A panel of thought leaders describe the optimal sample to use in molecular testing in NSCLC and the appropriate time to perform testing in clinical practice.
Ben Levy, MD: Tim, as the lone pathologist here, walk us through the testing methods you use to identify key molecular alterations and how pathologists integrate into the work-up for patients with lung cancer.
Timothy Craig Allen, MD, JD: There are various ways of testing. At the beginning of our journey with EGFR, and then without, there was good reason to do single-test kits. Sanger sequencing exists, which was the original. It’s rarely done anymore. Today polymerase chain reaction [PCR] tests, the kits, immunohistochemistry, and FISH [fluorescence in situ hybridization] play a role. These withstood the test of time. In some studies, they’re the gold standard. But in fact, they’re limited. There’s a higher false negative rate and lower throughput with time issues. Limited mutation panels are becoming more of a problem. Of course, there’s a need for subjective eyes for judgment of the results.
Then there’s next-generation sequencing [NGS], which is coming more into its own as costs have gone down. Because of the recent approval of targeted molecular therapies, there’s more of a need for a panel-based testing platform like NGS to be incorporated. It gives us broader options and a deeper depth of coverage. It can detect low and uncommon mutations that you won’t be able to find otherwise. It can cover hot spot regions. With the use of separate DNA-based testing for gene rearrangements, and then also NGS for ALK and other tests, we’re seeing less of a problem with diagnostic difficulty. As you said, we’ll discuss the tissue issue further, but tissue is the issue and it uses less tissue than a series of kits. For that reason, we’ll see more NGS being done around the country.
Ben Levy, MD: At your institution [University of Mississippi Medical Center], are these send-out tests? Are you doing them in house? Based on that, what’s the turnaround time for getting these results?
Timothy Craig Allen, MD, JD: We built a molecular laboratory, which should be going into solid tumor testing within the next 3 months. We’re very excited about that. The goal is to keep the finances intact, because sending it out costs money vs doing it in house. Also, and maybe most important, we can control the turnaround time ourselves. That being said, there are good options with send-outs. Reasonable turnaround times are possible.
Ben Levy, MD: I have 1 more question, and then we’ll get to a few more. But on this point, when you get that tissue, how do you prioritize between testing for PD-L1 vs doing next-generation sequencing? How does this work out? Is tissue exhaustion a problem that you see?
Timothy Craig Allen, MD, JD: Tissue exhaustion is always an issue that we worry about. We try to minimize the problems by having good communication among us. For example, we’re not going to do NGS if it’s a squamous cell carcinoma, for which we just want to do maybe EGFR and PD-L1. But if it’s a non–small cell or adenocarcinoma, then we’re going to do NGS and do a full protocol.
Ben Levy, MD: I want to go around the room. I’ll start with you again, Tim. Do you test all adenocarcinoma, advanced-stage and early stage [patients]? Do you see a lot of testing done where it’s a patient who’s on a particular therapy and it’s a rebiopsy at the time of progression? How often does this come through your laboratory?
Timothy Craig Allen, MD, JD: It certainly does. We test upon progression, because you don’t know what you’re going to see, and it will affect potential therapies in the future. Unfortunately, we don’t do reflex testing for various reasons, mostly financial. But we have a tight tumor board team that keeps things moving quickly. We’ll discuss exactly what to do with the patient in terms of testing. If necessary, we won’t do the NGS platform, but we’ll use a kit or something that will give us something faster. We discuss liquid biopsy. It’s not done on every patient—we still focus on tissue—but we’re seeing it more.
Ben Levy, MD: Hatim, I want to get your thoughts. We got data from the FLAURA study. A lot of us may extrapolate to other genotype-directed therapies in the adjuvant setting. What’s your take on how you’re testing your patients at UCSD [University of California San Diego]? Then we’ll get everyone’s perspective.
