 We have an interesting afternoon, and we'll go ahead and start. Sean Tunis, CMTP. Well, good afternoon, everybody. I do want to thank you all for inviting some unruly guests to your party who may not be speaking on exactly the same wavelength. And we'll see if we understand each other after the next set of talks. Just so you know a little bit more about my background, I'm trained as an emergency medicine physician, and then background in health services research and clinical research, then kind of went off the rails and worked in Washington as an aide in the US Senate for a few years, and then ended up as the chief medical officer at the Medicare program. So probably a slightly different pedigree than some of you all in here. And that's probably why what I'm about to say will seem as incomprehensible to you as a microassay photograph seems to me, which just looks like a bunch of gray and blue and stuff. So we'll just try as hard as we can. I do think it's important to try to have the conversation, because I have a little bit of a feeling listening to this that I'm just hoping that we're not doing one of those transcontinental railroad things where we're building from the east and the west and just hoping the golden spike actually falls in the middle. But I do worry a little bit that we are in some danger of ending up not completely aligned when it comes from some of the perspectives of the way the, well, certainly the payer community, but some of the regulatory community as well is looking at this. So I'm going to try to talk about developing evidentiary standards for determining clinical utility of genomic based diagnostic tests. And I think already in that title, it's clear to me from listening to folks today that we think of what the word standards means differently. So this is not the kind of standards that I think you generally talk about. So this is evidentiary standards in terms of the design of prospective clinical studies. And I also think the notion of clinical utility you'll see is quite a bit different. But we'll see where we go from there. So let me just start with a quote from a paper that I think, Ned, you might have been a co-author on. But it was from the EGAP working group and sort of an assessment of essentially the quality of evidence for clinical decision making and policy decision making in the field of genomic tests. Of most concern, the number and quality of studies are limited. Test applications are being proposed and marketed based on descriptive evidence and pathophysiologic reasoning, often liking well-designed studies, but advocated by industry and patient interest group. So kind of a fairly blanket grade of C minus for useful evidence across the field. And I don't know, Ned, you might say whether there's a feeling at EGAP that this has improved significantly in the last four years. But my general sense is that maybe there's still some view of a room for improvement. One of the things that you may have heard of, and I'll be talking a little bit from the perspective of comparative effectiveness research. How many folks in the room have heard the term comparative effectiveness research? Oh, that's very encouraging. Good, good, good. Well, there's a lot of ways to understand comparative effectiveness research. But one way that I would characterize the premise behind the entire enterprise is that gaps in evidence will be reduced by greater engagement of decision makers, such as patients, clinicians, and payers in selecting and refining research questions and developing study protocols. In other words, the notion is if you get decision makers more involved in the clinical research process from the beginning, there's less risk that the research that's produced won't be actually addressing the critical questions from the point of view of decision makers. That's a premise. Now I thought for this audience, you're unlikely to believe anything I would be able to say unless I could explain it at a molecular level. So I actually have come up with the molecular basis of uncertainty. And this reflects molecular biology as it was taught in 1981 when I went to medical school. So I apologize to those of you who know that this is somewhat amateurish. But let me show you the way this works. So you have your decision makers inside the cell here on the left, and then your clinical research enterprise out there in the extracellular milieu. And this is a bit exaggerated, but it's to make a point. So you have that little orange cloud of intellectual curiosity that drives clinical research. Its investigator-initiated studies leads to published evidence in that little purple oval that slowly diffuses towards the decision makers via KT1 that stands for knowledge transfer 1, but it's just meant to sound molecular and sophisticated. So that's a slow diffusion process. And then you have a field called health technology assessment, which is what Naomi Aronson and her colleagues do, which is packaging published evidence so that it's more digestible for decision makers. So that's a active transport mechanism via KT2, which was cloned in a lab, I think, at Hopkins. But anyway, that's the mechanism, accelerated transport. And then that goes to the decision makers who are coated with these little yellow balls, which are low affinity receptors for evidence. So that's called the knowledge translation blockade. But really, the real issue is that most systematic reviews and assessments actually identify, they basically say things like eGap said, which is we don't have the evidence that