 exon 14 skipping mutations. Obviously, you get more approvals and options. It’s exciting. It’s sometimes difficult to identify 

 exon 14 skipping mutations. Nevertheless, we’ve got 2 new drugs approved. Do you want to walk us through some of the data?

 We have a number of options here. Some are approved, and some aren’t approved but are available, which makes it a little more exciting. Obviously, there are diagnostic challenges with splicing products, or theoretically if you don’t capture the DNA intron splicing site, less sensitivity. Tissue is preferred for an RNA-based assessment. We have 2 oral therapies approved in this space: tepotinib and capmatinib. They both have very interesting data. The VISION trial that led to the approval of tepotinib showed a response rate of almost 50% in follow-up and an additional 20% of disease stabilization, with a duration of response of a little over a year. That’s a reasonable option. They’ve also submitted the data that are now available for intracranial activity that look almost identical between intra- and extracranial space. This is a very good option.

We have capmatinib based on the GEOMETRY mono-1 trial that looks similar. There are some differences in terms of handling. Tepotinib has once-daily dosing, which makes it easier for patients to manage. But overall, you have 2 very good options in this space that are available for patients. The question is, where do they fit in? It isn’t an 

 slam dunk with 80% response rates. There’s also a question of the value of immunotherapy in 

, especially with smoking. That’s a topic of hot debate.

There would be many who wouldn’t automatically extrapolate the data into the first-line space. There may be some tolerance considerations. A head-to-head trial data set against chemoimmunotherapy in the first-line setting is needed to determine the optimal sequence. It’s obviously a rare mutation, but in our area, it isn’t quite as rare. We have a lot of smokers, and we find [it] more commonly. Interestingly, the data here [were] as seen in a broad background of patients, and the geriatric population did as well or better.

 exon 14 skipping mutations is we don’t have any specificity for targeting. There isn’t a mutation or an ATP [adenosine triphosphate] pocket configurational change that could be specifically targeted. We’re suffering the full width of on-target effects, and vascular integrity and edema are at the heart and center. There are many ways of approaching this. I’m not sure that there has been 1 standard recipe, from extremity elevation to physical measures, compression stockings, or diuretics. They have all had limited opportunities.

It’s interesting. When I look at GeoMETry-III, which was at least in concept presented as an option, we’re looking at a trial design that aims more at placing the drug confidently in the second-line setting. Those data are still evolving, and it has to be adjusted to the individual clinical situation where it fits in.

We do have an additional option. We talked earlier about amivantamab[-vmjw] as a bispecific antibody in the 

 exon 20 space. With extracellular domain targeting, you have a great deal of versatility. You’re using both effects of targeted therapy as well as some immunogenic effects that may be helpful. We’ve seen some very small data sets from the CHRYSALIS trial in the 

 exon 14 cohort that showed response rates in the 60%-plus range as well. Those were very small numbers, and the handling here is obviously more complex than for an oral TKI [tyrosine kinase inhibitor], but I do think most of these patients will eventually see 1 oral agent followed by amivantamab or vice versa in sequence.

I’m not sure that there’s going to be a clarification scientifically on this question. I don’t think there’s going to be a lot of exchange unless it’s for toxicity between the 2 oral agents. It comes down to convenience, availability, and access that will make the decision on what’s going to be utilized in either the first- or later-line setting. The approvals are pretty permissive to initiate those patients. But the question is whether they’re superior to chemoimmunotherapy in the first-line setting. They’re probably better tolerated, but we’re not sure if they’re necessarily going to be superior in efficacy. For me, the standard right now is chemoimmunotherapy in the first-line setting and waiting for more data to change that sequence, even though they’re tempting and the data look quite good.

 Correct me if I’m wrong, but for a patient who’s [positive for] 

 exon 14 skipping, you would start with chemotherapy–IO [immuno-oncology] first, and then how do you choose between capmatinib and tepotinib?

