quick, simple, inexpensive way of doing it. We’ve got all these stored samples. We don’t have to go the trouble of collecting the samples. They are all there. It’s just going to be a nice, efficient way to do it. And one of the things you learn doing this kind of research is when the community is involved, there is no quick and easy solution. So, what the community typically says is that we have no problems with this being an important question. We have no problems with your integrity as a person to do this research. We have a lot of problems with what this means, what the misunderstandings might be, what the implications might be. The issue isn’t are we better off doing this research, or not doing it. The real issue is, can we do this in a way that addresses some concerns and questions we have. And, typically, what evolves out of these meetings is the sense that it may be cheaper, in terms of, if you don’t have more efficient, you don’t have to collect more samples. But the researchers end up doing a lot of things with the community. So, they go to community meetings, which may be open community meetings, where they just explain what BRCA1 is, what it means. Does it mean I’m going to die of cancer? Should I be tested? They probably are going to be asked to have more specific meetings with the rabbis, or people who are in community-based organizations that serve that community. Because it can be expected that when the results are known, people similar to the people of whom the studies — will have questions about what it means. And that’s clearly not information that the community knows, but it’s information the researchers know. And the researchers get educated, in that they have an obligation, not just in the research and get out of the community, but to make sure that the people who have—will be directly affected by the implications of the research understand what it means, and have their questions and concerns addressed. So, you end up saying, "No, you don’t have to collect the samples," but you spend a lot of time going to community meetings. And it’s part of the price of doing research on a population that’s particularly suited for the type of work you’re doing. And just educating researchers that that is part of their role—it’s an obligation that’s important—and that really only comes from these community consultations. There is also—another thing that happens is the researchers begin to broaden their role to include, not just doing the research, but how they write it up, how they talk to the press about it, and how they get involved with policy implications. So, again, going back to the AIDS example, early on the community said, "In addition to our concerns about how you design new studies, and your lack of providing education and information to the community, we’re concerned that you guys aren’t speaking out on the real policy issues, which have to do with discrimination, confidentiality, funding for AIDS research, all of those other issues that are — you know, in the minds of some community leaders, linked to the outcomes of the research. But researchers typically say, "That’s not our job. That’s policy. We’re just scientists in pursuit of the truth." So, I think to me the true value of the community consultation isn’t that you collect a bunch of people who vote, yes, and it’s okay. But that from having asked the questions, and had people talking to you, you have hopefully changed what the researchers regard as their role, and their obligation. And it’s particularly important in a situation where you are going to be going back to these people just because of these genetic quirks, and this population that made them so efficient a group to study for these diseases. And it’s like the international researchers we talked about yesterday. You don’t just do the study, take advantage of the community and get out, you have an ongoing relationship, and researchers need to learn that. And, frankly, having real live people in the community, these are not representative, and they may not represent precisely anybody but themselves. The fact that they are talking to the researcher face-to-face means the researchers are more likely to listen. So, I think we need to — you know we sort of not define what we mean by community consultation. I think it’s not really analogous to community approval or consent, the way it is for individual consent.

DR. SHAPIRO: Okay. If understand what you’re speaking about, Bernie, which I—in which you’ve been quite eloquent. It really is—what you have described takes place in a process of study design and formulation of the study, and it has impacts on everybody in that process, I understand. Carol?

DR. GREIDER: Bernie made the point that I was going to make and he made it very eloquently, so I won’t go much further, except to say that we did have in the Genetics Subcommittee when we met, Dr. Killen, I believe. He was from the NIAID, who spoke very eloquently about this. And, in fact, completely changed my mind about the issue of community consultation. I was struggling with the kind of things that Alta was raising about how do you get people to actually consent for something, for a group, and that sort of stuff. And when he presented what Bernie just summarized to the Genetics Subcommittee, I was really impressed with the idea of involving a community in the research, in the way that Bernie just said. And so, perhaps, my fellow Commissioners could go back and look at the transcript from that meeting, and look and see exactly the details, and also for the report about community consultation. I think he was very eloquent on this topic, and maybe we can just review that.

DR. SHAPIRO: Okay. Kathi, do you have something you want to say?

DR. HANNA: I just wanted—some of the clarification I was going to ask for has already been made. I just wanted to raise two points. One is that the Commission think about consistency between what they’re going to say about community consultation in this document, and what they anticipate they might say in the international document, and there not — that there not be two sets of standards, one respecting a democracy, and the other not being as sensitive. But the other thing was I wanted to just ask how—since we’re on the issue of community, how we should proceed in preparing something for the next report, whether it would be useful for staff to go ahead and start drafting what we know from the literature, what’s been reflected in the discussions, and proceed from that point of view. Would that be the most useful approach?

