David Hyman, MD: There are multiple technologies capable of detecting TRK fusions. You can look for TRK fusions at the DNA level, and that’s what most next-generation sequencing tests do. You can also look for expression at the RNA level, and there are now commercial-grade assays that can detect at the RNA level. You can also look for expression at the protein level with immunohistochemistry. On top of that, there are some orthogonal assay approaches like FISH, which have been used historically to detect other kinase fusions like ALK and can be deployed in a directed manner.
The really important message here is that there is no single perfect test for detection of TRK fusions. There are particular challenges with detecting TRK fusions in a purely DNA-based test. These fusions typically involve the introns of genes; the introns of NTRK 2 and 3 are quite large, and that can make it difficult to detect TRK fusions. RNA sequencing is, of course, an excellent way to detect TRK fusions, but it doesn’t have the same penetrance yet in the marketplace that DNA-based tests do, and they can’t necessarily detect other coding mutations that we can readily detect with DNA. Immunohistochemistry is also potentially a very good test, but because of the number of tissue types involved, there is always a possibility of normal TRK expression in certain tissue types that is not caused by an underlying TRK fusion.
We’re really in a situation where the physician needs to be aware of the test they’ve ordered and whether it has the ability to detect a TRK fusion. What I can really say at this point is that if they want to know, they may have to pick up the phone and speak to their pathologist or look on the website of the company. If they have a particular preferred vendor, they need to do that only 1 time. I think it is important because with some of these tests, physicians are asked to fill out these requisition forms and there are multiple checkboxes. They need to find the checkbox that says, “I want to look for fusions in my patients,” depending on the assay. Unfortunately, this is a diagnostic area that is going to require education of the oncology community over time.
David S. Hong, MD: I think there’s debate right now in the community about how valid and cost-effective widespread testing is. There are some data out there reflecting that there are only a handful of truly actionable molecular alterations. But I think what that criticism forgoes is the fact that we are making progress almost every year in this area of precision medicine. If larotrectinib is approved, it will hopefully be one of many other possible tumor-agnostic precision indications that will come in the next several years. I think as we move forward in the context of precision medicine and oncology, next-generation sequencing or profiling will become increasingly more important.
How will NGS impact rare tumors versus more common tumors? I think it will definitely have a much more direct impact with rare tumors. Often, these rare tumors do not have options other than surgery or maybe some radiation. They often do not respond to chemotherapy, and if these alterations are present and actionable, this may be a really good option for these patients if there is evidence that they have activity in the clinical trial setting. For more common cancers, I think it will just depend on what the data show in further studies down the line. Obviously, for many common cancers—for example, lung—we now know that immunotherapy plus chemotherapy has significant advantages and benefits in these patients. If the patient has an NTRK fusion, would I give them an NTRK inhibitor before that immunotherapy combination? Probably not. But if they had no other alterations and they progressed on that combination, I would probably prescribe larotrectinib right after that.
So much of this is really tumor dependent and involves to what extent first-line, second-line, and third-line therapies impact overall benefit relative to some of these agents such as larotrectinib. With that said, larotrectinib, even given the small number of patients we’ve had in lung and in colorectal cancer, has shown durable responses and durable benefit in those patients who do have NTRK fusions. Even in aggressive tumors like pancreatic cancer, we’ve had only a handful of those patients, but they’ve had significant and durable responses and benefit.
David Hyman, MD: What I could say is that one of the features we’ve seen with patients whose tumors harbor TRK fusions is that these fusions appear to be clonal events. What I mean by that is that they’re present in every cancer cell at every time point, at every site. And as a result of that, it really doesn’t matter which test you use and when you test. If you can test the patient’s primary tumor, you can test the metastatic site. What’s really critical is that you perform the test at a time when the turnaround time doesn’t prevent the use of that information: Can a TRK fusion or other actionable alteration be identified?
Transcript Edited for Clarity