Biomarker assessment for patients with metastatic non-small cell lung cancer (NSCLC)

DR LEVY: The critical aspect of all of this is genotyping the patient. And I think that this has evolved over time. And for these patients where you have enough material from a biopsy, I think, clearly, you want to send tissue off for comprehensive genomic profiling that includes all relevant biomarkers, but you want to do a next-generation sequencing panel.

I think the big question here is the utility of what I call layering in the liquid or doing a plasma biopsy at the same time. And I think my current practice is to do both. I will tell you there are times I don’t do both, and this is patients that I’ve already received the result from tissue, or we get increasingly more pushback from insurance companies about doing both at the same time.

So as a general rule of adenocarcinoma, I will do both tissue and plasma on most of the patients, based on some recent data that when you layer in the liquid you increase your chances of identifying an actual mutation. For squams, I generally don’t though. But there are scenarios where I may not do both and that is when I’ve assured enough tissue on a biopsy and we can wait, and insurance is pushing back. Those would be some of the scenarios where I may not do both.

DR LOVE: So Leora, I’ve been kind of surprised to see how much testing gets down when you present — when you say it's a squamous case. I mean is it really worth doing if it's clear cut scam —squamous? Is it worth doing NGS, for example, Leora?

DR HORN: So we actually don’t necessarily do NGS on all of our patients with squamous. We’ll do it on the younger patients, on the never smokers. Now that capmatinib is approved for patients with MET exon 14, that has been identified in patients with squamous lung cancer, so that may be a reason to consider doing the NGS in patients with squamous, though the incidence is pretty low But we don’t routinely do NGS in patients with squamous histology.

DR LOVE: Melissa, Any comments on the use of liquid assays in the up-front setting?

I guess you get the results quicker, which is beneficial. But in what situations are you doing both?

DR JOHNSON: I guess when I answered this question I was thinking about the fact that each hospital in our, in Sarah Cannon’s network, does the Reflex testing for the FDA-approved oncogene — FDA-approved therapies based on the oncogenes that will result in their therapy a little differently. And so, we’ve struggled across Sarah Cannon’s network to find a way to make sure that we’re getting adequate testing. And so often, when patients come to us initially, that testing is cooking, whatever it is: hotspot panels, NGS panels — and the liquid testing is a way to make sure that nothing got missed quickly, as you’re getting ready to start front-line therapy, Neil.

So I agree with all of my other colleagues that there is a benefit of tissue-based NGS testing and liquid and tissue are not analogous. But I like the rapid turnaround time of the liquid test. Basically, it's a second check to make sure none of the oncogene drivers have been missed. And it saves tissue for trial and that's the other thing I’m always trying to do.

DR LOVE: So John, in other cancers we have, for example, prostate cancer, you have PSA as a tumor marker. There’s a bunch of tumor markers out there. I’m starting to see — I’ not sure I understand the technology, but assays looking at genomics as a way to get a handle on tumor load, so to speak — kind of like a tumor marker. Any comments on that kind of strategy and whether there is any serologic way you can actually follow? I’ve seen, for example, allele frequency if you have an abnormality.

John, anything there?

DR HEYMACH: Sure. So, with the emergence now of liquid-based, next-generation sequencing, this does give the opportunity you might be able to follow blood over time. I don’t think it's quite ready right now to use routinely as a marker of tumor bulk the way you would use a CA125 for ovarian cancer or a PSA for prostate cancer, and this is for a couple of reasons.

One is that tumors have different degrees of shedding. So, for example, we’ve done a recent analysis, we’ve seen EGFR mutant tumors tend to shed more. This looks like it's mainly because they have more copy number. But there are some tumors either that don’t have detectable mutations, or they just don’t shed as much. So, you may be misled if you don’t see anything detectable in the ctDNA to thinking that you had a great response.

The other thing is that — so the measure that we use, we call this the VAF, the variant allele frequency — so that's just measuring out of all given allele like EGFR, what percentage of it is the mutant EGFR, so presumably from the tumor, as compared to everything else in your blood.

Now, we know if you’ve got a good targeted therapy like an EGFR inhibitor, that goes down when you respond. But we also know tumors can transform, they can lose different genes. So, we’re not ready right now to use VAF as a marker routinely, but I think in specific oncogene-driven subgroups like EGFR or ALK, it may be a more useful marker.

One other place where it may turn out to be useful is in the setting of adjuvant therapy. There’s a couple of studies, including 1 from our site, that if you resect a tumor and there’s detectable ctDNA in the plasma after resection, that's associated with really high risk. So those may be patients that merit different treatment. But you still don’t detect it everybody, and in early stage you’re only detecting ctDNA in maybe half the cases.

So I think VAF in plasma DNA has potential. Thera are certain things it may be used for in the future, but it's not quite ready yet to be used a general marker for tumor load.

DR LOVE: So general medical oncologists are used to hearing about MRD, particularly in hematologic cancers; myeloma you hear about it, CLL, etc. We’ll see if that ever — that kind of concept ever comes into solid tumors.

Selection of first-line treatment for patients with metastatic NSCLC without actionable tumor mutations; role of immune checkpoint inhibitors

DR LOVE: Let’s talk about first-line therapy, again in patients without targetable mutations. We’ll talk later about what that list is, as it's growing.

This has really been a controversy for a few years now, the issue of how do you factor in PD-L1 level, Ben?

And we’ve seen over the years, in general is kind of reflected here, that, in general, people want to use combinations under 50%, in general, they want to use monotherapy unless the patient is symptomatic or needs a quick response. But we are seeing people —
Govindan, as well as Melissa, who kind of think more about the combination, even in higher — patients with higher PD-L1s. How do you factor — and, of course, obviously, Ben, you have to consider the patient and how well they might tolerate chemotherapy, but right now how are you factoring in the decision?

Does it make a difference if it's 60% versus 90%, Ben?

DR LEVY: Yeah. Well, to answer that first question, I think we do have recent data that the — this is the PD-l1 as a continuous variable, and the higher the PD-L1, even above 50, the more likely you are to respond. So, patients with 90 may respond better to checkpoint inhibitors than patients with 60. And this is some interesting that's emerged.

