Jonathan W. Riess, MD: In terms of best approaches for molecular testing for patients with metastatic non–small cell lung cancer, there are different methodologies that could be used with either plasma or tissue. My general approach is to look at broad molecular profiling because we have 7 approved targeted therapies for eGFR mutated lung cancer matching the TKI (tyrosine kinase inhibitor) to a great targeted therapy. We have osimertinib for eGFR, among others; alectinib and brigatinib for ALK; crizotinib and entrectinib for ROS1; selpercatinib for RET fusions; larotrectinib and entrectinib for really rare NTRK fusions; capmatinib for MET exon 14 skipping mutations; and dabrafenib and trametinib for BRAF mutated lung cancer.
To capture all of those, I look at broad-based genomic profiling, next-generation sequencing (NGS). You could look at tissue. You could look at plasma. There are pros and cons for each one. If I need something quickly, I will often send plasma since you don't need to set up a biopsy if you need additional tissue, and you get the results in a week. If there's a big surgical specimen sitting in the bank, I'll often use tissue. Where I don't have tissue, I'll first try with plasma as I'm setting up the biopsy.
There are methodologies, such as IHC (immunohistochemistry) and FISH (fluorescence in situ hybridization), that could be used for ALK. For IHC, there is an FDA-approved test, and those can come back quickly. There are certain situations where, if I really suspect there's an ALK fusion and I need to start treatment quickly, I will occasionally send that as part of a targeted panel to get the results quicker. Generally, I send next-generation sequencing for broad genomic profiling.
Lyudmila A. Bazhenova, MD: My personal approach is to use both liquid and tissue next-generation sequencing at the time of the diagnosis. I believe that those tests are complementary. An alternative approach would be to do next-generation sequencing on tissue and save a plasma NGS, or so-called liquid biopsy, for patients in whom tissue was insufficient. I wait until I have all molecular testing before I initiate therapy. In general, most of your patients with metastatic non–small cell lung cancer will also have PD-L1 testing done.
I expect that PD-L1 testing will come first before I have the results of the molecular testing. I always wait for the molecular testing results before I start chemotherapy or any kind of treatment. Occasionally, I have to start therapy before NGS results are back because the patient is highly symptomatic and rapidly declining. In this case, I would recommend not giving immunotherapy with the first cycle. The reason to hold immunotherapy is to avoid emerging adverse effects, such as pneumonitis, if immunotherapy was given before tyrosine kinase inhibitors.
If you think about testing for our patients with metastatic non–small cell lung cancer, you need to remember to test. There are several ways of testing for molecular alterations. In some instances, immunohistochemistry or FISH are appropriate. I use next-generation sequencing, which has both a DNA and RNA capacity. I know that some institutions will have a rapid immunohistochemical screen for ALK, NTRK, and others. It doesn't really matter what system you use as long as you use the one that assures that your patients with metastatic non–small cell lung cancer get tested for all the abnormalities that have FDA-approved drugs.
NTRK, ALK, and ROS1 could be a part of NGS. Alternatively, you can order those genes separately. The most important thing when deciding to test and what company to use is to check your platform, if you're using an NGS platform, and make sure that the company you're using is testing for NTRK. NTRK is a new addition; we weren't testing for NTRK fusions 3, 4 years ago. In my practice, I prefer NGS over single-gene testing for 2 reasons: (1) I believe NGS testing saves tissue, and (2) it's likely to save costs because now we have to test for 7 molecular abnormalities. I think the cost will probably add up if you use single-gene testing.
If you identified a molecular abnormality that has an associated targeted therapy, you should go ahead and treat. For example, if you use the single-gene testing and find that you have an eGFR mutation, that's enough for me to start treatment. Our patients with oncogenic drivers are unlikely to have 2 separate oncogenic drivers in the same patient. Those cases have been described, and at this point, we do not know if they should go for ALK or eGFR first, for example, for patients who have both of the alterations.
It is not necessary to complete the NGS testing if you already know that your patient has an eGFR mutation. NGS testing does give you some additional information. We know that patients who have, for example, a TP53 or RB1 mutation have a worse prognosis than patients who don't have either. They're also more likely to transform into small cell lung cancer. Even though I'm not making any treatment decision based on that knowledge, I do find that it does give me some prognostic information. I might also look at that patient a bit more closely.
Another tip that I have for next-generation sequencing is if you have a patient who is never a smoker, or is a former light smoker, who has a higher incidence of molecular abnormalities… If you did not find it on your DNA-based next-generation sequencing, you could consider RNA-based testing, which would be more sensitive to fusion.
Balazs Halmos, MD: It is great to discuss the proper molecular testing paradigms nowadays for our patients with non–small cell lung cancer, especially when we face situations where the tissue is either somewhat limited or very limited. Fortunately, we have a lot of different ways to facilitate that effort, to enhance the completion of molecular testing, which is critical to put our patients on the best treatment path and achieve the best long-term outcome. We want to make sure that we understand the patient’s tumor the best upfront. I complete biomarker testing.
Now, how do we deal with this situation of having potentially limited tissue? Issue number 1 is to make sure that we have good communication between members of the treatment team. This is where a multidisciplinary tumor board is so important to educate each of our colleagues. For example, with our pulmonary colleagues, we know that every time they do a biopsy, their biopsy might be needed for molecular testing. They do their absolute best to do a number of different passes, build a specimen that’s…enough to allow molecular testing, even from a cytology specimen. It’s the same way for interventional radiologists. We know that they need to take a couple different cores, as long as they’re safe from a particular patient, again, to allow the need for molecular and biomarker testing through the patient’s treatment continuum.
Still we’ll face situations when what’s obtained is limited due to safety concerns. Lymph nodes or nodules might not allow much of a specimen to be obtained. The next issue is having the pathology department ensure there’s appropriate tissue stewardship, that the tissue is prioritized for testing that’s necessary, as opposed to potentially wasting tissue for immunohistochemical stains that might not be absolutely necessary. Both coordination and stewardship are very important.
Lastly, we’ve been greatly assisted through the development of ctDNA (circulating tumor DNA) technology, liquid biopsy technology, where you have a patient with very limited tissue where we cannot perform all the molecular tests. We can utilize this platform, ctDNA testing, to complement the tissue testing we’ve been able to do. It is dependent upon the tumor burden of the patient, the stage of the patient, the yield of ctDNA testing, but in the metastatic lung cancer case, 70% to 80% yield is expected. A large majority of the patients will have informative ctDNA results, and they are strongly reliable enough to guide treatment. For example, if you detect an eGFR mutation or MET exon 14 skipping, we have strong data to say that the patient should benefit from the targeted treatment just as much as if we had detected the same mutation abnormality from tissue-based testing approaches.
Transcript Edited for Clarity