 They’re both good agents—there’s no question—and have shown similar levels of efficacy and safety. The data are comparable within the margin of error of cross-trial comparison. The once-daily dosing is probably the main argument. Especially practicing in Florida, having a robust geriatric data set is reassuring, but I’m not sure if I’d expect the data to be much different with capmatinib. It comes down to provider choice, but typically once daily is a meaningful argument for me.

Videos

Dr Martin Dietrich provides an overview of how to treat MET exon 14 skipping mutations in non–small cell lung cancer.

Recent Advances in Treatment of MET Exon 14 Skipping Mutations in NSCLC

 exon 20, which isn’t to be confused with 

 is exon 20, but most are. Misako, do you want to walk us through this? We’ve got some very exciting data coming down the pike for these mutations. Walk us through where we’re at with some of the drugs that are emerging as potential therapeutic strategies.

 Sure. There are 2 important studies in this space. In particular, the DESTINY-Lung01 trial is potentially practice changing. DESTINY-Lung01 was a multicenter, 2-cohort phase 2 trial enrolling patients with 

-mutant or -overexpressed non–small cell lung cancer refractory to standard treatment to receive [fam-]trastuzumab deruxtecan[-nxki], or T-DXd, which is an ADC [antibody-drug conjugate] with 3 components—the humanized anti-HER2 IgG1 mAb [monoclonal antibody], the topoisomerase I inhibitor payload and exatecan derivative, and a tetrapeptide-based cleavable linker.

The primary end point was objective response rate per RECIST version 1.1 by independent central review [ICR]. The initial presentation of the preliminary results [occurred] at ASCO [American Society of Clinical Oncology Annual Meeting] 2020, and the updated primary analysis results from the 

-mutant cohort of the DESTINY-Lung01 study were reported as a late-breaking presentation at ESMO [European Society for Medical Oncology Congress] 2021 and simultaneously published in the 

. The total number of patients reported was 91. In patients who have progressed following 1 or more systemic therapies, the confirmed objective response rate by ICR was 54.9%. Disease control rate was 92.3%, with a median duration of response of 9.3 months, median PFS [progression-free survival] of 8.2 months, and a median OS [overall survival] of 17.8 months.

Any treatment-related adverse events [TRAEs] were seen in 96.7% of patients, with grade 3 or higher TRAEs at 46.2%. Importantly, the adjudicated drug-related ILD [interstitial lung disease] was seen in 26%. Of the 26%, approximately 19.8% had grade 1 or 2, and approximately 6.6% had grade 3 or higher. The efficacy results—with the objective response rate of 54.9%, median PFS of 8.2 months, and median OS of 17.8 months—are certainly promising, but I feel that more data will be helpful to identify patients at risk for developing AEs, in particular ILD.

The other important study to highlight in this space is the ZENITH20 trial using a TKI [tyrosine kinase inhibitor], poziotinib, for patients with 

 exon 20 mutations. ZENITH20 is a global, multicohort phase 2 study. Cohort 4, which enrolled first-line patients with non–small cell lung cancer with 

 exon 20 insertion mutations, was just presented at ESMO Targeted Anticancer Therapies Congress 2022. The primary objective was objective response rate per RECIST by ICR. In a total of 70 patients, there was an objective response rate of 41%, a disease control rate of 73%, a median duration of response of 5.7 months, and a median PFS of 5.6 months. Of note, 48 patients took 16 mg [daily], and 22 patients took 7 mg [twice a day].

Grade 3 or higher adverse events were reported in 69% of patients. The common ones included rash, dermatitis, and diarrhea. While the safety profile was consistent with the TKI class, the tolerability, dose reduction, and interruptions were reported to be improved with the [twice-daily] dosing.

 Misako, you elegantly reviewed both of these trials. How do you treat a patient who has a 

 mutation? I assume you’re testing with comprehensive genomic profiling and identifying these. Trastuzumab deruxtecan is potentially one that could be approved. We didn’t get a chance to talk about T-DM1 [ado-trastuzumab emtansine]. That’s another ADC targeting 

 I try to put these patients on clinical trials whenever available. But if none are available, I’d do platinum-based chemotherapy up front because the current data for the DESTINY-Lung01 study is post–first-line standard-of-care platinum-based therapy. I’d be willing to use trastuzumab deruxtecan in the second line. It’s on the NCCN [National Comprehensive Cancer Network] guidelines with T-DM1, as you just mentioned. I have used afatinib in patients with 

 mutations when the performance status was borderline or when I thought chemotherapy was challenging.

