 It's a common mistake, Teri. Gail from University of Washington. Where is the whole, like, there. OK, so I'm here to talk, really, on behalf of a very large group of people at the University of Washington, including Debbie, who's here to correct me, right, hit me, possibly. She's close enough. On behalf of Seattle, I will just say that it's a beautiful place to have a West Coast September meeting. No agenda there. OK, so I'm going to talk about a bunch of things. Briefly, I'm going to talk about Debbie's Genome Center, and I should say Debbie and Mark Reader, and the new Mendelian grant that starts, I believe, today, of which the PIs are Debbie Mark J. Does it start today? I think it does. Today or tomorrow? All right. No, no, the announcement's tomorrow. It starts today. Jay Shenduri and Mike Bamshad. The Seattle Seeks Exome Variant Server, which is a great resource for people who haven't been using it. One slide on the Emerge Consortium, a little on CLIA sequencing opportunities at UW, and then a little bit more time on the clinical sequencing study, which I'm the PI of, but is a huge group effort at UW that is also new. If there's a minute, one slide on the Northwest Institute of Genetic Medicine. So the next generation, Mendelian Center is, there are several centers that were funded to identify Mendelian diseases. And Debbie has, I think the largest, I'll say, since you're not the one talking, maybe by a dollar, but in any case, the idea of these is to do the unsolved Mendelians. And of course, there's been really impressive progress in the last couple of years in unsolving the Mendelians that people had in their freezers. And so we're hoping to find more of those and really make a lot of progress. Not every pedigree that we can solve will look like this one. And in fact, this one may not even be solvable. We like recessives better than dominance. And so I'm going to talk quickly about two other models that have worked. Take a minute to say, though, that question of how do you know you really have it is a really important question. And so validation or replication is important, but also those functional studies become important. And definitely an advertisement for if you have these in your freezers, let Debbie know right away. So here is one of the ones that was solved under the A-R-A small. We called it little Mendelian. Now we have big Mendelian. And this is a de novo mutation in autism. And I think this is what Debbie was referring to when we don't always have to look for a family history to find a gene. And actually, it was very powerful in the autism. Out of 20 children with autism, they found de novo mutations in four. So I advise you to look up that paper. Similarly, we also solved through this project a quantitative trait locus for phospholipid transfer protein activity level, which was at last force, ceramide synthetase. So not a place we would have thought about except that there was a linkage signal in this region. And we were able to take the exome data, lay it on the linkage, and come up with the variant very, very quickly. And then of course, it took a lot longer to confirm it. But it was really quite remarkable having had the linkage signal for some time to go through and actually find the variant in weeks. The exome variant server is the web addresses here. If you haven't used it, you should really go play with it. They have 5,400 exomes up there right now. And I have found it extremely useful when I'm interested in a gene to go see what else is there. The phenotypes are from the NHLBI studies. So the large cohorts that are listed here. So there are early MI, hypertension, et cetera. But for any phenotype you're interested in, it's interesting to look at the gene. What I will tell you is that they get like a request a week. Can we find out about this phenotype in this person? And the answer to that right now at least doesn't know. So here's some of the other features that are up on the exome variant server. And you can also get the allele counts, which are a new feature. And so it's a good place to go play. Thank you, Debbie. I would have drunk your water, but I know you had a cold recently. It's not my water. OK. All right. Emerge. So several people have referred to emerge. And the seven sites that are in Emerge 2 are up here. The idea for the three of you in the room who don't know about Emerge is to take places with electronic medical records where the phenotypes can be raised from those medical records. And pair that with genome-wide association data, and then share. So kind of I show you yours. You show me mine. And so the phenotypes that we picked, for example, we picked white blood counts. Someone else picked eye disease. Someone else picked QRS. We all pushed the phenotypes and the genotypes across so that we have a large cohort of people that were analyzing. And that's been very powerful. The Emerge is now looking at deploying the new pharmacogenetics array across all sites. And that will be a new component. And then the bioethics group has been very, very productive in Emerge to look at very, very important issues with what patients want out of these new genomic technologies. Cleo-sequencing at UW, there are actually lots of places. You can get cleo-sequencing done. Peter Byers historically has a collagen lab. He does arrays for collagen disorders. But he'll do anything you ask him. So it's great. So we use Peter, and even confirming variants. We use Peter a lot for things like that. Mary Claire King's lab, and in particular, Tom Walsh developed this array. He's a really great junior guy. They have a 29 gene cancer array. And they actually make these primers available. So if you want to have their array, they'll tell you how to do it. And they're happy to export it. And they've exported it to several