Health Plan Coverage of Diagnostic Tests for PD-L1 Expression

Michael Kolodziej, MD: The current state of affairs in immunotherapy biomarkers is that pembrolizumab has a companion diagnostic and nivolumab has a complimentary test. And our current coverage policy does require at least an attestation that a companion has been done for prescribing pembrolizumab, and there is no such requirement for nivolumab. I am quite well aware of the massive controversy about interchangeability of IHC tests, different antibodies, different cutoff points, etc. I can assure you that there is no health plan in America that has the bench strength to go in there and analyze each of the individual assays, antibodies, performance characteristics. It doesn’t work that way.

And so, we look to the guidance of the thought leaders in the area to help us identify what test and when, what disease, what cutoff, and how to use that information. The truth of the matter is, I’m not very confident that IHC is ultimately the answer, and there are a lot of people working in the space trying to find alternative biomarkers that will be either predictive of response or a lack of response. And if I come back and talk to you next year, and we’re still talking about an IHC, I’ll consider that a failure.

The answer regarding whether or not the health plan views the level of evidence adequate to write it into coverage policy: in the case of the companion diagnostic for pembrolizumab, yes. Because, we have accepted, as we do, the expertise of the FDA in approving that companion diagnostic test. In every other case, where we’re talking about IHC for clinical decision making in the treatment of patients with immunotherapy, the consensus is that there’s just not enough evidence to use it to write policy. Because, remember, policy is designed to say, “That’s appropriate use, that’s not appropriate use.” It’s a yes or a no. It’s not a maybe. And I think the truth is that for most IHC tests right now, what you get is pretty much a maybe.

So, the specific clinical characteristics, in which biomarker testing is something embraced by the payor, is largely dictated by the package insert, if you will. We do know, for example, that there is some evidence that there may be a tumor-specific variability in terms of IHC expression and the benefit in terms of predicting whether or not immunotherapy will work. Like I said, we don’t have the bench strength to figure that stuff out. We look to the regulatory agencies and the clinical community, NCCN, to help guide us in terms of our clinical policy.

And I might add that in all complex testing in oncology, by and large, the payors require attestation. That is, we ask the doctor if they did the test. We expect them to be honest with us. To the extent that we can, we might look for a bill for the test. But it’s really, really hard to know that the specific antibody used for that specific patient is the one that was used in the registration trial as part of the companion diagnostic piece. We know that there are LDTs that are being substituted for the companion diagnostic that may not have the same performance characteristics that people are making decisions based on. We may not like it, but there’s not a heck of a lot we can do about it.