 This is the challenge. I’ve fortunately been able to use trastuzumab deruxtecan on clinical trial. My experiences mirrored what we’ve seen in the trials. The response rate is north of 50%. Some of these are very durable. We haven’t had time to talk about intracranial activity of these potential drugs, these new drugs, these ADCs. We don’t know. The molecule is quite large, so theoretically and scientifically, it shouldn’t elicit responses in the brain. But this is what we’re looking for. I also had the opportunity to use poziotinib. Misako, I don’t know whether you’ve used that drug. Do you have any thoughts on that?

 It’s the exact same situation as yours, Dr Levy. I have clinical trial experience with both trastuzumab deruxtecan and poziotinib, and some of my patients experienced both. My personal impression is that trastuzumab deruxtecan is basically chemotherapy given in a targeted fashion. You need to monitor for cytopenias and your usual chemotherapy adverse effects like nausea, vomiting, and hair loss. Poziotinib, on the other hand, is an oral option. While patients like having oral options, this drug is pretty challenging in the sense that there are very high rates of high-grade rash and diarrhea. It’s very difficult to adjust the dose, hold, and be in close communication with patients. Your nurses have to be aware of all of these issues. It’s doable, but it isn’t that easy.

 Yes, it isn’t the easiest to get people through.

 No. But the [twice daily] dosing is potentially a little better, and that’s promising—although the ZENITH20 cohort 4 presentation was for the first line, and the response rate was only 40%, if I’m remember the data correctly. I understand that the number of patients was still limited. But with this amount of toxicity, we have to balance out the efficacy. It’s a tough field.

 Yes. That was a nice overview. We still have a lot of unanswered questions for this genotype. My hope is that trastuzumab deruxtecan gets approved very soon, and we can leverage it. I’d like to see it used in the first line. It’s a pretty good drug. Obviously, the ILD that you mentioned can be a challenge.

Dr Misako Nagasaka reviews clinical trial data presented at ESMO 2021 on therapies for NSCLC harboring HER2 exon 20 mutations.

Updates in Treatment of HER2 Mutant NSCLC

 exon 20 insertions. It’s a slightly different alteration, difficult to drug, has a narrow therapeutic window, and it’s very difficult for these TKIs [tyrosine kinase inhibitors] to fit into the pocket. But we’ve had some new data and drug approvals in this setting. Fernando, do you want to walk us through the amivantamab [Rybrevant] and mobocertinib [Exkivity] data?

 Yes. It’s a breakthrough. In 2021, we had 2 approvals for these very heterogeneous groups of mutations. It’s very important to know that unlike 

, they’re born the same, are very heterogeneous, and you can have rare mutations with different sensitivities. Of note, 

 exon 20 insertions have a very low response rate to first- and second-generation TKIs. We had only standard of care chemotherapy as of last year, but now we have 2 new approvals, amivantamab and mobocertinib.

 antibody that works by inhibiting the ligand binding and with receptor degradation, ADCC [antibody-dependent cellular cytotoxicity], natural killer cells, and macrophages. It’s a very powerful drug that acts against 

We recently had the report of the data from the 

 exon 20 insertion cohort with patients post–platinum therapy from the CHRYSALIS trial, which is a phase 1 dose escalation expansion trial with 81 patients. Fifty percent of the patients had previous IO [immuno-oncology], 20% had brain metastases, 25% had already used a TKI, and the median number of previous therapies like 2. They showed an overall response rate of 40%, with 3 patients having complete response, median PFS [progression-free survival] at 8.3 months, and a median duration of response of 11 months.

