David Hyman, MD: The data with TRK inhibitors and with other tumor agnostic biomarkers, such as checkpoint inhibitor therapy, in patients with microsatellite instability suggest that to me that to deliver optimal care to patients these days, we should be offering them a broad next-generation sequencing test at least once in their course of their treatment. And, of course, because these tests can sometimes take time, if we want to be in an optimal position to utilize the information from this testing, it does behoove us to do this testing early in the course of their diagnosis.

We know that all patients with metastatic disease will, unfortunately, progress on therapy, and so once you have a patient in that situation, that’s when I consider sending for next-generation testing. Because we believe these alterations are early and clonal alterations, we can really test any tumor material available on these patients, be it a biopsy for metastatic disease or even a primary tumor if that’s all that’s available. And I think we can feel comfortable that the results, at least as it pertains to TRK, from that testing will be relevant to the patient throughout the course of their treatment.

David Reardon, MD: I recommend next- generation sequencing be done as soon as possible when a patient is diagnosed with cancer. It’s important to get access to that tumor tissue and take advantage of the state-of-the-art technology that’s routinely available now, to learn as much as we can about what may be unique and important to the biology of that individual patient’s tumor.

If we wait until the tumor has recurred, the biology may change, and we may learn some additional important information that has evolved as the tumor has recurred. So it often is worthwhile to consider obtaining a tumor sample at the time of recurrence to conduct an additional analysis, next-generation sequencing analysis, of the tumor. But I think it’s very critical to be thinking about obtaining these important data, performing the next-generation sequencing analysis, as soon as possible when a patient is initially diagnosed with cancer.

David Hyman, MD: So one of the areas certainly of confusion in the diagnostic space around sequencing in general and TRK in specific is whether the individual platform that you may order for a patient is a platform capable of detecting TRK fusions, and really what is the sensitivity for the detection of TRK fusions? In my mind, broadly speaking, I put this in a couple of groups. We know that, in general, plasma-based testing, the so-called liquid tumor biopsies, can be very useful in detecting point mutations. We do know that the sensitivity for the detection of fusions is relatively lower. And I do think that it doesn’t represent the optimal technology platform currently to identify TRK fusions.

In the tissue testing space, broad next-generation, DNA-based testing platforms are certainly capable of detecting TRK fusions. But, because of some specific characteristics around NTRK, which really have to do with the size of the genes as well as the number of different fusion partners that can be involved in TRK fusions, purely DNA-based testing, although again it’s capable of detecting TRK fusions, will miss patients who have productive TRK fusions and would benefit from TRK inhibitors.

So, really, the optimal testing strategy for the detection of kinase fusions in general, including TRK fusions, is tissue-based RNA sequencing. And there are certain platforms like technology offered by a company, ArcherDx, that have the ability to detect any TRK fusion in a patient, regardless of what the upstream partner is.

There are several other diagnostic approaches for the detection of TRK fusions. These include things, such as FISH [fluorescence in situ hybridization] or immunohistochemistry, and obviously these are single-analyte tests that you’re looking specifically for TRK fusions. But we actually have, in our practice, found that TRK immunohistochemistry, which is a very simple and relatively inexpensive test, can be a useful and rapid screen for the presence of TRK fusions, especially when it’s applied selectively. And this can be, for example, in a tumor type that has a high prevalence of TRK fusions. An immunohistochemistry test could be a very rapid way to suggest that a TRK fusion is present.

In addition, if a patient has had a test like a liquid tumor-based test or a tissue-based test that is a sequencing test that doesn’t include the ability to detect TRK fusions, if that testing does not identify a driver alteration—for example, if you have a patient with lung cancer who doesn’t have ALK, or ROS, or BRAF, or KRAS, or EGFR—that would be a patient who you might secondarily screen, either by one of these RNA-based platform tests or through a single analyte test like immunohistochemistry or FISH. So there you’re really heavily enriching your screen for the patients that are most likely to harbor TRK fusions.

I think one diagnostic paradigm for TRK fusion detection is to apply testing early that’s capable of detecting TRK fusions in all patients. And the other is to use a step-wise approach where you do an initial screen and a more limited test, and then you selectively screen with a more focused test, those that have so-called driver-negative cancers. So I think there’s going to be a lot more literature in the pathology community with these diagnostic pathways for the detection of TRK fusions. There will be nuance depending on the prevalence of TRK fusions in that individual cancer type and the environment in which the oncologist is practicing. So there’s unfortunately no way to get a really simple answer to this question.

Transcript Edited for Clarity