I think this treatment decision has be individualized. And I think we all do this. And it's tough to dichotomize this, but I do look at patients’ performance status. I do look at patients’ — I try to steer away from age as a factor. I just try to, instead, use performance status. I look at disease burden, I look at their symptoms to make a decision, specifically for those patients over 50%, whether they should be getting single-agent pembro or triplet therapy. I mean I think we do cross-trail comparisons from KEYNOTE 189 and 024, showing that the hazard ratio may be a little bit more robust with the triplet therapy versus single-agent immunotherapy. Again, these are cross-trial comparisons. But, overall, my decision for patients greater than 50%, I generally do default to single-agent pembro.

Patients in which I would consider a triplet therapy are those with high disease burden. Patients, where I’m not sure they can get to second-line therapy, meaning they’re sicker and they need to be, what I call, systemically debulked. And patients who may be a little more fit, I would consider using triplet therapy.

DR LOVE: So Melissa, we saw a very similar reaction conceptually when we said it was a squamous cell, just a little bit different in terms of the chemo, obviously more taxanes basically, either of the 2.

Any comments? Again, it seems like maybe you kind of think more about the combination. Can you talk a little bit about how you approach patients in this way?

DR JOHNSON: Yeah, Neil, I agree with Ben’s comments, that it's not a dichotomy and every patient is an individual.

But I’ve been burned more than a few times with a PD-L1 above 50% read from the pathologist, that is a semi-objective call on their part, and the patient’s response to single-agent pembro. And in retrospect, sometimes we understand that it's because of STK11, for example, or some other resistance mutation that the patient harbors in their tumor. I get nervous when the patient has had a need for radiation prior to starting systemic therapy and has had some amount of steroids in their body before they’ve started treatment, brain radiation in particular. And I worry that pembro alone won’t be good — won’t be effective for a while. And so, the chemotherapy, sometimes I rationalize we’ll begin to cytoreduce the cancer while the T-cells are revving up.

I think, no question, patients are interested in a chemotherapy-sparing regimen. And so, I consider it. And now that nivo/ipi potentially is another option, we certainly should consider chemotherapy-sparing regimens. But in general, I find patients want the best therapy and, in my mind, the chemo-containing regimen is preferable.

DR LOVE: So maybe the most common question that we got, Leora, from these docs, at least related to immunotherapy, is actually kind of great question to even think about, which is what happens in the patients who are doing really well in terms of potentially discontinuing therapy? And it looks like our faculty, depending on the situation, the response, for example, maybe is a little bit split about whether or not to keep in going indefinitely or to stop at any particular time.

It also, Leora, brings up the other question that I don’t know that there’s enough visibility about, which is that it seems like we have a fair number of people who are actually making it to 5 years now. And the question of, how are we really altering the biology and the long-term outcome of this disease?

What’s your approach, Leora? It looks like you’re open to stopping therapy.

DR HORN: Yeah, and I wonder if people would change their answers if you asked them, now with COVID. But the reality is these patients have metastatic disease. And part of the reason that I even said in the PD-L1-high I’d give pembro alone, it's hard to continue doublet therapy with pemetrexed and pembrolizumab for 2 years. In fact, I have very few patients. I always end up dropping 1 of the drugs.

We have a whole cohort of patients who are now long-term survivors. We were lucky, we were part of all Phase I with nivolumab, pembrolizumab and atezolizumab. And I have patients now who’ve been off therapy for 7 or 8 years and I’ve never had to retreat them. And I am in the camp that if you give chemo — if you gave IO alone, I’m not really sure that chemo/IO would have been any better for them. I’m still waiting to see that 5-year data and see if we still only have 15 to 20% of patients who are alive at 5 years. And have we really shifted the curves that much with the chemo/IO combos? And who wants to come to the cancer center every 3 weeks for 5 years or 10 years if you can stop them at 2 years and they still have good quality of life.

And so, I start preparing my patients at about 18 months and say, hey, you’ve only got 6 months left. As opposed to suddenly saying, oh, your tier is up. You’re done. So, mentally, they’re ready. I’ve had 1 or 2 patients who want to continue, and 6 months later, when their scans look fine, they’re like, oh, okay, fine. I’ll stop now.

But these drugs are expensive. My brain is still very much Canadian. And I think 2 years is enough.

DR LOVE: So Melissa, Leora referred to q3wk, but we just saw the FDA-approval for q6wk. Are you — for pembro. Is that what you’re doing, period?

DR JOHNSON: I think we will. We’ve been waiting for that data. And it begs, I think, a good question, Neil, that we don’t have much data support to suggest how often we really need to be dosing. The indication is for a higher dose once every-6 weeks. But what if you gave the same dose every 6 weeks? That trial has not been done; probably won’t be. But I’ve often wondered, how much is really enough?

I have some patients actually on q2wk avelumab, from the Phase — from that Phase I trial. During COVID, they have not been willing to stop because they’re not going to be able to restart that drug, right, and they can restart others. We’ve had this discussion several times. And if it ain’t broke, don’t fix it is, I think, their philosophy.

We did, for the first time, begin to delay their therapy during COVID. And so, I think we’ll all have that experience, to be able to see what that really does. And I think that maybe will help us point patients that, I guess I have a lot of them, that have been unwilling to stop at 2 years, to maybe evidence that it's okay.

DR LOVE: Ben, putting aside patient preference, when I saw that indication I was thinking, everybody is just going to do it. Is that the case, or not really?

DR LEVY: I agree with Melissa. We’re not there yet. And to be quite honest, we’re governed, oftentimes, about when is entered into our EMR as an option. Of course, you can change it, the dose and the schedule, in the EMR as an adjustment. But we’re not there yet, but I would agree, that COVID and other issues, may allow us to realign our practice standards, and we may start doing this. We’re not currently do it, the 400 q6wk, but something that we are starting to think about doing.

DR LOVE: Interesting.

So, John, Leora was referring to this issue of maintenance and the pemetrexed and pembrolizumab. And, again, we got a bunch of questions about that. But 1 thing that, actually Ed Garon brought up to me, when I did the session with him, that I didn’t think about. He said that he thinks a little bit about the PD-L1 level. And if he wants to drop 1, he’s more likely to drop the pembrolizumab if the PD L-1 is lower, and more likely to drop the pemetrexed into the fire if he wants to drop 1. I hadn’t really thought about that. Do you do that, John?

DR HEYMACH: In general, I’m more likely to drop the pemetrexed, almost in any case. Because I can tell you that after a year, patients just get weary with pemetrexed. You often see slowly rising creatinine over time. That's actually probably one of the most recent — common reasons, or progressive fatigue, and it just impacts the quality of life over time.