It’s important to note that the activity was across the spectrum of the different 

 exon 20 insertions. No MTD [maximum tolerable dose] was identified. It’s important to note that with the adverse event profile, you can find a little from the 

 world, like rashes, dermatitis, diarrhea, paronychia, and some peripheral edema, but it was quite safe. The incidence of grade 3 or higher adverse events was about 35%, and rash was the most common infusion-related reaction. It was very common. More than 60% of the patients developed infusion-related reactions. It’s usually in the first 2 days of the first cycle.

There are several ways to mitigate these infusion reactions. The first is by splitting the first dose in the first week in 2 days, a very slow rate of infusion, and premedication. Most of the patients didn’t have any subsequent infusion reactions in cycle 2 or 3. It’s something that we need to try to understand in all the settings to know how to best manage this infusion reaction.

Mobocertinib is an oral drug. It’s first in class, very powerful, and selective against 

 exon 20 insertions. It was also approved in 2021. We have reported data from the 3-part phase 1/2 dose escalation, expansion, and extension with more than 110 patients. Overall response rate was 28%, which is a little lower than amivantamab. There was a good disease control rate of 80%. The median duration of response was 17.5 months, which is a little longer than amivantamab. In terms of adverse events, diarrhea is the main issue, with more than 30% to 35% of the patients with grade 3 or higher diarrhea, and 17% of the patients had to discontinue the drug because of that.

We now have 2 good drugs in a space where we were lacking targeted therapies. There are several challenges on how to best sequence it, if it’s durable, and if 1 has activity against the other.

 How do you choose between the 2 in your clinic off of a clinical trial? Because the data is a little different and we don’t like to do cross-trial comparisons, but we do. What’s your go-to for an 

 exon 20 patient? Would you use it first line? What drugs are you using first line and second line?

 It’s important to note that all this data is in the post-platinum setting. The approval is after progression on first-line platinum. Chemotherapy is still the standard first line. I usually don’t add immunotherapy in this setting, even though we have some retrospective data showing that the exon 20 insertions might have a little better response rate. But it’s more or less the paradigm of all the oncogene-driven cancers.

I’d go for chemotherapy only first line. In the second line, we need to discuss with the patients how the toxicity profile is different. One is an oral drug, while the other is an infusion. We still await data from CNS [central nervous system] control, but we don’t expect any of these drugs to have very good CNS control. It’s sharing with patients. I’ll probably go for mobocertinib as the first drug. There’s a lot of push that the amivantamab data show a greater response rate, but mobocertinib showed a good duration of response. As far as sequencing, there’s a good rationale on placing the antibody after the small molecule. If you think about resistance mechanisms, like C797S, where the mobocertinib binds, amivantamab might have some activity after that. There are still a lot of unanswered questions.

 That’s a very nice overview. The thing you hit on that we’re all struggling with in regard to all these genotypes is, what’s the optimal sequencing? How do we move forward with combination strategies? How do we sequence in the way that can give the best quality of life and improve the outcome for patients? It’s getting incredibly complicated for each of these genotypes as the field is getting more crowded.

Dr Fernando Santini explains EGFR exon 20 insertion mutations in NSCLC and the best approaches for treating these cases.

Recent Advances in Treatment of EGFR Exon 20 Insertion Mutations in NSCLC

 Let’s move on to our next section, which is on the specific gene mutations in lung cancer. We’ll start out with 

 mutations. We’ve come a long way. We have first- and second-generation TKIs [tyrosine kinase inhibitors] like erlotinib [Tarceva], gefitinib [Iressa], and afatinib [Gilotrif]. They were compared with chemotherapy early on and showed improvements in response rates and in PFS [progression-free survival] specifically with patients with sensitizing mutations, and also improvement in the quality of life.

More recently, we have a head-to-head trial comparing a third-generation TKI, osimertinib [Tagrisso], to erlotinib and gefitinib in the FLAURA trial. That showed improvement in PFS and in OS [overall survival]. Based on its tolerability, OS, intercranial response, and proclivity to penetrate the blood-brain barrier, osimertinib is now considered the standard of care for at least sensitizing mutations in exon 19 and 21. It seems pretty simple, but it isn’t. I tell fellows that osimertinib is just the beginning of the story, that there are going to be combinations, and that we’ve got to understand how to tackle resistance.