we really need to make good decisions. So that's the gaps in evidence. And then you have defective transport at KT3, which means that those gaps in evidence don't get systematically communicated to the research enterprise. So you have defective transport, accumulation of ignorance inside the cell, which is toxic over time. So this is actually the notion of comparative effectiveness, or what I sometimes call decision-based evidence making, where the knowledge gaps as perceived from the point of view of those who makes decisions are not consistently acted upon or don't become the agenda or the guiding force for the next generation of research. So hopefully, if you didn't believe me before, you're now convinced because it's a molecular model that's been published, I think, in the Asian Journal of Hospital Management, as far as I can tell. I think it's the only place this appears in the literature. So let me talk a little bit about it then. What is it the decision makers are looking for with respect to evidence of clinical utility in the domain generally of molecular diagnostic? So I'm not going to talk very much about the FDA regulatory regulation of in vitro diagnostics, particularly because there's a FDA person here. But I'm mostly using this for contrast to what the payers are looking for. So here you have they're looking for reasonable assurance that the benefits outweigh possible risks and reasonable assurance that the use of the device of the diagnostic will provide clinically significant results. So keep that in mind. So what there is not in here, you will notice, is something that definitively states some kind of evidence that suggests that health outcomes are improved, that patient health is improved as a result. This is reasonable assurance that benefits outweigh risks that could be thought of as health outcomes, but clinically significant results. So here, let's go and look at what Medicare says is the kind of evidence that they're looking for when it comes to diagnostic tests generally. So this is back from, I think, 10 years ago when they actually articulated their framework. But basically, this is not really much different from any other payer, which is two questions. Is there adequate scientific evidence to determine whether the test provides more accurate diagnostic information than an existing diagnostic? So is it accurate? And then the second question, which is the hard one, is if it changes accuracy, is it adequate to determine how that change in accuracy affects health outcomes? So looking actually for does the information change how the patient feels, something quality of life, morbidity, mortality, something of that nature. So that's kind of Medicare's view of clinical utility of a diagnostic. And then Naomi may talk more about this, but this is Blue Cross Blue Shield Association, one of their criteria for determining whether something is medically necessary. And I would point to the second paragraph, which is evidence should demonstrate that the technology can measure or alter physiologic change related to disease injury illness. In addition, there should be evidence or a convincing argument based on established medical fact that such measurement affects health outcomes. So again, there's an emphasis on does this information influence presumably through some kind of clinical management decision impact on health outcomes. So this is a slide that's kind of meant to illustrate that across multiple payers, you're going to see that they come to this determination of what's adequate evidence of improved health outcomes at quite substantially different times. And I don't really have a working pointer, but this is the rate of adoption or reimbursement for the Oncotype DX in node negative breast cancer. So it goes from 2004 all the way through 2009, 2010. And basically, at least the way this was compiled, Kaiser was the first payer group to actually adopt Oncotype DX, followed by a Harvard Pilgrim, Premier and a couple of Blue Cross Blue Shield plans. And then there was a sort of the percentage of the population that have covered lives that were actually had access to reimbursement for Oncotype DX sort of sat stable for a while until the Medicare contractor, a local Medicare contractor called the NHIC, it's the National Health Insurance Corporation, which is the California Regional Contractor for Medicare decided that they would pay for Oncotype DX. And because the company that does the tests is in California, that effectively became the national policy for all of Medicare. And once that happened, I think what you're seeing here is that's the domino effect, because once Medicare is paying for something, it's virtually impossible for private payers to no longer pay for it. But this has a laser on it. So following that, you have Aetna here, and United is there, and Reed, for whatever it's worth, Aetna beat United. So at least in this race, whatever that's worth. Anyway, and Naomi, I'm not sure if you know when was it that the Blue Cross Blue Shield Association technology assessment? But it was somewhere closer. Oh, sorry. I'm not sure. Eh. OK, so probably over in here somewhere where it was sort of determined formally to have met the criteria of the Blue Cross Blue Shield Association. But as you can see, there were some blues plans that were already paying for it down here. But really the point of this is just payers are different in terms of when they decide that the scientific evidence is compelling about clinical utility. So it suggests that there's not a great