Selecting a Molecular Panel in NSCLC

Transcript:

Jared Weiss, MD: There are multiple commercial molecular platforms out there, and as practicing oncologists, we’re very busy. We don’t always want to delve into the details of particular tests, but unfortunately, it is actually quite relevant toward the results you’re going to get and how you’re going to use them for your patients. When looking at tests and comparing them to one another, what you need is a test that has good operating characteristics, good sensitivity and specificity, especially for the common FDA [US Food and Drug Administration]-approved actionable changes. You need a breadth of testing where you’re going to cover everything that you can target with an FDA-approved therapeutic, that you believe to be actionable through an off-label drug, or that you have availability of a clinical trial for. This includes thinking about the patient’s ability to travel, which would induce a need for broader testing.

There are many commercial platforms out there that do this well. The most commonly used test in the United States is probably the Foundation Medicine test. It’s a good panel. It has good sequencing. It has good bioinformatics. It catches all of the common stuff and most of the rare stuff. I do like the way that they format their report. They format their report in a way that helps you to know what to do with it and make it easy to find additional information.

Gregory J. Riely, MD, PhD: One thing I like a lot about tumor mutation burden testing in patients with non—small cell lung cancer is that I can get tumor mutation burden along with all the other molecular markers I care about. In the test that I use to identify patients with EGFR mutations, ALK gene rearrangements, and ROS1 rearrangements, we get tumor mutation burden as a consequence of that. And so, it’s all rolled into 1 assay, which I think is really appealing.

Jared Weiss, MD: Another question that arises is the turnaround time for next-generation testing. In reality, that turnaround time is about 3 weeks. That’s a long time for a lung cancer patient. In 3 weeks, lung cancer can grow causing symptoms and even risking death, so it’s a big deal. There’s a certain friction in clinical practice between wanting to do something right away to protect the patient and wanting to wait and do the right thing. I think there are tough decisions that a patient and doctor have to make together around this. In my practice, the 2 core elements that go into that conversation with the patient about whether to treat right away or wait are the risk level of the disease and the probability of finding something useful on testing.

At one extreme, for a patient with bulky central chest disease causing respiratory symptoms and pain, I might feel that I can’t wait or that I need to treat right away. Then, I reserve the right to change things as the molecular report comes back. But in contrast, a less symptomatic patient is this other extreme, maybe with a high probability of having an actionable molecular change: for example, a never-smoker. Most of these changes—not all of them, but most—are much more common in never-smokers. Or, for example, an EGFR mutation, the most common actionable change, is statistically far more common in Asian women than in men or people of a non-Asian race. If you have those characteristics with a high probability of finding something actionable, and I have a patient whose disease status is such that I think they can afford to wait a little bit, then it’s absolutely worth waiting for all the human advantages of targeted therapy when available.

Greg Riedlinger, MD, PhD: In using tumor mutation burden as a predictive biomarker, one advantage is that it has an absolute score. There are different algorithms at Rutgers CINJ [Cancer Institute of New Jersey]. We’re typically calculating tumor mutational burden, or tumor mutational burden is calculated from a Foundation Medicine test. We base it on that. We have absolute, well-defined cutoffs with tumor mutational burden from Foundation One and good results, or clinical data, supporting who is potentially going to respond or not respond based on other tumor mutational burden scoring.

The turnaround time for tumor mutational burden is typically about 2 weeks. So, we’re sending for this Foundation One assay, and it’s typically a turnaround time of about 2 weeks; whereas PD-L1 [programmed-cell death ligand 1] immunohistochemistry is somewhere closer to 3 days.

Transcript Edited for Clarity.