But first, let’s talk about these first-line combination trials. Hatim, walk us through some of the ones that you think are important to look at and where we’re heading in the field with combination strategies for sensitizing 

 It’s a great question. Thank you, Ben. The first-line management of 

 TKI. That has been the historical standard. Some of the ways that are being explored in clinical trials are combinations of TKIs plus chemotherapy, combinations of TKIs plus anti-VEGF–directed therapy, and even combinations of 

 in the frontline as well. The frontline management is getting somewhat complex, and combinations may have some importance. The critical question is whether there are a population of patients who are going to be best served with a combination vs monotherapy with a TKI.

At ESMO [European Society for Medical Oncology Congress] 2021, there were a few abstracts presented around anti-VEGF combinations with erlotinib therapy, including the final results of the BEVERLY study, which looked at frontline erlotinib and bevacizumab [Avastin] in an Italian population and showed improvement in overall survival with the addition of bevacizumab in that study.

There was another trial presented that looked at how 

 could segregate patients from erlotinib and ramucirumab [Cyramza] frontline. This was an important study to emphasize the importance of co-alterations with 

, or other co-alterations, that may segregate 1 combination, perhaps 

 plus VEGF. The jury is still out in that regard. The last abstract that also came out at ESMO 2021 looked at osimertinib plus bevacizumab. It didn’t show an improvement in PFS. We’re eagerly awaiting the chemotherapy plus TKI data from FLAURA2, osimertinib plus chemotherapy, and also a little more clarity with TKI plus VEGF.

 The VEGF story is a little confusing to me. We’ve had some data that looked like improvement in PFS, but we haven’t seen that in combination with a third-generation TKI. The chemotherapy plus osimertinib is intriguing. We’ve got a Japanese study and an Indian study showing that chemotherapy plus first-generation TKI is better than first-generation TKI alone. We’ll have to see how this this plays out. If FLAURA2 is positive, I’d like to know who benefited, because it would be nice to not have to give chemotherapy to everybody. We’ll have to see how this shakes out over time.

I’m going to deviate a little here. Patients who are on osimertinib who develop disease progression are a challenge. This is a real challenge as to what to do. Martin, how do you handle patients who are on osimertinib who develop disease progression? What do you do with these patients?

 The proper answer is that we don’t know. There’s a big difficulty in guiding this, and everything that’s post–

 progression is basically off label, with the exception of standard chemotherapy. The first step is to understand why it’s progressing, so I get at least a liquid biopsy, preferably a repeat biopsy. We’ve seen fusions and amplifications as resistance mechanisms. We’ll see more of this as we get into more precise testing. We have a bit of a diagnostic disadvantage by not being as aggressive about tissue biopsies, amplifications, and fusions, which are very difficult to see on liquid biopsy, depending on the circumstance.

We try to see if there’s a straight progression with 

 overactivity that maybe overrides the activity of osimertinib. Looking at clinical trials, looking at some of the new allosteric ErbB inhibitors that we have in clinical trials. And certainly, in combinations of TKIs, we’ve seen cMET as a frequent escape mechanism overriding 

 in combination therapies, both in trial as well as off label. That’s a consideration. If none of these are options, my standard would be the IMpower150 regimen. We’ve seen sustained benefit with the addition of immunotherapy, interestingly so and not quite well understood, with a differential effect between the 

 seems to benefit. There was an update as well that showed that it seemed to maintain the benefit overall.

IMpower150 would be my standard answer. They were included and were also approved in Europe for the use for patients with 

. The FDA [Food and Drug Administration] excluded them, even though they were permitted in the clinical trial, which I thought was an interesting decision. That would be my standard approach. First, testing, then evaluating clinical trial opportunities, because that honeymoon period of osimertinib, which is both very effective and well tolerated, should be followed up by genomic guidance in the second-line setting. For the all-comers, chemoimmunotherapy plus an antiangiogenic agent is the current and most available standard.