uniformity across different payers. And I couldn't resist actually giving you a little bit of insight. This is actually one of my favorite slides, maybe the only one you should remember, which is actually how the National Health Insurance Corporation, the California contractor, made their decision on ankyotype DX, my friend Bruce Quinn, who was the carrier medical director at the time, made this slide. So basically there were some meetings with genomic health, with CMS folks. This was the New England Journal paper, the first major paper, I think, on the prospective retrospective study of the clinical utility of ankyotype DX and no negative breast cancer. Bruce looked at all this information, as well as, I think, an early Blue Cross Blue Shield tech assessment, and said, OK, we're not going to pay for that. That was down here in early 2005. The response to that was the California branch of the Clinical Oncology Association went berserk. And going berserk meant they went to Sacramento. And what also happened is that Medicare coverage decisions can be appealed to administrative law judges, lawyers who can review. And every single one of the denied claims was overturned by an administrative law judge, a lawyer, or a judge. And so Bruce decided at this point, what's the point of having a non-coverage decision? So he changed it to a coverage decision in early 2006. And that's what launched that entire cascade of the rest of the private payers following along. So what I'm really trying to illustrate for you is there's some work to be done in terms of getting our arms around a genuine scientific evidence-based framework for what constitutes sufficient evidence of clinical utility in this domain. And I'm mindful that I don't want to go too far over, but let me try to do a couple more slides. This was around, I think, well, several years ago, the Secretary's Advisory Committee on Genomics, Health, and Society, which was doing a lot of work around personalized medicine. And Reid, I think you might have been the chair of this committee when this report came out. So you got the right people in the room. One of the recommendations in a report that this committee did on coverage and reimbursement of genomic tests was information on clinical utilities, critical for managing patients, developing guidelines, making coverage decisions. So HHS should create a public-private entity of stakeholders to establish evidentiary standards and levels of certainty required for different situations. So the notion was what we ought to do here is get the test developers together with the payers as well as the professional societies and patient advocacy organizations and actually try to come up with some guidance, some description of the evidence necessary to define clinical utility. As far as I know, Reid, this never happened. There isn't anything of that type that's moved forward. But at our center, we thought that was an interesting idea. So we're trying to do it in a kind of a private sector way. So we're a private independent nonprofit that is now trying to work on developing guidance for how to design prospective studies of the clinical utility of, in this case, we're using molecular diagnostics and oncology, and that would reflect the information needs of patients, clinicians, and payers. So by talking to patient groups, bringing in the ASCO, the other relevant professional societies, as well as Medicare and private payers, the notion is, can we come up with something that's analogous to a regulatory guidance document? And for those of you who are familiar with this, if you want to get FDA approval, you can generally find a guidance document that would say, well, except for the case of in vitro diagnostics for and in this space, there's not yet a final guidance. But generally, you can find an FDA guidance that says, if you want to get regulatory approval for an asthma drug, you have to study it in this kind of population compared to this alternative intervention and measure these outcomes. What we're trying to do is develop a comparable guidance document, but that's written from the perspective of what kind of evidence you should bring forward for purposes of reimbursement. And the notion would be to try to describe studies that would provide a reasonable level of confidence of improved health outcomes. Several folks in the room are actually on the technical working group, the group of experts and clinicians that's developing these recommendations. And so actually, you can see the names of a number of the members of the technical working group that's helping come up with these study recommendations for molecular diagnostics and oncology. And just to finish up, we're jointly hosting with the Institute of Medicine Roundtable on Genomics later this month a kind of a multi-stakeholder meeting to present the draft recommendations for study design, for input, refinement, et cetera. But what we're trying to do here is pretty much what the Secretary's Advisory Committee suggested, which is to get stakeholders together and define the kinds of evidence that would be necessary to establish clinical utility. Hopefully those would become widely, at least consulted in the process of developing these tests, and would reduce the amount of variability in terms of how, reduce the uncertainty in terms of how test developers would need to go about demonstrating clinical utility. So I'll stop there. Thanks very much. And then. Questions? Mark. Mark retained Chicago. So that was very helpful, but at least in the context of