Tissue Versus Liquid Biopsies in BRAF+ NSCLC

Transcript: Joshua M. Bauml, MD: Any patient who is diagnosed with nonsquamous non—small cell lung cancer should have molecular testing performed. The reason for that is we know that targeted therapies for patients harboring such molecular alterations are the best approach for the treatment of those cancers. It’s associated with the best response rate [and] the best quality of life, and we’ve seen improved progression-free survival across multiple studies when compared with chemotherapy. So this is the best approach for patients. To get those patients on these treatments, though, we need to get the tests done.

There are multiple ways to identify molecular targets. Historically, we did individual single-gene assessments—things like PCR [polymerase chain reaction], and FISH [fluorescence in situ hybridization], and immunohistochemistry [IHC]. The problem with such approaches is that each time you do 1 of those tests, you use a little tissue and a little time. Patients with lung cancer do not have a lot of either of those.

In my mind, the most tissue-efficient approach, particularly given the large number of molecular drivers that we have to test for, is the use of next-generation sequencing. Personally, at Penn [the Perelman School of Medicine at the University of Pennsylvania], we use a DNA-based next-generation sequencing platform to identify mutations, and we use an RNA-based multiplex PCR approach, a so-called Archer-based platform, to identify fusions.

We do this on all patients with nonsquamous non—small cell lung cancer. For patients with squamous disease, it’s a little more controversial. However, it’s important to remember that if you have a patient with squamous non–small cell lung cancer who has an otherwise atypical history or a light smoking history or maybe is a very young patient, he or she is a patient for whom I will do testing. Because remember, the biopsy needle is a small needle in a large tumor. It can sometimes have sampling error, where things look like squamous disease but are, in fact, largely adenocarcinoma, and you can miss some molecular target and the opportunity to help that patient.

Circulating tumor DNA, or so-called ctDNA, can be identified in a plasma test. It has a lot of appealing options. You don’t need to put the patient through another biopsy, if the biopsy was inadequate, to identify molecular drivers. If you identify something off a ctDNA assay—and those tests, by the way, can either be single-gene assays, even PCR, or they can be a more comprehensive next-generation sequencing platform—it’s real. You can act on it. If you identify an EGFR mutation, you can start them on a TKI [tyrosine kinase inhibitor]. If you identify a BRAF V600E mutation, you can start them on TKIs. The problem becomes if you don’t identify something. The negative predictive value, so to speak, of plasma testing is not very good. Basically, what I say is, “We’ll do this test. It often comes back quicker than the tissue test. But if it’s negative, it’s really important to go back to the tissue.”

And if the tissue was inadequate, I will often send patients for repeat biopsies to identify the sources. It’s also critical, when doing those biopsies, to avoid bone biopsies. The decalcification process can degrade the DNA and can lead to false-negative results.

I think both tissue and liquid can provide valuable information for the diagnosis of molecular drivers in non—small cell lung cancer. And I think that for the appropriate patient, we may use both. If a patient has a positive result on either tissue or liquid, I feel comfortable acting on it. If the biopsy was inadequate from molecular testing, then doing the liquid test is better than not having any testing. But if it’s negative, it really is important to go back and get another tissue biopsy, because the liquid testing, the negative predictive value, is not terribly reliable.

Transcript Edited for Clarity