So I think it's tolerability, but pemetrexed, as long-term maintenance, does take a toll. So, I tend to be quicker to drop the pemetrexed.

Now, if somebody was a PD-L1 level of 0, I might be slower to drop it. If somebody is higher PD-L1, then I’m much quicker to drop the pemetrexed. But it's pretty rare that I would drop just the pembrolizumab.

Part of the rationale is also that chemotherapy that you got upfront, the 4 cycles of platinum doublet, may have activated the tumor. And we’ve seen that it actually does turn up PD-L1 level. So, that chemo may have converted a low PD-L1 to an intermediate or to a high, that may be why you get more of the benefit.

DR LOVE: Oh, that's really interesting.

Case (Maria Picton, MD): A man in his mid-80s and previous smoker with metastatic adenocarcinoma of the lung and pleural effusion receives carboplatin/pemetrexed/pembrolizumab followed by maintenance pembrolizumab

DR LOVE: So I want to now go to the first case.

And basically, this is an elderly patient, 84 years old. A bunch of comorbidities. Presents short of breath.

Got a big pleural effusion. Mediastinal shift. And on cytology, has an adenocarcinoma compatible with lung cancer. Past history of smoking, as well. Lot of problems with the pleural effusion. And actually, the patient, in spite of his age, was given carbo/pem and pem, an did pretty well, but he — and he actually stayed on therapy for more than a year, including maintenance pembrolizumab. At which point, his pleural effusion became hemorrhagic and unable to be managed and he ended up going to hospice.

But the question that Dr Picton, who — Maria Picton is a general medical oncologist in North Carolina, the question that she had, and, actually, I thought about you John, when she brought up this question. She’s saying she has a lot of positive experiences with bevacizumab in patients with effusions. And that was this man’s main symptomatic problem. But she didn’t know, like how — can you bring bevacizumab in upfront along with chemo/IO?

So here’s Dr Picton.

DR PICTON: One of the thoughts that I had with this patient is, I wanted to give Avastin because when patients have pleural effusions, pericardial effusions, a VEGF inhibitor usually tends to cause re-absorption of the fluid, they respond better. So I have this question in my mind, should I give carboplatin, pemetrexed, pembrolizumab and bevacizumab? At that time, we did not have the regimen with, atezolizumab and bevacizumab.

For this patient, my thought was to add bevacizumab to his regimen, but since he had a high PD-L1 status, I also wanted to give the immunotherapy. So could I have given Carboplatin, Pemetrexed, Pembrolizumab and Bevacizumab? Is that a regimen that would be valid on this patient?

DR LOVE: So John, any comments about that? And, also the issue of the bevacizumab advantages in patients with effusions?

DR HEYMACH: Yeah. It's a really interesting idea. So just to remind everybody. Bevacizumab blocks VEGF. And VEGF is known as an angiogenic factor, but it was actually initially identified as VPF, that stands for vascular permeability factor, and then, later, they realized those 2 things were the same molecule. So one of the things VEGF does is it controls the permeability of blood vessels. And that's why people have seen improvement, for example, in ascites after giving bevacizumab, where you seem to be getting less leak out.

We actually did a study with a different VEGF inhibitor, looking to see whether it could slow pleural effusions from accumulating, just from VEGF effects here. So, I think it's a really interesting idea.

Now, the only regimen that we have approved for using both immunotherapy and bevacizumab in the first-lien setting, is the IMpower150 regimen. And so that, just to review, is carboplatin/paclitaxel/atezolizumab and bevacizumab. That's a regimen I use a fair bit. Now, if I had my preference, I like carboplatin and pemetrexed as the chemo backbone better than carboplatin and paclitaxel, but I do think, in some patients, having the bevacizumab on really can add benefit. So that's a chemo regimen I use for EGFR mutants, for example, because the pembrolizumab regimen is not approved for patients with EGFR mutation.

So in somebody with a pleural effusion, where I think that is a persistent problem, I will think about the IMpower 150 regimen instead of just a carboplatin/pemetrexed and pembrolizumab, just as your — the questioner was asking.

DR LOVE: So Leora, any comments about this question?

DR HORN: One thing that I have seen is, for patients with high PD-L1 score and who have pleural effusions or pericardial effusions, when I’ve given them single-agent pembrolizumab, several of them have ended up back in the hospital with a quick reaccumulation of their effusion, and then, later on, went to do really well. So those patients, I actually will add, regardless of their PD-L1 score, chemotherapy upfront, but then I’m more likely to drop the chemotherapy and maintain them on single-agent pembrolizumab.

I think it's also interesting that this patient had chronic kidney disease and they were given pemetrexed. I’m curious how poor their renal function is. Because there’s so much hesitation to use pemetrexed because of renal function. But I still use it quite a lot and just dose reduce the pemetrexed in those patients to 400 instead of 500 and use this regimen.

DR LOVE: That's interesting. What kind of GFR would get you to not use pemetrexed, Leora?

DR HORN: Probably less than 30.

DR LOVE: Interesting.

Case (G Richard Polkinghorn, MD): A man in his early 50s with metastatic squamous cell carcinoma of the lung receives ramucirumab and docetaxel as second-line therapy

DR LOVE: So I want to bring up another case — we can come back to this one if you have any other comments.

But, Ben, this is a patient who’s 51 years old, a young patient. This patient has actually metastatic squamous cell, presented with mets in the brain, liver, and bones. PD-PD-L1-negative.

Got radiation therapy to the brain. Got put on carbo/paclitaxel/pem. A little bit of a response. Really didn’t do that great and then had disease progression. And currently receiving a second-line therapy with docetaxel and ramucirumab. And the questions that Dr Polkinghorn, from Maine, had, was 1 of the questions that you kind of a little bit, why is that we can use ramucirumab in squamous cell, but we can’t use bevacizumab, and is it safe do that? This patient seems to be progressing very well. So that's 1 question: Is it safe to use it a patient with squamous cell? Do you use it?

So here is Dr Polkinghorn on Zoom from Augusta, Maine.

DR POLKINGHORN: So,= I have 2 questions. First, in the setting again of squamous cell carcinoma, metastatic, how comfortable are investigators using ramucirumab in a disease where we do not, by guidelines, use bevacizumab — it's not approved?

And my second question is, if he progresses through docetaxel and ramucirumab, what would be your 3rd line option for this relatively young gentleman? Would you consider gemcitabine, even though he has really blown through cytotoxic chemotherapy? And would afatinib be of interest to you?

