Liquid Biopsy Use in Metastatic Non-Small Cell Lung Cancer - Episode 3

The Role for Liquid Biopsy in NSCLC

Transcript:

Sandip Patel, MD: With the community oncologists that I work with, the NCCN recommended panel for non—small cell lung cancer is often utilized for testing. This includes EGFR, ALK, and ROS1, as well as now PD-L1 immunohistochemistry. I think these panels are particularly helpful, because they capture the majority of patients who have driver mutations. However, there are a variety of different mutations that can be detected both by liquid biopsy approaches as well as tumor-based NGS. These are aberrations in HER2, BRAF, and MET exon 14 skipping, and a variety of other rare mutations that may show sensitivity to targeted therapies as opposed to chemotherapy. And so, in general, my experience has been that in the community setting, these tests are typically done: EGFR, ALK, ROS1, and PD-L1. And, often for a never-smoker, a patient with a unique history, they’re often referred to the academic setting for a more broad-based profiling.

Geoffrey R. Oxnard, MD: The thing we don’t talk about all the time is how tricky it is to get genomics. And, how often do we meet a patient and we aren’t sure of their genotype? We have tried to get tissue. We have tried to get results for them, and it hasn’t panned out. These are patients who maybe have a tumor specimen at diagnosis that’s too small, and it has been exhausted just figuring out what kind of cancer it is. Maybe they have a hard-to-biopsy lesion that’s in the bone, primarily, or in the brain. Or, maybe you’ve gotten the tissue and it’s just necrotic and it’s fallen apart and it doesn’t produce the results you want. And so, I routinely see patients where I need a result that day and the tissue I have available is not adequate for getting me genotyping. That’s where I think about liquid biopsies or noninvasive testing as an alternative to tumor genotyping—trying to get me that genotype and get targeted therapy options for our patient.

Sandip Patel, MD: There have been several studies that have assessed the concordance between tumor-based sequencing approaches to cell-free-based DNA approaches in cancer. In particular, the specificity, meaning the ability for a mutation that’s detected in liquid biopsy to find a concordant mutation in the tumor, is relatively high. However, there are discrepant results between the sensitivity of these assays, liquid biopsy assays, in terms of detecting mutations that are present in the tumor. Now, this bias may work in both ways, depending on the tumor type that’s assayed. For tumor-based NGS approaches, oftentimes this may be related to very small biopsies done by endoscopic ultrasound or endobronchoscopic methods, which involve very small amounts of tumor. That assays a specific part of the tumor within a wider, more heterogeneous tumor.

And so, whether or not that represents the gold standard, necessarily, in many of these retrospective studies is not entirely clear. However, for a mutation that is detected by cell-free DNA, the specificity in terms of finding that mutation on a larger biopsy is quite high. The sensitivity for cell-free DNA tends to vary by the stage of disease, with improved detection for more advanced stage disease with tumors with higher turnover. However, for patients with lower stage disease—for example, stage 1—the ability to detect DNA that is shed into the blood is lower. And so, for those patients, while an initial cell-free DNA strategy may be preferred to avoid an unnecessary biopsy, the overall yield may be lower.

Geoffrey R. Oxnard, MD: When I’m thinking about for whom to send a liquid biopsy, the question I am thinking for an individual patient is, does their cancer shed DNA? We are looking for DNA that is effectively a needle in the haystack—the haystack being all the patient’s germline DNA, and the needle being the tumor DNA floating around there. There are certain characteristics that are associated with DNA shed. What we found in our study so far is that the more metastatic sites of disease the patients have, the more DNA is shed. If they have liver or bone metastases, there is more shedding of DNA versus, if this is a lung cancer patient with small lung nodules—asymptomatic—the chance of that being a shedding cancer with a productive liquid biopsy is smaller. And so, this is something we each need to really develop a sense about as clinicians, which is, in which patients is this tending to be a useful assay? In which patients is this negative, and falsely negative, because there was simply no DNA in the specimen for us to assess?

If you think about a biopsy, there is a pathologist looking at that specimen telling you, “Yes, it’s got tumor in it, this is a good one for testing.” But when you do a blood test, there is no one telling you, “We’ve got good tumor DNA, this is adequate.” And so, you have to trust, is this a cancer that’s shedding DNA? Because if it’s negative, you’re falling back on tissue testing, the standard, as our backup test, because sensitivity is not perfect with these assays.

Transcript Edited for Clarity