places so far. There are seven colon cancer genes among those genes. And that panel is available as a clinical diagnostic test that insurers will pay for. It's called Coloseq. And it's kind of like a chemistry panel. You get the report on the seven. But actually, they're running all 29. And so you will get the rest of them only if there's something important there. So for those of you who don't know how chemistry panels work, if you order a sodium, everything gets done. But you don't hear about the rest of it unless something bad is in there. So this fits in that model. OK, so they have a current trial actually going on where they're sequencing these 29 genes in all new cases of invasive breast cancer. And they've been looking at colon cancer for some time. Laboratory medicine at the UW now is setting up Mary Claire's array as a fee for service. And then coming soon, Debbie and Mark are going to have a CLIA certification of their lab. We're moving down that pathway. And so we're really looking forward to having CLIA-certified exomes, genomes, exome chips, and everything else over there. All right, moving on to the clinical sequencing and cancer project that, again, is one of the new U01 projects that's just started, which will technically be announced tomorrow. Our project relates to colorectal cancer and polyposis. The project one is the genomics study with an economics overlay to look at utilization. Project two is the technical whole exome sequencing, but also we do intend to return those incidental findings. So we have a group of clinicians that are going to help us determine which are the results that we should be giving back to patients. And then project three is an ethics project. Separately, there was a return of results. R01 funded to Holly Tabor at the UW for electronic return of genomic results. So our study rationale, I mean, we sat there kind of long and hard and said, well, if we were going to deploy an exome in clinic, what would be the best case scenario? And we just went with colorectal cancer for these reasons. First, there are a bunch of genes that are known to affect it. Second, the process, which I'll show you in a minute for coming to those diagnoses, is actually quite complex and involves multiple visits. And finally, in our clinic, we solve less than half of these patients. They look like a Mendelian colorectal cancer, but they don't have the known genes. So we thought that whole exome sequencing might help us solve more and solve them faster. So this is the usual screening for people who come in looking lynchee. So they have colorectal cancer early age. They have family history. They have a couple of polyps, but not too many. And you go through and you actually look at the tumor pathology, use that pathology to help you order genetic tests. When those tests are negative, then you do some more tests. And you can go through four clinic visits easily and not find the answer. And then eventually either the patient, the doctor, or the insurer drop out of the arrangement. And since many of these remain unsolved. So this is the schematic for this new study. Basically, it's a randomized control trial where we're taking these patients that we would have ordered some kind of colorectal test for, even the tumor marker testing. And we're going to randomize them to usual care or usual care plus an exome. And this is how IRB let us get around the not telling people anything. We are going to draw the blood in the usual care no exome people, but we are actually not running the exome until in the end if we can cross them over we will. So they will get in a very short amount of time, four to six weeks. The colorectal results that are identified, those will be CLIA certified. So initially before the whole exomes are CLIA certified, we'll repeat everything in a CLIA lab. And hopefully the CLIA exomes will be on board pretty soon. The patients will be surveyed. And then after that, we are going to return the incidental findings to that group. We had a hard time coming up with a matched intervention for the people who didn't get the exomes. They're basically going to come back and talk more about their family history and other clinical risk estimates for them. Survey people again, survey people again. And one of the important parts of this is that we're going to follow economic measures as well as other kinds of measures. So the study is randomized. We are going to control for things that we find as confounders. For example, we have a very low number of non-caucasian patients. So those will be randomized tightly to make sure that they end up in both arms. The blinding will end at the second clinic visit. So at the first clinic visit, the patients won't know what arm they ended up in. But when they return for their colorectal results, the clinician will know at that time, and then the patient will know. So when the physician makes their plan at the first visit, they will not know if the patient's been randomized. The physician who will see most of these patients is Fuki Hisama, who is fabulous. So we are very, very lucky to have a strong clinical team. We see about 600 cancer genetic cases a year. So we believe we'll have enough over four years to accrue the 220 colorectal cancer patients we need to be well-powered. The unsolved cases will move to a discovery arm, will collect their family members, and try and find new genes since the old genes will not have accounted for those. We're going to collect a bunch of psychosocial and economic outcomes on these patients. The health care utilization will be followed by postcards where the patients just say what they're doing. And the economists, Dave Vienstra, who's fabulous. And I like talking to the economists. I'm a mathy kind of person. They put a price on the activities. The patients