DR LOVE: Ben — this is a younger patient: What might be next? I don’t know what’s going on exactly right now with clinical research, but I know it's definitely hit an obstacle, but are there trials, at least that were available prior to COVID, that a patient like this could consider? And what about off trial, how do you approach third-line therapy?

So Ben, can you take a crack at this discussing that?

DR LEVY: Yeah, I think that this is a patient that has been appropriately managed. I think that we at least know from the REVEL trial, that roughly 20 to 25% of patients in that trial — this was a trial evaluating docetaxel/ramucirumab compared to docetaxel in platinum-refractory patients. We know that a quarter of these patients were squamous cell and that the drug — we didn’t see any bleeding events that would prohibit giving this drug in a squamous cell patient population.

I think, and I’ll defer to John on this because he certainly has a tremendous amount of experience on bev, but clearly, we saw some safety signals in the Phase I/II with bleeding. And I’m not sure we have seen these with ramucirumab.

So I think, based on at least the Phase III data, we certainly, I think, a patient with squamous cell should be offered the opportunity to see the combination of docetaxel and ramucirumab. And, Neil, as we’ve talked about before 1-on-1, I think the challenge with this regimen is not the ramucirumab, it's the docetaxel. Which kind of leads me to the next point of, what do we really need in this space and really good clinical trials. And I think clinical trial opportunities exist of course in the second line for patients who are progressing on platinum-based chemotherapy. And we try very hard to put all of our patients, second line, on clinical trials rather than default to standard of care. And that's because every subsequent line of therapy you lose an opportunity to treat the patient. So, I would say, even in the second line here, our default would be to do or try a clinical trial.

But, in this patient who got docetaxel/ramucirumab, what I would do next off study would most likely, if the patient wasn’t responding to docetaxel/ramucirumab, would most likely be something like single-agent gemcitabine, given that that is an active agent in squamous cell. I think, on study, the sky is the limit. I think there are multiple competing strategies out there. I will just throw out some of the ones that we’re doing, and I know is being done across some of the institutions that these speakers are from.

But bottom line is, this patient is being appropriately treated with docetaxel/ramucirumab. Off study, third line I would offer gem.

DR LOVE: So Leora, this patient did pretty well on immunotherapy with chemotherapy. What about post-immunotherapy/immunotherapy? In other words, for example, ipi/nivo or a combination like that?

DR HORN: I think everything is being looked at. I’m not sure that I am excited about any of it. I think we’ve just had so many big failures, that the post-IO, every single combo is out there. But there’s nothing right now that hits me as a clear winner. Probably what I’m least excited about is nivo/ipi, and we’re still seeing that.

I’ve actually seen patients come and getting that post-failure of IO. We have a bunch of trials going on looking at combos post-IO, but nothing seems to be hitting it out of the ballpark the way the single agent drugs just did on their own in the second line and chemo failures.

DR LOVE: So interesting about ipi/nivo. So you’re saying you’re seeing people having that suggested to them. When they come to you for a second opinion, do you go, well, that's an evidence-based strategy but I don’t really do it. Or, that's just not a good idea.

DR HORN: Probably more excited about docetaxel/ramucirumab than I am about ipi/nivo. I do use a reduced dose of docetaxel, the 60, the Japanese dose rather than the 75 mg/m2. But I generally will talk patients out of that and talk them into a clinical trial if they don’t want to go to a chemotherapy strategy.

DR LOVE: So Melissa, one thing — again, just from the general medical oncology perspective because Ben’s comment about the docetaxel there is well taken, older patients in particular. But general medical oncologists are used to using ramucirumab and paclitaxel for gastric cancer. They use it all the time in second-line therapy. It seems like it would be a natural substitution, but I know there’s a lot of reluctance based on data. Is that something you ever consider?

DR JOHNSON: It's a good point, Neil. I think the paclitaxel dose, if I’m not mistaken, is lower and I think it's weekly for pac plus ramu as opposed to the docetaxel that's every 3 weeks. So that’s one reason — that’s one way that taxanes are more common.

I agree with Ben and Leora’s comments. I tend to think twice about docetaxel/ramucirumab. It adds 1 month of overall survival, 1 month of PFS and it burns up some — I think the combination effort does have toxicity.

I think the community, oncology community, is borrowing the data from melanoma when the use nivo/ipi in the second line. I think there, we do see that you can cycle through immunotherapy drugs. It does appear to be more difficult to do that in lung cancer. And as Leora said, some of these efforts to re-activate the immune system or reverse immune evasion or resistance have not yet been promising.

I did want to add, to all of the trial comments that antibody drug conjugates are a particular strategy that I’m more excited about in the second line and beyond for lung cancer. In breast cancer they're using them. They are FDA-approved there.

We have a lot of antibody drug conjugate trials ongoing, many versus docetaxel. I think that’ll be something that we talk about more in the future, on these meetings as well.

DR LOVE: We did a whole entire symposium at San Antonio on HER2-positive disease. And

I was shocked when I started to see the data in HER2-low and that it was being used outside of breast cancer. Have you seen actual responses in lung cancer to this drug?

DR JOHNSON: So that drug targets HER2. It's an antibody drug-conjugate. Exatecan is the payload. And it is, as you said being used for a lot of breast cancer patients with HER2 expression. In contrast, the trials that are evaluating this in lung have been for HER2 mutated cancer and that's a much smaller patient population.

We are participating in that trial. We have other antibody drug conjugates with other targets that are more prevalent in lung cancer. We have a HER3 ADC, for example, a TROP2 ADC,

There are other, PTK7 ADC. The common denominator among all of these trials and the field in general, is that the target is super relevant. And so, beyond oncogene mutations, it appears that we will need to do biopsies and look at tumor tissue for patients after front-line therapy to understand what the best next treatment will be.

I think that's the next wave. And I’m really excited about it. We’ve seen some great responses.

DR LOVE: John?

DR HEYMACH: So I agree that the patient’s been managed appropriately this far with the right approved agent. But at this point, I would really reconsider 1 important strategy that hasn’t been mentioned, which is actually doing next-generation sequencing here.

And, in fact —

DR LOVE: Oh, for sure.

DR HEYMACH: — we would typically do it for squams upfront. Now, I know there’s some controversy about do you do it routinely for all patients with squamous histology or just squamous non-smokers?