Hatim Husain, MD: We’re trying to do NGS as often as possible for all patients with non–small cell lung cancer. That includes nonsquamous as well as squamous. We try to do NGS in early stage. Sometimes we have issues acquiring the approvals for NGS, because as we know, EGFR testing is approved in the early stage. But if we have a high suspicion and still a negative result, we’ll also consider RNA-based testing for fusions. These are the general principles we use. With next-generation sequencing for all non–small cell lung cancer, you’re trying to do this in the early stage as well. Upon clinical suspicion, even being a little more aggressive about either repeat testing or trying to do an RNA-based test to make sure we’re not missing a fusion event.
Ben Levy, MD: Fernando, at Memorial [Sloan Kettering Cancer Center], I’m sure you do a lot of testing. Is there a particular algorithm that you’re wedded to? Is it similar to what Hatim said? Are you doing all patients early stage? How common are you rebiopsying patients at the time of progression, even if they’re not a candidate for a clinical trial?
Fernando C. Santini, MD: I echo that we’re trying to do it in every stage: the early and advanced-stage settings. We try to do a lot of rebiopsies. Whenever needed and whenever feasible, we try to do rebiopsies. If it’s negative for any known actionable alteration, we have the reflex testing for diagnostics for the first line, and sometimes even EGFR in the EGFR setting postprogression. If there’s no known resistance mechanism, we try to do Archer [testing] for these patients as well.
Ben Levy, MD: Misako, are there any other experiences at your institution [the University of California, Irvine, School of Medicine]?
Misako Nagasaka MD, PhD: No, I echo most of what has already been mentioned. I send both tissue and liquid biopsy at the time of advanced metastatic disease for all patients with non–small cell lung cancer. At the time of progression, I also try to repeat a biopsy, but sometimes it’s difficult. They might not be feeling well or the disease might not be under control. But at the very least, I’ll try to do liquid biopsies at the time of progression.
Ben Levy, MD: Martin, what are your thoughts on how you guys do it?
Martin Dietrich, MD: We simplified it by saying we’re going to be sequencing all early stage tumors. This changed a little most recently with the approval of neoadjuvant chemoimmunotherapy. We basically have to go back to needle biopsy specimens even in earlier stages.
But the idea is to unify the testing pattern, not having the pathologist be in the predicament of having to do a stage-adjusted testing panel. I don’t think anything is lost. We have a large clinical trial drug development unit that’s able to channel many patients into different clinical trials in addition to what we know based on standard-of-care markers. Early stage, typically tissue based. The sensitivity for early stage liquid biopsy obviously isn’t as good, depending on tumor volume and shedding status. But that certainly has been a discussion for us. Do we use liquid biopsy if neoadjuvant chemoimmunotherapy is desired where EGFR and ALK are excluded? This is still going to be an evolving pattern, but squamous and nonsquamous are going to be tested by NGS panels independent of stage at this point.
Ben Levy, MD: Yes, that’s where we’re heading here [Johns Hopkins University School of Medicine]. [There are] clearly more identifiable potential alterations in squamous cell that we can wed to particular therapies. Before we segue into the challenges of testing, Tim, I want to get your thoughts on what you’re routinely seeing in terms of samples. Are you getting mostly FNAs [fine-needle aspirations] at your institution? We’ve heard different ways to do this, as far as core biopsy or surgical samples. What do you typically see from a pathology perspective at your institution in terms of tissue, how it’s coming to you?
Timothy Craig Allen, MD, JD: We get about 80% cytology specimens and 20% core biopsies. We’re working with our interventionalist to increase the number of core biopsies, because that gives superior results. That said, cytology clearly works and will work as long as the cell block and the spin down material that can be used for testing is appropriate. If it isn’t, we can take glass slides of smears and use those for testing as well. It takes a little time to select which ones, but it’s certainly worth it. We’ve had little problem with the inability to get enough tissue as long as we’re not dealing with things like a very small amount of tumor or lots of necrosis.
Transcript edited for clarity.