DR. SHAPIRO: I think it would be useful. I don’t want to get us off to just discuss that particular issue now, since there are other cases we’ll want to get to. But I just want to indicate that, in my sense, I think I support everything Bernie said, or, at least some of it, but I want to support everything he said. It really is—there is something in my mind which is qualitatively different from that, then what we have normally been calling protections, even though I understand that it may be even more important than the protections. So, it’s no question, one being more or less important, but it’s just qualitatively different. And as we think about it, we have to think of the right conceptual apparatus, as to where to put it, and how it should be reflected. Because I think what I sense around the table is everyone is concerned with the issues Alta raised, and we don’t see any resolution to those issues that Alta and others have said. But this is really a qualitatively different dimension of the issue. And so, we’d have to think of the right conceptual framework just to incorporate it. Tom?

DR. MURRAY: Some of the elements, obviously, of the comments Bernie has made, the kinds of things that Jack Killen shared with us... the pieces of Alan Buchanan’s paper, which are making I think a very good effort to tease apart the notion of the interest of a particular group of community, and begin to understand that there is really several interests which are distinct — related, but distinct. There is also a concern, a paper I will share with you, recently given at NIH by a colleague of mine, Eric Juengst, which I think expresses some concerns, which we—not today, but we’ll need to think about before we bless the notion of community consultation—about concern that we not symbolically or otherwise contribute to a kind of genetic reductionism by assuming that genetically defined populations’ themes are the same as culturally defined populations. Put those pieces together, and I think a draft at this point would be quite helpful. Alta?

PROFESSOR CHARO: First, are you sure Winston Churchill wasn’t quoting Mark Twain? It is so unnerving when you were sure you knew something.

PROFESSOR CAPRON: I only know it because Zeke agreed with me.

DR. EMANUEL: It’s just confirmatory data.

DR. SHAPIRO: What do these men know anyway?

PROFESSOR CHARO: I’d like to ask whether people have in mind that community consultation is something that is a backup with individual consent and dissent was impossible, or whether it’s supposed to be present, even when individual decisionmaking is possible. The reason I asked that we reflect on this is because there is going to be a great deal of opportunity for individualized consent and dissent, I believe, when we begin to break this out a little bit. First, for all sampling of human material from here on out, even before this Committee and many others come up with recommendations, I think any sensible sampler would take advantage of the most broadly drafted consent form possible, and I have seen examples of ones that are very good on this, in which they ask people if they’re willing to have their sample viewed indefinitely into the future for a variety of as yet unknown research protocols, including some that might implicate group stereotypes, and it gives people an opportunity to say, yes or no; and then the repository to have that consent form and that particular human biological materials is ever after able to know whether or not that material should be released for particular protocols, which means that we’re really dealing only with existing collections, and those people who are silly enough not to start doing this right now. For existing collections, there are good reasons why existing collection repositories are doing that now because it’s a way to avoid an awful lot of the messy problems that are going on and they can have a reconsent for a variety of future things that allow people to say yes or no to indefinite, future, wide-scale things. Which leaves only the residual of those people who have been unfound, whose materials are in existing collections as the ones for whom we don’t know what their individual attitudes are. And there the question becomes, should you have a presumptive consent or presumptive dissent for any particular kind of use, and that’s the setting in which, if community consultation is a backup for lack of individualized decisionmaking, that’s the place you’d be using it. I don’t know what the actual number of those people is going to be, but it’s certainly smaller than the complete universe. On the other hand, you might all be thinking you want community consultation in conjunction with individualized decisionmaking, and that’s a whole different story. It would help me to understand how tolerant I am of the problems associated with community consultation if I knew the scale on which we were going to use it, and the degree to which we’re going to let it substitute for making a genuine effort to get individualized decisionmaking.