So we just reported a study last year, Vincent Lam is actually the first author of this who’s now at Johns Hopkins, and more than 400 patients with squamous carcinoma, if you do NGS, about 10% of them had actionable alterations. And so, this includes about 3% that have EGFR mutation, about 2% have ROS1 or ALK. And the big one, the one that’s going to become most relevant this year, is about 5% have MET exon 14. Because we expect, in the next year, it's likely that capmatinib or tepotinib is going to become FDA-approved for MET exon 14.

So now, if you’re talking about 10%, that's a substantial group that could really get benefit from targeted agents.

So I would just try to remind everybody, don’t forget about NGS for squamous patients, particularly given that this case, the patient was younger, age 51. We don’t know if he was a smoker or not, but I wouldn’t let that impact. Smoking doesn’t prevent you from getting an EGFR mutation or a ROS1 mutation. So don’t forget about NGS in those patients.

DR LOVE: So just got to follow-up though. When you say you see 10%, that's — in the past, as I said earlier, I’ve heard people say, okay, well, maybe it's mixed. There’s a small tissue. You’re not really sure what it is. But 10% for real squamous cell?

DR HEYMACH: Yes. That's a great question. This is patients who were diagnosed at our cancer center as squamous histology is the bulk of that dataset. Now, is the diagnosis always perfect? Well, sometimes you only get a needle biopsy in the real-world and you’ve got to make a call on it. But this is what happens in the real-world. So this is a real-world experience — patients who are being called squamous are being treated as squamous, and 10% of them have an actionable alteration.

Now, you might say, that's lower than non-squamous, which is 30 or 40%. But I don’t think that's the right comparator. I think you’re saying, okay, does doing this potentially help a substantial number of patients where they could get a targeted therapy?
 
And I think the answer is yes, especially as we have MET exon14 drugs becoming available. To my mind, that's where it becomes even more important to do NGS for your squamous cell patients, when we have the MET exon 14 drugs available on (fades out)…

DR LOVE: So before I ask —

DR JOHNSON: I believe capmatinib was approved this week.

DR LOVE: Yeah, capmatinib was approved like yesterday, I think. Yeah.

DR HEYMACH: Oh, it was yesterday. Okay. I’m a day behind.

DR LOVE: They come flying in.

But just to go back to the question that came up before, John, since you’re the angiogenesis person, why is it that it's safe — I mean empirically they didn’t see bleeding when they gave ramucirumab in squamous — but is there a biologic reason, or just the way they did the trial? Why is it?

DR HEYMACH: Yes, good question. So first of all, empirically, we know it's true from multiple studies. So not only the REVEL study that Ben mentioned with ramucirumab plus docetaxel. We did the ZEST study with vandetanib plus docetaxel with squamous and there was no bleeding signal there in the squamous. There have been studies with nintedanib in the squamous, where it's been safe. All these in the second-line setting, it looks like VEGF inhibitors are safe. There isn’t a big dataset for bevacizumab.

So it’s hard to know if it's a bevacizumab issue or it's a second-line setting. I suspect, I the second-line setting, you could do this safely as long as patients didn’t have big, bulky, central tumors.

I think what happened in the first-line setting, the tumors are very fragile before they’ve gotten any therapy where the vasculature is very leaky, especially when you’ve got large, central vessels. But after patients have gotten initial chemotherapy, you’ve had some debulking, a lot of those fragile vessels that are prone to bleeding are probably weeded out.

So we’re not sure the reason, but there’s multiple studies that say squamous patients can be treated with angiogenesis inhibitors in the second-line setting.

Use of immune checkpoint inhibitors for patients with a history of autoimmune disease

DR LOVE: So I want to move on now and talk a little bit about the issue of toxicity, autoimmune toxicity. In particular, I want to get into the issue of how you deal with patients with prior autoimmune disease. And Ben covered that in his talk, and it's a great talk. It goes through the data that we have on this.

And we put together a few scenarios. And Melissa, maybe you can comment on these first 2, the patients with Crohn’s disease. So, one where it's well controlled, and we see that in terms of the — but they’re on infliximab. So, it looks like only Ben would use an IO in a patient like that.

But, Melissa, can you comment a little bit about how you approach people with a prior history of autoimmune disease in general? And specifically, I don’t know if you’ve had patients with Crohn’s disease, how you figure this out?

DR JOHNSON: Yeah. I think, just like PD-L1 level is on a spectrum, I do believe that autoimmune conditions concomitant with lung cancer are also on a spectrum. And so, I guess this question is getting at how significant is Crohn’s disease on your autoimmune spectrum? I guess what I picked up on, and maybe wrongly since Ben did the research and can tell me why, but I picked up on the fact that the patient was being treated for his Crohn’s disease.

And sometimes I look to a patient that has a history of an autoimmune condition, but is stable, not on therapy, as a reason that I could apply the immunotherapy, in addition to the chemotherapy.

But here, because not only does this patient — this patient doesn’t just require budesonide or something local for the intestines, to keep the intestines quiet, but as systemic therapy? I thought that was a reason that I’d be reluctant to add pembrolizumab. I look at — I guess the risk is potentially that you’d add fuel to the fire and make the Crohn’s disease worse.

DR LOVE: So Ben, it also gets into the mechanisms of how these autoimmune diseases actually occur and how the therapies work, which certainly I don’t know too much about. I’m sure each one is somewhat different. But any broad conceptualization that you have, Ben, in terms of the pathophysiology that you’re dealing with and how you adapt to that?

DR LEVY: I think that we know that PD-L1 and CTLA4, both are probably in play with self-tolerance and they are upregulated on a variety of tissue. And manipulation or alterations of these can lead to autoimmune phenomenon and may explain, a very basic level, why patients who are receiving checkpoint inhibitors may have a flare of a pre-existing autoimmune phenomenon. And the data is — we just don’t have a lot of data. And I agree with Melissa, I think there are autoimmune diseases and there are autoimmune diseases — there’s a spectrum within each autoimmune disease, and then there are differences between them. And I think it has to be an individualized treatment decision for these patients and really deep digging on the severity.

I mean if you look at the data, at least the retrospective analysis from Memorial experience, it looks likes like, even for patients with pre-existing autoimmune phenomenon who are being treated, about 50% of those patients, small numbers, will develop a flare, of which most of those will still be Grade I or II. And when you have that, with the idea that at some point a patient is going to need a checkpoint inhibitor at some point in their treatment, at least I believe, in their treatment continuum for their lung cancer, I would probably consider — I don’t say I would do it all the time, but I would consider offering a checkpoint inhibitor with a very close monitoring. But I don’t think there’s any right or wrong here with — it looks like 80% or a large majority are avoiding the use of immunotherapy. I would certainly give it a go.

