Standardizing Testing May Aid Sequencing-Based Test Reimbursement

Q&A with Kathryn A. Phillips, Ph.D. The translation of personalized medicine into clinical practice remains in its infancy. Many large challenges remain, including development of a standardized metric to evaluate the clinical utility and cost-effectiveness of molecular-based tests; ethical considerations surrounding return of results in a field where genomic interpretation is rapidly evolving; and establishing a reimbursement framework for large sequencing-based tests. Rather than explore sequencing from a technical angle, this quarter DTET examined the transition of sequencing-based tests into clinical practice from a policy perspective. DTET recently spoke with Kathryn A. Phillips, Ph.D., founder and director of the Center for Translational and Policy Research on Personalized Medicine (TRANSPERS) at University of California, San Francisco, which focuses on developing evidence-based information to guide the use of personalized medicine. Phillips discusses TRANSPERS' work on increasing transparency of reimbursement policies, developing approaches to evaluating the downstream economic impact of sequencing, and understanding variance in test adoption. TRANSPERS' Evidence and Reimbursement Policy Advisory Council brings together a variety of key stakeholders. How far apart are industry and payers from establishing requirements for evidence of clinical utility? It is really critical each party understands where the other party is coming from. My sense is that often that doesn't occur. Industry doesn't understand why payers won't pay and thinks they need to be more specific in requirements. Payers think that industry just does not provide the data they need and that test developers want them to cover anything new and exciting, whether it has been proven or not. Of course, in reality, the truth is somewhere in the middle. We are seeing these issues come up even more with sequencing technology, partly because it has received so much hype. There has been a lot of attention being paid to the good impact sequencing can have on health care. We are also seeing more patient and provider demand for it. We see this technological imperative taking place and there is no doubt that over time we will use panels and sequencing and in many cases it will be a great advance over single gene tests. On the other hand, it is very murky how to evaluate these technologies given their complexity and that they raise all kinds of issues for payers, such as the distinction between what is being done for research purposes versus clinical purposes. Sequencing is often used for research as well as clinical purposes, but payers don't usually pay for research. It is not their mandate. We have to figure out what part of the results are going to be returned to patients; what to do with information that could potentially lead to inappropriate or even harmful care or variants of unknown significance where we just don't know what they mean; and the tension between patients who often think they want to know everything, and the need to protect the public interest. We do not want to give people information that is going to cause harm, or risk, or cost without the relevant benefits. Is the health care system ready to adopt sequencing into clinical practice? Personalized medicine is here. It is moving forward and is not going away. The trend is an upward trajectory in terms of use and adoption. But it will be selective as to how and when. I think single gene tests will still have a place. The move to panels is taking place in many areas. That trend is moving forward. In terms of size of panels—the number of genes included–it will vary by clinical context. Every situation is unique and of course that makes it complicated. You can't have a one-size fits all policy. I think whole-exome and whole-genome sequencing is something that has some specific uses now, but this idea we are going to sequence everyone at birth, I think, is a ways off in terms of being a feasible proposition to consider. How big of an issue is test cost in adoption of sequencing-based tests? Sequencing illustrates what's happening in our health care system as a whole, and it brings certain issues to the forefront. Right now it is murky as to who is going to pay for what and when. For example, BRCA testing is moving to panels. Right now insurers generally do not have positive coverage policies to pay for the BRCA panels. That doesn't mean they won't cover them, but there is not clear guidance out there on which panels they will cover and when. How do we best evaluate the economic impact of sequencing-based testing? There is going to need to be a variety of approaches. For example, one could look just at what is the budget impact of using a panel versus a single gene test. That is more of a straightforward evaluation. Here is the clinical pathway and here are the costs associated with getting to a diagnosis. What is trickier is when you have to look at the downstream implications. That is what people tend to forget. Just because you do a test, give back results, and get a diagnosis out of it, it doesn't end there. There is a pathway of events that occur after you return results and that is where we don't have much data yet to support testing for these genes where we think they have a relationship with a condition, but it is not well demonstrated yet. Over time that will become clearer. But right now we are looking at things like where is the tipping point. If you do three single gene tests sequentially versus a panel with six markers, you can identify the tipping point if you assume the panel is providing more information. But what if the additional information actually costs money and does not have benefits? Do you still come out ahead with the panel? Over time we will increasingly reach the point where you might as well do the panel because when you look at costs you are better off. Adoption of molecular testing is non-uniform. How do physicians evaluate the benefit of one test over another? We recently looked at the evidence and indeed there is a lot of variation in terms of what they order, when they order, how they order. One key is going to be standardizing how these tests are ordered. That has got to be done by professional groups. It is not realistic to expect physicians to keep up with the field. A lot of companies are stepping into this void and providing standardized panels, or offering interpretation, or offering reports that tell a physician what to do. Right now it is really pretty chaotic. That will shake out over time and we will move towards a more standardized, guideline-driven approach, instead of the Wild, Wild West we have right now. Hopefully the move toward regulating laboratory-developed tests will provide some standardization and some consistency as well. From a policy perspective there have been calls for greater transparency related to reimbursement of molecular testing. Tell us how the Genetic Testing Reimbursement Registry will aid in these efforts. Right now there is no place you can go that synthesizes different payer coverage policies in a standardized way and that is not part of a for-profit, proprietary database. We have been funded by the National Institutes of Health to develop this registry to better understand what policies are out there; whether there are gaps in coverage; how the policies compare; do they have a positive or negative coverage decision; what evidence is cited in those policies; and how can one predict if there is going to be a positive or negative coverage decision. This is ultimately what industry is interested in. What factors lead to a coverage decision? Using this, they can develop better evidence focusing on the factors that will have the greatest weight. We are now just starting this project, but we are hoping that it will provide more transparency and understanding of coverage policy. What advice would you give to stakeholders in order to improve trust and transparency during this transition period? Industry needs to be sure that they are providing the data that demonstrates the value of their products; not just the technical benefits, but how it meets a clinical need. Payers are really under the gun to figure out how to evaluate these new technologies in a consistent way. Some payers are starting to establish frameworks. But a lot of these tests have been under the radar and payers have made ad hoc reimbursement decisions, as opposed to putting out policies, and I don't think that can continue as these tests emerge into clinical care.