Steven Peskin, MD, MBA, FACP: In the absence of a biomarker to guide selection, how do we formulate coverage decisions? Well, again, in the majority of situations, we do not have biomarkers. So, in the majority of situations, we’re looking at situations in cancer where there are established protocols or established regimens.
These regimens are defined by the National Cancer Institute. They’re defined in refereed, peer-reviewed journals, with clinical trials that demonstrate that this particular chemotherapy, or immunotherapy, or targeted therapy, is effective in a particular situation (in a particular type of breast cancer, for example). So, certainly, our guide prior to the emergence of biomarkers and molecular diagnostics has always been, and continues to be, looking at peer-reviewed literature, looking at evidence for the effectiveness or relative effectiveness of particular regimens. We leave a lot of this in the hands of our clinical partners in terms of the decisions that they make as it relates to a particular regimen versus another. We certainly have looked at National Comprehensive Cancer Network (NCCN) Guidelines as another yardstick that we use, if you will.
[When making coverage decisions about companion and complementary diagnostics for PD-L1 expression for] 2 particular immunotherapies in the treatment of non—small cell lung cancer, certainly, the FDA approval will be a major decision point for us. FDA approval would naturally move toward that being an appropriate or approved therapy from the standpoint of any kind of prior authorization or evaluation for medical policy. Then, beyond that, we’re going to be looking at real-world evidence. That evidence that’s generated here in New Jersey or elsewhere, and that longitudinal look at the effectiveness or relative effectiveness over time, would be really the ways that we would go. Further, as we’re working more deeply with our clinical partners in what are called “episodes of care” or “bundled payment models,” our clinical partners have some skin in the game, so to speak, to look with us and evaluate with us the relative effectiveness and make wise, appropriate, and effective, or value-based, treatment decisions.
[Is there] a level of data that provides information around false positives and false negatives in diagnostic testing? The short answer is, no. I can’t give you a particular number in that regard. Again, we’re going to be looking at the FDA as a key regulatory agency to guide us. We’re going to be looking at the NCCN as an authoritative source. We’re increasingly looking at some of these value frameworks—the Drug Abacus and the American Society of Clinical Oncology Value Framework. These are all guides that we use. But we don’t have any specific [data]. I can’t give you a number on that.
[How is coverage managed] around first- versus second-line treatment, and where a companion diagnostic might be associated with a particular tumor type [or] particular cancer? Where oncologists, as they have for many decades, [consider] off-label [uses] of FDA-approved drugs? We are going to, sometimes, look at that on a case-by-case basis. Other times, again, in the context of cancer care, if it’s not clearly inappropriate, then it’s going to be clinical judgment that’s going to dictate and determine how that might be used—maybe in a similar type of cancer where there are indicators that that particular pathway is as relevant in kidney cancer as it is in melanoma, for example. So, those kinds of decisions are getting into judgment and some of the emerging science.