don't have to report actual monetary values. We are additionally asking them about their insurance, what changes. So if they suddenly add a bunch of disability insurance, we want to know about that. And we are asking them what they're telling their family members. The economics group is also giving subjects in both arms. People who got exomes and people didn't. What's called discrete choice experiments, which you pick between two things to try and put a value on items. So then we come to the return of incidental exome findings. And I was really hoping that Howard could help me with what to give back to people. That's why I was asking the question. So we have a group of fabulous people that are going to help us to decide what are valid and with clinical utility, actionable, whatever terminology you want to use. I did change my slides since Friday. And basically the binning kind of worked at Jim Evans and his group do. The only two outside people outside of Seattle are Jim and Jeff Murray. So I just sort of highlighted them in blue. And then we do have several ethics people who are physician ethicists. And of course, Wiley Bird is the co-chair of the group, is a medical geneticist and ethicist. So that'll be an interesting process. Alone there is the third ethics project is basically focus groups in patients who got exomes, focus groups in patients who didn't get exomes, but weren't part of the study at all. So outside of the study. And then also we have this CLIA aim where we are having Barbara Evans, who's a lawyer and professor, actually look at the First Amendment issues related to return of non-CLIA results. So there's a big argument that I'm not even going to get started here on whether for research, not for clinical medicine, it needs to be CLIA, but for research, can you give back? If you have a non-CLIA result that you believe is important to the patient, can you even give it back is an argued point. And so Barbara's going to look at the First Amendment freedom of speech issues around that question. So that'll be interesting. The Northwest Institute of Genetic Medicine's the last thing that I'm going to talk about this, and these are the usual suspects. This is money awarded to us by the state of Washington to help investigators move into translational genetics. We have today worked with 180 different investigators over three years and helped a lot of people with either bios, with all of these things. So either, I don't know really what I do. I just started all. But Debbie does the technology. Mike helps with sample collection. I help with the IRB, I guess. Peter does the informatics. And Bruce does the biostatistics. And then we've started a biorepository with Eric Larson at Group Health. So we've been able to help people with a variety of ways. If they need a 20% biostatistician to do genetic analysis. And the most popular person has been our full-time genomic IRB person. And she's, I think, going to be a very crucial person as we move forward. So this is just a very, very partial acknowledgment of the faculty involved in these projects. And the staff is also amazing. And I felt bad leaving them off. But I got tired on the last slide. So if anyone has any questions, thank you. Great. Questions for Gil. So Mark and then Erwin. So your randomized trial, again, what's the primary endpoint? The primary endpoint is actually the percent solved. The percent of patients that we find the genetic cause that we believe accounts for their colorectal cancer. That's the primary endpoint. For their cancer. For their. For the colorectal cancer. So they're coming in. But these are patients at risk of cancer? These are patients with colorectal cancer who also have enough of a family history or presentation that we believe they have a Mendelian condition that we would have ordered a genetic test of some kind in. I guess. But not the APCs. How do you know that if you find something that that's the cause? I think that is a very important question. And so part of it is there are genes that we know that never get tested in these people, that we know cause colorectal cancer, because we do it kind of one step at a time. And so I think that that's a really important question. And I think that we'll be very conservative about telling a patient, these are known genes that we're going to return to the patient. The discovery arm is separate. I don't understand why you need a randomized trial for this, because basically you're saying in patients with colon cancer and a family history that there's some percentage that will have abnormalities that can only be picked up by XM sequencing. And that doesn't require randomized trial. You really don't have a cause and effect here and you really aren't doing any intervention, right? It's an incremental genome. But why not just do it stepwise? Why not? Because of the high... So that's a major part of the hypothesis is, is it faster and cheaper just to do the whole exome the first day you see them than to walk them through the tumor testing gene by gene by gene? Is it just faster and cheaper to get it done the first day? Yes. I hope so. Well... I mean, the end of the day, you know you're going to find stuff and you know it's going to be faster and cheaper than going step by step. But you won't know whether it's really beneficial because you don't really know what you're finding is important. Well, I would disagree with that. I think that we, you know, finding these genes is beneficial to the, not just the patients, but to their families. Maybe, you would need to have tested their other family members to show that there's linkage. Well, we... So you're talking about new discovery genes, but many of these genes are going to be known, you know, clinically, you know, clinical validity and utility known genes. Okay, that