And Neil, you and I spoke about this last parting shot. I mean I think there are some autoimmune phenomenon that I feel very hesitant about giving. I mean we’re talking about, in this dataset, myasthenia patients being involved, multiple sclerosis patients being involved. Those are patients I would not consider. For this patient, I think I’m on the fence, but would certainly consider it.

DR LOVE: And as you stated, for example, when we present a patient who has psoriasis, people in general are going to be using checkpoint inhibitors. Although, I’ve heard of a case that ended up in the ICU.

And multiple sclerosis, well, it looks like a couple of people would go with carbo/pem and bev.

So multiple sclerosis. And everybody I’ve asked about multiple sclerosis over the years has said the same thing, that they wouldn’t treat it.

However, I have a case that is our case incidentally.

What about transplant, John? Again, when you think about how the drugs that are used in transplant, how they work and how checkpoint inhibitors work, what would you expect and what do we know? Hopkins has done a lot of research on this. But, what’s the bottom line, John, in terms of giving IOs to people who’ve had transplants.

DR HEYMACH: Yeah, that's another population that we just don’t have a big dataset, but it really concerns me. I tend to be a little bit more aggressive using it in the case of autoimmune disorders. There was a study that was led by Mark Wadd, where it looked like about 20% of people with autoimmune disorders, a little bit more had a flare, but they were manageable and controllable, so you weren’t killing people if you gave the tumors, they had underlying autoimmune disorders. But multiple sclerosis, myasthenia and transplant are the 3 conditions where I’m really concerned about it.

It may emerge that there will be data, slowly, over time that tells us it's safe. But just because the consequences of losing a kidney or something like that, I tend to really use it as a last resort. I think I’ve done it once for somebody who had distant history of a kidney transplant and who was aware of the risks and had been off immunosuppressants at that time. But if somebody was actively on immunosuppressants, that means their immune system is still trying to attack. And so, unleashing that potentially with immunotherapy is a bit of a frightening prospect for me. So I’m still cautious in that setting if they’re still on immunosuppressants.

Case (Dr Polkinghorn): A woman in her early 60s and current smoker with multiple sclerosis and COPD receives nivolumab for metastatic NSCLC and attains a complete remission

DR LOVE: So here is Dr Polkinghorn presenting another case.

DR POLKINGHORN: So this is a 62-year-old woman, longtime smoker, significant COPD — is the mother of one our providers — who has underlying active multiple sclerosis.

So she initially was treated when she was Stage III, with chemotherapy and radiation therapy, chemotherapy being carboplatin and paclitaxel. Durvalumab was not provided at that time at Stage III because of the concern about the multiple sclerosis. However, she then progressed with bilateral lung metastasis, progressive dyspnea, required oxygen and she did seek an opinion regarding the use of nivolumab with underlying multiple sclerosis.

She clearly does not have great oxygen carrying capacity with her ambulation. Her COPD is very significant. So that's why the —

DR POLKINGHORN: — the critical part of this any worsening of her lung cancer in the lungs and she’s in big trouble.

The decision was made because of the lack of options and potential benefit to proceed with nivolumab. The neurologist reduced a lot of her medications of multiple sclerosis that she felt might conflict with the nivolumab, and she had a dramatic response until she obtained autoimmune pneumonitis from the nivolumab. And it definitely had to be stopped. She received corticosteroids. You could see the pulmonary effects of the autoimmune pneumonitis resolve. You could see the radiation pneumonitis continue. But remarkably, now many months after stopping all of that, there are no further lung lesions.

So my questions to the panel would be how do you make these tough decisions with people with autoimmune diseases, yet really potentially could benefit from the checkpoint immune therapy?

DR LOVE: And Leora, just to cap it off, the patient was PD-L1-negative. I actually presented this to Julie Brahmer, and I said have you ever seen this? And she said, well, yeah. I have 1 patient I treated also. And I said, what happened? And she said, the multiple sclerosis stayed the same and she went into CR. So, I don’t know if that's a series of 2. But it just seems like a very deeply difficult situation for an oncologist to be in, to face this decision.

DR HORN: It’s hard. I can add a third patient to that, although I didn’t treat them, but they came to me for a second opinion after their MS got worse on their checkpoint inhibitor. Their cancer is great now, but their MS has required additional therapy.

It's hard because in patients with metastatic disease, it's a quality-of-life issue, and in addition to longevity. And I’m a big believer that quality-of-life trumps longevity. And so, that's great. You can cure their lung cancer, but suddenly they’re in a wheelchair or depending on family for full care. So, I think it's a hard decision, difficult discussion to have with patients.

This patient that I saw as a second opinion, had really pushed their oncologist after chemo hadn’t worked, and they wanted something. And they said they didn’t really care if their MS got worse, if they were alive. But I’ve had those discussions with my own patients, and I’ve said, well, there’s this drug. It's got a 10 to 30% — it's not like a TKI, it's not like this 70% chance of working for you. It's a 10 to 30% chance of working, depending on which second-line trial you look at. And so, I think it's an important discussion to have with those patients. Because those patients are fortunate, but there could have been a far worse outcome.

DR LOVE: Yeah. I think certainly should not be thought about in kind of a cavalier means at all. It also brings up the question, and certainly, maybe hesitate clearly in a Stage III situation where they could be cured. But I think even first line and maybe hesitate quite a bit.

Treatment of metastatic NSCLC with actionable tumor mutations, including ALK rearrangements

DR LOVE: So I want to move on now and talk about next topic, which is targeted therapy. And the other group’s going to do EGFR. But I wanted to present a couple of cases and go through a bunch of the other ones, because there’s so much happening now.

We’re going to start out talking about ALK, and I’m going to let Ben comment on ALK.

As you might expect, alectinib is now first-line therapy for patients with ALK in spite of TPS. I don’t know whether — I assume people are using targeted therapy. I don’t know if you all see people who get IOs or chemo/IO instead of targeted therapy for high TPS. Maybe, Ben, can you just comment a little bit about where we are today with first-line therapy? We have brigatinib out there as well.

Is that a consideration? And what does TPS mean in a patient with a targetable mutation, for example, ALK? And what does it mean in terms of benefits from checkpoint inhibitors.