 That’s a great summary. We’re still in a vacuum as to what to do for these patients. I agree that tissue and liquid makes a lot of sense here, understanding mechanisms of resistance, wedding those mechanisms of resistance to particular therapies.

A discussion on EGFR mutations in non-small cell lung cancer and the combination treatment strategies for sensitizing mutations.

Updates in Treatment of NSCLC with Classical EGFR Mutations

 Let’s talk tissue. Obviously, procurement and tissue stewardship is a big issue. It’s travel time from the needle to the slide to the pathologist to the NGS [next-generation sequencing]. There are a lot of challenges and hurdles that can happen along the way. Hatim, is tissue sufficiency a challenge at your institution [the University of California San Diego]?

 Tissue insufficiency has been less of a problem, partly because of education among our multidisciplinary team, specifically involving interventional pulmonary and interventional radiology, who have also understood the importance of genomic profiling. When you have the whole health care team concerted around a focus, it helps facilitate some of these barriers. It requires working [together]. Dr [Timothy Craig] Allen elegantly and appropriately made a comment about communication among the team and how it can drive optimization. In our place, we’ve been trying to do that by including the whole multidisciplinary team, including interventional pulmonary and interventional radiology, in some of the education about the molecular drivers.

 Yes, communication is key, and even having a physician champion who can make sure that everyone is on the same page at the institutions is critical to getting this done. It can be a challenge. Fernando, rounding us out, we’ve talked about some of these strategies, but what are your thoughts on strategies to help improve testing rates for molecular biomarkers in lung cancer, both in the academic and community settings?

 One very important point is to make sure it’s tissue negative for drivers and not just insufficient tissue. We can’t take for granted that the tissue is negative. We must make sure we have enough tumor content and enough cells were analyzed. Read more lines from the molecular report. It’s important in making sure you’re giving the best care. Multidisciplinary awareness of the importance of the molecular analysis is key.

 Yes, it takes a village to get testing done. Everyone said that it takes a concerted effort to make sure this is done efficiently and effectively, with results that are timely. Liquid biopsy plays a huge role. It doesn’t solve all the issues, but it clearly can help with rapid turnaround time, especially with patients who shed. You can identify an alteration and act on it without having to wait for the tissue, for the most part. There are clearly challenges with liquids, and we don’t have time to go into all the things that liquid biopsies can discover in addition to circulating tumor DNA, but it certainly has revolutionized the diagnostic algorithm.

A look at the challenges associated with different molecular testing methods and the barriers that patients with non–small cell lung cancer face in getting tested.

Challenges Associated with Molecular Testing in NSCLC

 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.

 There are various ways of testing. At the beginning of our journey with 

, 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 

 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.

 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?

 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.

 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?

 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 

 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.

 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?

 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.

 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.

 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, 

 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.

 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?

 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 

 setting postprogression. If there’s no known resistance mechanism, we try to do Archer [testing] for these patients as well.

 Misako, are there any other experiences at your institution [the University of California, Irvine, School of Medicine]?

 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.

 Martin, what are your thoughts on how you guys do it?

 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 

 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.

 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?

 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.

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.

Molecular Testing in NSCLC: What is the Optimal Sample and Time of Testing?

 [Turning to] the structure, I know this isn’t what the guidelines recommend. They still de-emphasize liquid testing—more in a second—and reflex testing with some exclusions. In a real-world setting, there are so many obstacles to obtaining appropriate tissue at the time of first visit that it’s just a simple clinical requirement for us to have the information as soon as possible. Liquid biopsy has filled in very nicely and has become a cornerstone of diagnostic [testing] up front.

 Yes, I agree. We’ll get Tim’s perspective on tissue, but I want to touch on liquid and get some perspectives around the room. Tim, what are your thoughts on how you incorporate ctDNA [circulating tumor DNA] in the diagnostic algorithm for patients? Then I want to get Fernando’s view on what’s done at Memorial [Sloan Kettering Cancer Center].