DR. SHAPIRO: David wants to say something in a moment. I just want to make a comment, at least on how I’m thinking about it. I simply don’t look at it as a substitute for consent, except in very unusual and particular circumstances where it might be appropriate—I haven’t thought that through. But I’m not thinking about it at all as a substitute for individual consent or as, well, just say that. The question is, in my mind, can the kind of interactions that Bernie described improve the overall quality of what’s going on and answer the needs and maximize the benefits, minimize the harms that attribute to any particular groups by behaving the way Bernie indicates. That’s my own view. But now, it’s not clear to me if that’s the case, if that’s the model we’re thinking about, it’s not clear to me just what focus it should have in the report. That’s why I mentioned it - I’m not quite sure what the conceptual framework is, because a large part of it goes throughout the process, as Bernie said. Both planning it, carrying it through, follow up. It has impacts every where, it impacts on researchers’ ideas and so on and so forth. That’s of a different dimension—that’s the only word I can think of right now—than the kind of thing that’s consent that we normally talk about. So it’s a different, it really is a -

PROFESSOR CAPRON: It really is a consultation.

PROFESSOR CHARO: Perhaps I just did not make the question clear. Let me try very briefly to clarify what I’m asking. There are going to be many opportunities in the future and still today to go back to individuals and simply say, "What do you want? Do you want to allow your stuff, even if used in an anonymous, unlinkable-back-to-you fashion, do you want to allow it to be used for things that you may find pernicious?" Yes or no. And I’m asking if we could do that for most people. What role does this community consultation still play after these people have been asked what they want. I understand the debate about community consultation where you can’t ask people. You can’t find them to ask them. But I’m not sure I understand the role of community consultation where you can ask them.

DR. EMANUEL: Maybe I can—I hate to jump in here, but I will. Sorry, Eric. As I believe the Subcommittee was discussing it a long time ago, it was in addition to any potential individual consent in the context where research might not implicate anything about an individual but might have implications for a community. So even though you couldn’t identify the individual, as David has made quite clear, you could identify a community where your research result will have implications and could adversely or positively affect their interest. And the idea of community of consultation, which has evolved, was an effort to understand and also to revise, reform, refine the research and maybe not even to do it. Depending upon whether it affects, and how the community views the effect. It is not a substitute for individual consent, but it mainly arises in the context of research, which may not be individually identifiable, but as community, it is targeted at particular communities that are definable.

PROFESSOR CHARO: It’s not about linkable to people—I understand that. But you could ask me whether I want my stuff to be used in the future, even when it can’t be linked back to me. I can already have told you what I want, whether I want to be implicated in this research or not. I’m asking why then do these arbitrarily selected representatives get to decide?

DR. EMANUEL: Here’s the issue, because you are convinced that any time that you can decide and your rights are implicated, that whatever the group is, if you give the group rights, that’s just going to be an unworkable solution. I don’t think that everyone necessarily shares that view. And what you decide as an Ashkenazi Jew could have implications for me or other people and they could be quite adverse for me. I think the question was, beyond the individual, there are significant group concerns. I guess, at least as I heard it from the Subcommittee, there wasn’t a shared view that, "Well, we understand it might have implications for the group, but we’re going to rest everything on individual consent." There was a sense that the community needs to be involved in participating in the design and decisions as well because the community is going to be implicated as well.

DR. SHAPIRO: Let me make a suggestion here since this is an issue which we’re going to have to resolve and think about carefully and articulate. And I think the next best place to articulate is in writing and see how it looks and let’s see how we feel about it. It is a very difficult issue; it’s not an easy issue. But let’s try to articulate that in writing. There is quite a fair amount of literature in this area by now and let’s see how we can attach it and see if it makes sense to us. And I don’t mean to so arbitrarily cut off, but we just have to get on. But we’re going to be returning to it—it’s not resolved, we’re not resolving it. But I want to just give David one last shot since it was his case he presented.

DR. COX: And I’ll make this very brief. It was a case study and I presented one side. Now, I’d like to make it very clear what I actually believe in the case of community consultation. The discussion was extremely interesting to me and Bernie’s comments, which resonated with everyone, didn’t have anything to do with the case that I presented because they dealt with something that could be dealt with prospectively. If you’ve already used retrospective samples where you can’t identify the people, they can’t be involved in the research design. So I really support the idea of community in the context Bernie was talking about. I use this example to try and, in the best way that I could, talk about how you would have community involvement with retrospective samples where you didn’t know who the people were. And for precisely the reason that Alex says, I have difficulty with this because it confounds who the individuals are versus who is speaking for the whole group. So for me out of this discussion, it reinforces my own belief that community consultation in the case of coded, nontraceable samples where you don’t know who the individuals are, I don’t know how the hell you do it. But I do understand how you do it in other boxes. So that’s why I use this example because I wanted to try to have some better insight and I didn’t find it.