DR LEVY: I think that, clearly, we’ve got data in the first line, both with alectinib and brigatinib in the ALEX and the ALTA trials, respectively. The most mature data we have is with alectinib. We’re seeing a median PFS versus crizotinib, PFS, not OS, of close to 32 to 34 months. I mean that's pretty remarkable. And that's not overall survival, that's PFS. And I think for these reasons, the more mature data, given the improvement in PFS that was seen, given the intracranial activity of this drug and response rates north of 60 to 70% in the brain, alectinib is a standard of care — is a standard.

I think brigatinib has also been compared to crizotinib. We don’t have as mature data, I believe. Maybe we’ve had some updates recently. But we’ve seen some PFS numerically that is not as robust as the ALEX trial with alectinib.

Both are reasonable options, but I think alectinib, as you can tell from the answers here, seems to be the first line preferred choice.

I think the big question is what do you do next? And that, we’re still wrestling with optimal options post-alectinib. And, of course, and do you biopsies? Liquid or tissue? And then how does that inform treatment decisions with lorlatinib or not, another next-generation sequencing ALK drug?

Now, one quick parting shot about PD-L1. I think that what we’ve learned in retrospective data, at least, is that patients with ALK rearrangements generally don’t respond, at least to single-agent, immunotherapy. We have the immunotarget data out of France, a very large study, looking at response rates to single-agent immunotherapy for ALK rearranged lung cancer. And it generally does not elicit meaningful responses.

So I think that immunotherapy combinations may play a role at some point, and I think there’s emerging trials that are looking at this. But, at least single-agent immunotherapy is off the table for me. And I think we need to optimize sequencing TKIs for patients with ALK rearranged lung cancer.

DR LOVE: So John, we asked people about, in general where they’re going to use IOs in patients with ALK and I think probably other targetable lesions, as well, as we’ll see in — you can see that it's certainly down the line a little bit.

One of the things I’ve heard, I’m not really clear about, John, is that some patients, particularly with ALK that I’ve heard — I think Ross Camidge was telling me about this — they have high PD-L1 levels but don’t respond the way you would expect high PD-L1.

Any comments on that, John, in terms of what’s going on there?

DR HEYMACH: Yes. We don’t know exactly about that population and whether that should drive you towards using immunotherapy. We have some data from EGFR that's somewhat similar, Justin Gaynor put this data together, looking at EGFR mutant patients who had high PD-L1. And even if they had high PD-L1, their outcomes were significantly worse than patients with — that did not have a driver mutation, that had high PD-L1.

So it looks like, in the presence of a driver mutation, even if you have high PD-L1, immunotherapy doesn’t work great. It has some activity. So that's why I put it after platinum doublet, as well as TKIs, further down the road.

Now, the reason for that is thought to be, at least in part, lower tumor mutation burden. So perhaps the high PD-L1s are the ones that have a little bit higher tumor mutation burden, but there may also be mechanism where ALK and EGFR are actively suppressing the immune system.

DR LOVE: So I have to say — I’ve got to ask you, John, to tell the group what you were telling me about, you mention EGFR mutated — I’ve got to bring this up, because it just was so interesting to me. John made the analogy between EGFR-mutated lung cancer and renal cell cancer. Have you heard that one, Ben?

DR LEVY: I haven’t. I’d love to hear it now.

DR LOVE: I asked Corey Langer the other day and he hadn’t — if ever heard of that. No. So John, I never heard of it. Nobody heard about it. But I think it's amazing. Can you explain it to them?

DR HEYMACH: Yes. And I guess part of this is connecting some publications that are out there. And there’s something we’ve presented at meetings that will hopefully be coming out soon.

So EGFR mutants actually, as part of their activation, constitutively drive the pathway called the HIF-1 alpha pathway. Now, HIF-1 alpha, hypoxia-inducible factor, is normally what gets turned on in hypoxia. So in the center of a tumor where there’s poor vessels going on — going in, hypoxia gets turned on. The VEGF gets turned on and so forth.

But in the case of EGFR mutants, this pathway is turned on all the time in all the cells, whether they’re hypoxic or not. In fact, EGFR-mutant tumors, they often have this lepidic growth in the lung; they’re actually less hypoxic than most tumors.

Now, the other tumor that’s similar to this is actually kidney cancer. So kidney cancer has a different mutation — there, they’ve got VHL mutation that leads to HIF being constitutively active. In fact, Bill Kaelin just got the Nobel Prize last year for uncovering this, how VHL activates the HIF-1 alpha pathway, for uncovering this biology with Gregg Semenza and Peter Ratcliffe.

So the reason this actually came to us — we had reported earlier about the EGFR mutant’s driving the HIF-1 alpha pathway. There was a paper a number of years ago, Li Xu, she was the first author, but — and others have observed that as well. When we were doing gene expression analyses of EGFR mutant, we thought they were going to be similar to HER2 amplified breast and other oncogene driven tumors, but actually, what they came out to be more similar to was renal cell.

So we think that for the same reasons that renal cell responds robustly to VEGF inhibitors, we think EGFR mutant does the same.

DR LOVE: I thought that was super cool. And I guess the other thing I was wondering about, after you mentioned that, John, is in renal cell — and other things, endometrial cancer — you’re seeing VEGF TKIs plus IO, so you have pembro/axitinib, for example with renal cell. Is that being looked at in lung?

DR HEYMACH: So we’ve had a couple of trials taking a look at that. Now, here the problem is that several of them have led to more pneumonitis. And in the case of osimertinib with immunotherapy, and that would be a natural — osimertinib is such a good drug in the first-line setting — there, there seemed to be an increased incidence of hepatitis seen when you combined immunotherapy with osimertinib.

So it doesn’t tell us that biology is wrong, combining immunotherapy with EGFR inhibitors, but there may be toxicities.

Now, for the VEGF inhibitors, we do have combinations now with bevacizumab and osimertinib, bevacizumab and erlotinib, and ramucirumab with both of those 2. The RELAY study was ramucirumab with erlotinib. That showed a significant benefit in that study, a PFS difference of about 7 months and maybe FDA-approved in the near future. It got a positive vote from ODAC.

DR LOVE: I’m out there with all the general oncologists, listening to all this stuff, trying to figure out as it comes together. I guess the other thing that's kind of interesting about IO-angiogenic is the atezo/bev combination that just came out with hepatocellular cancer. It's first-line therapy. Immediately. So, similar principle, but we’ll see where that goes.