 Liquid biopsy has come into its own from its origin a few years ago. It’s still correct to say that tissue is superior. But we have to yield to the practicalities of patients who are difficult to get tissue from, those who recur. Often, it’s very difficult to get tissue, and those are good candidates for liquid biopsy. Some folks are doing liquid biopsy at the time of tissue biopsy and work-up because they’ll get a faster turnaround on the liquid. The tissue comes later with the results from the tissue molecular diagnostics that can help supplement the liquid biopsy results.

 Hatim, what’s your perspective on how to incorporate liquid biopsies at UCSD [University of California San Diego] and how you’re using it in the diagnostic algorithm?

Thank you, Ben. Tim gave a wonderful answer. To add to it, I’ll say that we do liquid biopsy up front. We think liquid biopsy is an essential component of diagnostic testing, largely because of the turnaround time that can be achieved. With more sophisticated technologies, there’s good concordance from tissue to liquid. When there isn’t concordance, if we get a negative liquid test, we emphasize reflexing the tissue and not forgetting about the tissue as the gold standard.

 I agree. A negative tells you nothing. The sensitivity of liquid is around 70%. You get a 30% false negative, and a positive essentially rolls in. But at Johns Hopkins we’ve bought into the algorithm and drunk the Kool-Aid of using liquid with tissue. Given the challenges of tissue, we can talk about it later in this section.

Fernando, how are you routinely using liquid at Memorial? Is it every patient? I know you have an in-house liquid test. You have MSK-IMPACT, the tissue test. How do you molecularly profile patients?

 We have 2 broad and very good panels: MSK-IMPACT, which is the tissue NGS [next-generation sequencing], and a smaller but still very broad panel, MSK-ACCESS, for NGS and ctDNA. Fortunately, it reflects every lung cancer histology. [For tissue testing], non–small cell lung cancer goes to MSK-IMPACT reflexively. At every new visit, we try to collect with MSK-ACCESS the ctDNA as well. We’re trying to do as well like the Idylla for ctDNA for 

, which can sometimes be even 1 day or a few hours in turnaround time. With a good amount of patients, we can start treatment fast with the Idylla ctDNA, where it’s complementary. We’re trying to do both platforms in the first visit.

 All of us have started to do this. I want to reserve the challenges of tissue for the end, not only in the academic setting but in the community setting as well. Maybe we can revisit liquid in the context of those challenges.

Experts in lung cancer explain their views on the critical role of liquid biopsy in molecular testing in non–small cell lung cancer.

Liquid Biopsies in the NSCLC Molecular Testing Algorithm

® program titled “Actionable Mutations In Non–Small Cell Lung Cancer: Expert Perspectives in Testing and Targeted Therapy.” I’m Dr Benjamin Levy, and I’m an associate professor at Johns Hopkins School of Medicine [in Baltimore, Maryland] and the clinical director for the Johns Hopkins Sidney Kimmel [Comprehensive] Cancer Center at Sibley Memorial Hospital in Washington, DC. I’m joined by a panel of experts in the field of lung cancer. We have a lot of complexity to distill. It’s gotten complicated. This complicated field has generated a tremendous amount of enthusiasm, but it’s always nice to hit the pause button and review the data that we have so far with genotype-directed therapies, which have been on the leading of edge lung cancer and solid tumor oncology. I’d like to welcome my esteemed fellow panelists to introduce themselves.

 Hi, I’m Misako Nagasaka. I’m an associate clinical professor at the University of California, Irvine, School of Medicine, and I specialize in thoracic malignancies. I’m excited to join the discussion. Thank you for having me.

 Thank you. I’m Tim Allen, a professor and the chair of pathology at the University of Mississippi Medical Center [in Jackson, Mississippi]. I’m a pulmonary pathologist. This is going to be a great discussion.