germline genotyping and sequencing, I think CMS has it wrong. And let me tell you what my logic on this is. There's as much utility demonstrated for genotyping and sequencing now as hard evidence as there is, for example, for electronic health records. And CMS, of course, is requiring EHR for all practical purposes. In fact, paying institutions because of the belief that there's long-term value here. And I think you can make the same case for genotyping and sequencing. You have upfront costs to develop the infrastructure. You have costs per patient, one-time costs, essentially. And then lifetime value. And I'm not talking here about Oncotype DX and tests like that, which don't fall under that construct. But has there been any consideration at CMS for sort of looking at this kind of germline testing in a different context than a traditional diagnostic? Well, just a comment on the premise that one would assume some consistency from CMS in terms of policymaking. So whatever they do in EHRs, there's no reason to think that they would be consistent in terms of what they did about around genomic or genetic diagnostics. But the answer to your question is no. I mean, of course, they haven't thought about that. And part of it is in the same way that 95% of the Medicare agency is devoted to processing bills and paying claims. It's not really a policymaking organization for the most part. But you have the problem, which is stuff doesn't get paid for unless they decide to pay for it. So my point here wasn't to defend what CMS or Blue Cross Blue Shield or Etna or anybody else does. My point here is you all have the technical knowledge and the understanding of this field to sort of be able to propose what ought to be an intelligent, sustainable framework for thinking about what's clinically useful. And you're going to have to go to CMS and the private payers and have that negotiation and discussion yourselves. It's in the same way that there's an ongoing dialogue and interaction with the FDA around what the regulatory framework is. And the FDA may be a few years behind the cutting edge of technology, but however many years FDA is behind, I can guarantee you Medicare is twice that behind. And the private payers are probably somewhere in between, certainly better than Medicare. Again, all that's doing, I think the point you're making is you may very well know better what would be an appropriate framework for identifying what are clinically useful, important tests in this field. But there's not going to be some magic way that Medicare or any of the private payers understand that unless there's this dialogue and communication. Did that at least address your? Well, I'm trying to distinguish the one-off diagnostic test ordered by a physician for a particular patient for a particular clinical indication versus more of what many of us are talking about, genomic medicine, sort of preemptive sequencing or genotyping, which, as I said, is very similar to the EMR and the basic concept that there's long-term value here. And I guess the question is, how does one elevate the debate around this to the same level to get at the same visibility that the EHR has gotten to the point that actually institutions and physicians are being paid to do it, as opposed to being told they can't do it and nobody's going to pay for it? Yeah. Well, I mean, it is certainly true. And maybe Joanne or Reed will either want to pipe up now, because with Reed having chaired the committee, maybe this came up at the Secretary's Advisor Committee. But the framework for reimbursement decision-making for the most part in payers is around kind of individual specific technologies and services, and not in this kind of broad what you're describing. So I think the points of dialogue around that are probably going to have to be at senior levels in the administration, in HHS, possibly in CMS. But I honestly don't think that they're not really equipped with the policy mechanisms that would really enable them to do that at the Medicare agency level. I don't know, Reed, did you want to weigh in on this? You got a? OK. I'll just give you very two concrete answers. First of all, CMS is not paying for the electronic health records. DHHS is through HITECH and ARA. So it's not specifically that CMS made that decision. So that's one distinction. The second thing is that Medicare has to operate under its enabling legislation. And there's nothing in the enabling legislation that allows it to account for this type of an application. It's all on a test-by-test basis. And so even if they said this is a good idea, my understanding, and Sean, you can correct me if I'm wrong, is they would not even be able to consider the question because it's not reflected in the enabling legislation. Yeah, I mean, I think that is probably right. I mean, it would be on the order of a question like reimbursement for telehealth. It's sort of a body of related technologies. And it turns out that Medicare doesn't have the statutory authority, the legal authority, to pay for telemedicine. So that has to either be addressed through a legislative mechanism or possibly through a regulatory or a mechanism or a demonstration project. So it probably is more relevant there. So one way or another, again, the main point of the presentation was that in terms of a framework for clinical utility that fits what you all are talking about, I think we're a long way off from having something that's a hand in glove there. I think we'll move on, Ned, if we can have the other end of standard discussion.