many, much of what we're going to finding is going to be that. Patient and provider fatigue is a huge issue. And so that doing it all at once, to get the data all at once is actually going to be so much more productive in terms of actually getting results than when you have someone come back and they don't show up or you can't get insurance cover. That to me is actually a crucial part of this question. Getting an answer to begin with when you don't have to go through all the steps is just a big thing. I mean, this is what we do. I mean, I'm sure it happens to everybody. Right, and that's what I meant when I said either the physician, the patient, or the insurer can drop out along the way in the current workup. Yeah, Debbie, I wanted to... Gail. I flanked you with both. More about the trial design, which is interesting to me. It has been reviewed and obviously been considered well done and you're getting funded. The issue is the following. So you have apparently double blind, right? You have the physician and the patient until the first return visit. And then the physician must... The physician knows, the patient doesn't know till the return visit. The physician knows before... Does not know until they're done dictating the note from the first visit, but does know before the second visit so that they can appropriately cancel. Does the physician know that a patient is in the study but does not get exome? Yes. So doesn't that introduce a bias? The... The physician knows that this is a patient in the study but does not get exome. Would the physician's behavior be altered just by the fact that the physician knows that this patient is in a study where standard of care is compared with standard of care plus exome? Right, so the standard of care is actually pretty well set in our clinic. I mean, we really do use the algorithm that I showed. However, we did actually discuss this. And that's why we wanted them to dictate their first note with what their plan was before we gave them the results. But they should really be following the algorithm but not everyone does, I concede. So, and then, you know what Dave Vienster would say, I guess, is since we're just, you know, we're making the comparison within our arm even if people behave a little differently, they still don't have the exome and so that's the direct contrast we're making. So we're trying to minimize it. We can't get rid of it completely but he thinks within the economics that we're safe. It seems that, you know, you could go one step further and sequence the whole genome. You know, there are going to be a lot of instances of promoter mutations or alterations. You know, so especially when you're involving patients where you could get an answer, you know, more completely. Wouldn't that be a consideration? Right, and so we talked about that and it really was a cost issue that in order to be reasonably hard for the study, we felt we needed to get 220 people to do a genome just wasn't cost effective. It wouldn't surprise me if we end up crossing over to a genome during the course of the study. We'll have to do that carefully but as the costs drop we may, that, you know, I would prefer that to be honest but I think that exome will get us most of the way we need to go but I agree it won't get us all the way. So it's just sample size versus technology. Okay, further comments for Divi? Oh, for Gail? You can? Yeah. I can't go. Yeah, or whoever. Yeah, I would say the one thing I should have said for opportunities to interact, the one thing we're not doing because we couldn't afford it, which I would love to do is collect tissue on these patients at the same time. So if anyone wants to help us support that, I think it would be really, really interesting. Are you archiving? Sorry? Are you archiving tissues? That, you know, we're gonna try but we don't have any budget for it. So, yeah. Yeah, on the clear question, I can tell you about conversations that we've had sitting in a room full of NCI people with the OHARP, not with the OHARP but the CMS people. And they said, absolutely not that you, in other words, we've discussed with them whether we could write as eligibility criteria for research protocols, having to test positive on a, at some sort of biomarker. And they said, absolutely not if it's not a clear laboratory. Right, so if you intend to return it and use it for any kind of clinical use, including enrollment in a study, I think everyone agrees that needs to be clear certified. I think the real question is, when I pull my emerge sample out of the freezer and it's been sitting in the freezer for 20 years and the subject's not consented for return of results because no one even thought about it then, and I find something they actually should know. You know, it's not like I can call them back and ask for a new blood sample to run in a CLIA lab without telling them why I'm doing that. So can I, in that case, do it? I will say as a practicing clinical geneticist that we see patients frequently who come in and they say this is my result from a research study, now they say I need the test and then we run it in a CLIA lab the same test. CMS does not like that at all. But if you look at the CLIA law, it says very specifically there's a research exemption but they cannot be used for clinical care. So some people, but not everyone, think that if you take the research test, you can return it but then you have to get a CLIA test before you provide care to the patient. Based on that test, not everyone agrees. And so that's what we're trying to resolve but it's never been legally tested. So actually George Church's attorney wrote up a whole nice legal opinion of this for us which I'll be happy to shoot over to you. Okay, we ought to finish. Maybe this is a good topic for a breakout group. All right, thanks.