 Good evening, everybody. I’m Martin Dietrich from Florida Cancer Specialists & Research Institute and the University of Central Florida [Lake Nona] Cancer Center [in Orlando, Florida]. I’m an assistant professor, and I specialize in genetics and thoracic oncology.

Hello, everyone. It’s a pleasure to be here. My name is Hatim Husain. I’m a medical oncologist at the University of California San Diego. I’m an associate professor in the School of Medicine. I focus on lung cancer, genomically oriented therapies, and immunotherapy.

 Hi everyone, I’m Fernando Santini. I’m a thoracic oncologist at Memorial Sloan Kettering Cancer Center in New York [New York]. It’s a pleasure to be here.

 Welcome, everyone. Thank you for joining me. We’re going to discuss a number of updates in advanced non–small cell lung cancer. We’re going to discuss the data in the context of national and international guidelines, and the evolving treatment landscape and its impact on clinical practice. We can all agree that lung cancer has been the poster child for precision medicine. We’re going to review a lot of those data as they relate to not only medical oncology decision but also tissue interrogation. From the pathology side, we’ll get input on how these are best identified.

Let’s get started with our first topic. Chapter 1 of this personalized medicine story in lung cancer is probably 

 and all these others, let’s take a step back and talk about how we test for these and what we’re looking for. Clearly, 

 are chapters 1 and 2. But there are other alterations that have been discovered that are targetable that we can wed to genotype-directed therapies. It has changed the outcomes for our patients. It has improved outcomes in terms of not only PFS [progression-free survival] but also OS [overall survival], and perhaps underrepresented is quality of life. We’ve come a long way from 

. We have 9 and are maybe going to have 10 actionable mutations in lung cancer. In the next maybe 6 months, we’ll get another approval.

Misako, I’ll start with you. What are you testing for? What do you routinely test for at your center? How do you go about doing that?

 This is a very important topic. Guidelines recommend routine testing at the time of advanced metastatic diagnosis of all patients with non–small cell lung cancer with adenocarcinoma histology for 

, and that’s what I test for as well. It’s important to note that the NCCN [National Comprehensive Cancer Network] panel strongly advises broader molecular profiling, with the goal of identifying rare driver mutations for which effective drugs may already be available, or to appropriately counsel patients regarding the availability of clinical trials. I agree that broad molecular profiling is the key component of improvement of care of patients with non–small cell lung cancer.

 At your centers, I assume that this isn’t a single gene test but that you’re doing comprehensive genomic profiling. Is that correct?

 Shifting to Martin, there are a lot of questions about how this is done. We want to do comprehensive genomic profiling, but is this done reflexively at your institution? Do you have to get the medical oncologist to order it? How do you carry this out?

 We have 2 major hospital systems we’re working out of. One has established a reflex protocol on an Oncomine panel for 50 genes, which is a very efficient way of obtaining genomic results. We’ve been talking about advanced lung cancers. As you know, genomic testing and guidance has found its way into early stages as well, starting in stage I in the setting of adjuvant 

 and immunotherapy targeting. This is affecting early stage disease as well. For the most part, this remains an on-demand testing that we do on an NGS [next-generation sequencing] platform with both DNA and RNA chips. We’re trying to be as comprehensive as possible, but it typically defaults to our medical oncologists to obtain these tests upon seeing the patients in the first visit.

There are many reasons, which we’ll get to later, regarding why we don’t have reflexive testing through the many channels that we’re seeing those patients. It brings us to the question of how best to test. How we’re doing it is typically a question of what’s pragmatically done and scientifically pure. We all agree that tissue-based testing for the time being remains the gold standard when it comes to the sensitivity of testing. But for pragmatic reasons, including tissue availability and turnaround time, we’ve adopted a pattern where we integrate liquid biopsy in a virtually synchronous fashion of ordering at the time of first visit with patients to ensure that patients have at least the best attempt of having NGS results available at the time of initiation of therapy.

Oncologists provide insight into what actionable mutations they look for when they order molecular testing for patients with advanced non-small cell lung cancer.

Timothy Craig Allen, MD, JD, FCAP

Articles