Panelists: Luis E. Raez, MD, FACP, FCCP, Memorial Cancer Institute; Benjamin P. Levy, MD, Johns Hopkins Kimmel Cancer Center; Philip C. Mack, PhD, UC Davis Cancer center
Benjamin P. Levy, MD: The gold standard for identifying EGFR mutations has been tissue for years. Now, we increasingly have new platforms that are plasma-based in which we can identify many of these mutations. There’s now an FDA-approved cobas liquid EGFR test that can be done. Tissue still remains the gold standard for identifying mutations in a treatment-naïve patient. But, I think there are clinical scenarios where we can consider using plasma when you’ve done the biopsy, you’ve made a diagnosis, but you don’t have enough to do the molecular testing. This is where you may consider doing a plasma test and not a tissue test—or re-biopsy. You’d also certainly consider doing a plasma test in the resistant setting—patients who are on an EGFR TKI who are developing resistance—to identify that relevant actionable T790M mutation.
There have been some very good data looking at the specificity of plasma against tissue for EGFR in treatment-naïve patients. That’s quite good—the specificity is as high as 100%, meaning if it’s positive in the blood, there’s a 100% chance it’s positive in the tissue. So, things are changing. We haven’t sorted out how to use plasma in the treatment-naïve setting yet, but there are more and more scenarios that I’m seeing where plasma is being incorporated into decision making upfront.
Philip C. Mack, PhD: Regarding tissue versus plasma testing, the gold standard remains tissue testing for treatment-naïve patients. When they’re first diagnosed, you absolutely have to get a tissue biopsy if it’s at all possible. This is required for diagnosis, and it gives you a very good shot to identify molecular markers and perform PD-L1 staining, which can’t be done in plasma. However, if tissue is not available, a biopsy cannot be performed, or there’s not sufficient tissue remaining from that biopsy, then it is reasonable to do a liquid biopsy in a treatment-naïve patient.
What you have to remember with a liquid biopsy is that in some cases, it will miss the identification of some molecular markers. Currently, the sensitivity seems to be running between 75% and 85%. So, that if you know that the tumor is positive for an EGFR marker and you run a plasma analysis, you will see it in a treatment-naïve patient 75% to 85% of the time—depending on the technology that you’re using. You have to be cognizant of the fact that you may miss a mutation if you’re doing plasma testing. This is particularly true for very slow-growing indolent tumors, which may be the cases when tumors are EGFR mutant-positive.
If you are doing a liquid biopsy, you’re doing analysis in circulating tumor DNA in plasma, and if you have a positive result—for instance, you identify the presence of an EGFR activating mutation—then you can consider that as trustworthy, and you can proceed to treatment based on that diagnosis alone. However, if the liquid biopsy is negative for any markers, you cannot assume that that patient is negative for EGFR or ALK mutations, and a follow-up tissue analysis would be advised.
Currently, the most common usage for a liquid biopsy is in patients who are progressing on a targeted inhibitor. When patients have had a successful course on a front-line EGFR inhibitor or an ALK inhibitor, and they are beginning to recur, a liquid biopsy is commonly used to identify mechanisms of resistance in that setting. This may be more appropriate than a tissue biopsy in the sense that if you have a positive finding from the liquid biopsy—if it’s positive for an EGFR T790M mutation or an ALK mutation that confers resistance—then that obviates the need for performing a tissue biopsy. The advantage is that you’re going to spare the patients the trouble involved with biopsy.
If the liquid biopsy is positive for your marker, then you can go ahead to the appropriate therapy. However, if you do not see it—if it’s negative—then at that point, it’s the best practice to go back and perform a tissue biopsy, which might give you better information. You have to remember that in a liquid biopsy setting in patients who are just starting to progress, it’s very easy to miss the circulating tumor DNA. There’s a lot biology involved here, and it’s very dependent on how much DNA the particular tumor is shedding. If you have a very slow-growing metastatic lesion—which is common for a T790M—you might have very low levels in plasma. It might take 2 or 3 weeks for you to start detecting the presence of a T790M mutation. But, you should make no assumptions about the results of a liquid biopsy if you see no data. No data should be considered inconclusive.