And just to finish out on that, John, I guess also — we’re going to talk with the other group, but I was curious from this group about the issue of EGFR plus antiangiogenic. And I know there are trials going on there right now. Just kind of curious. Leora, is that a strategy you’d ever consider? Somebody’s progressed on osi, add bev, or that kind of strategy? Or even start up-front with that kind of strategy?

DR HORN: I’ve tried to add it and I hear people say they see responses. I have not seen any patients respond. I don’t start up-front because again, it comes back to quality of life — who wants to come back every 3 weeks for an infusion? It's going to make you a little bit more tired. It's going to bump up your blood pressure. And, eventually, you’re going to get proteinuria because you’re on these drugs for such a long time.

I think that, for that to become a good strategy, is to find an oral VEGF TKI that we can combine. Unfortunately, they tried that with afatinib and nintedanib and that was just toxic. But it's probably because of the EGFR inhibitor more than nintedanib.

And so, if we can come up with a good oral/oral combo, I think I’ll get excited about it. But, while we’re using an IV and oral, why come in every 3 weeks when you could come in every 3 months.

DR LOVE: Particularly nowadays.

Case (Joseph T Martins, MD): A woman in her mid-70s and previous smoker with metastatic NSCLC and a MET mutation

DR LOVE: So we had one other case that I just want your comment on.

This was a patient who actually had a liquid biopsy showing MET D1010N mutation, John. And Dr Martins wanted to know about what that means in terms of therapy.

So here is Dr Martins and his questions.

DR MARTINS: I'm curious about MET mutations and which, what drugs that they like to use for that and now versus then. And I guess I do wonder about blood testing. I know when I prefer to get the tissue when I want to, but it's not always an option. So when you have blood testing, do they have a preferred vendor. And do they how do you know, I guess when you get a positive, do they always trust it? I understand what you're saying with the false negative, we don't ever trust the negative or at least you have unless you have no choice but, but any false positives to be concerned about?

DR LOVE: He actually gave the patient crizotinib, which, on imaging, if the patient’s been stable, but the patient feels much better. But he wants to know A) what’s the status of current management of MET abnormalities, mutations and otherwise, and if you could just comment on where we are with that, as well as plasma. We talked before about using liquid biopsies, how effective is it in something like MET?

DR HEYMACH: Well, the MET space is getting more complicated. And you can think about 3 different types of alterations with MET. One is MET amplifications. Now, MET amplifications can occur at baseline or after resistance, for example, to EGFR inhibitors.

The second is MET exon 14 splice variants, or people will just call them MET exon — exon 14 mutations. Now, the way that those work, they’re often intronic mutations, so between the exons, and so there’s a whole exon that gets spliced out. And that exon that gets spliced out is one that's important for the normal degradation of MET; it's what turns down the MET signals. So with an exon 14 splice mutant, all of a sudden, the levels of MET get sky-high, it stays activated, and so they become very responsive to MET inhibitors like capmatinib, approved yesterday, and tepotinib, which we expect will be approved later in the year.

The third type are point mutants that may act like exon 14 splice variants and activate the receptor, but don’t work in exactly the same way. And we don’t know all of those, but we know some of them. I believe this one is the location just next to the membrane, that it does cause activation of the receptor. It may be responsive to drugs like tepotinib and capmatinib. But this isn’t part of the trial. So this isn’t what we consider a standard MET exon 14 splice mutant.

Case (Margaret A Deutsch, MD): A man in his mid-60s, a never smoker, with metastatic adenocarcinoma of the lung and a KIF5B-RET fusion receives cabozantinib

DR LOVE: And that kind of leads really into this next case. John, this is a 65 year old man with metastatic adenocarcinoma with disease progression after response to carbo pem pem…and then found to have a RET fusion

Here’s Dr Deutsch.

DR DEUTSCH: In terms of RET fusion mutation, I elected to put him on cabozantinib. I guess I would like to ask the experts, since I don’t have many of these patients, would that be — would that be their choice of initial therapy? And if not, what would they choose? And then, at the time of progression, would they go to a second targeted therapy or would they switch to systemic therapy? And if so, what systemic therapy?

DR LOVE: John, the patient is on cabozantinib right now, doing okay, the dyspnea a little bit better, but having a fair amount of side effects. The thing I thought was interesting was that Dr Deutsch, who’s a very knowledgeable medical oncologist in Raleigh, North Carolina, so a lot of academic places around there, understandably was not aware that there are a couple of drugs in development.

Ben, can you talk a little bit about where we are right now in terms of RET? Amazingly, one of these agents, selpercatinib, was just approved today by the FDA.

DR LEVY: I mean, we’ve been fortunate enough to have the LIBRETTO with selpercatinib here. And that's certainly been a great experience for our patients. And certainly, a good drug. A very good drug with minimal side effects and meaningful activity that is very — we started these patients and within 4 to 6 weeks, similar to what we see in osimertinib in the EGFR space, a very selective drug with minimal off-target side effects which elicits meaningful responses, both below the neck and in the brain. And so that's been a — we’ve had a lot of patients on that study. It's an oral therapy with very few side effects.

The other drug that I don’t have a lot of experience with pralsetinib — has also been presented and seems to have meaningful activity. I don’t have a lot of experience with that drug. But, clearly, we’re fortunate enough to be part of this study, and I think we’ll see a publication soon, both not only for lung cancer, but also for medullary thyroid cancer for RET mutations, not fusions. A really good drug.

And RET is one of these fusions where, yes, it's rare, but we have to think about, once you find one, and you get these patients on a drug, it's a different experience than being on cabozantinib. And two, this is a fusion. So you have to be careful about how you’re going to detect fusions and what next-generation sequencing platform you’re going to use to commonly get these identified.

DR LOVE: Just getting back to RET, Leora. I don’t know the extent of your experience with some of these new agents. But what do we know about the 2 new agents that are out there? How they compare to drugs like cabozantinib, both in terms of efficacy, as well as tolerability?

DR HORN: So we actually had the LOXO trial here, with 292. And the other drug is Blu-667. They both have names now. And I think there are 2 striking things. One is, their response rates are better with cabozantinib and vandetanib, their response were in the 30 to 40% range. And now we’re seeing those better responses in the 60 to 70% range with these drugs.

We’re also seeing that these drugs are much better tolerated. There’s less toxicity — cabozantinib is not an easy drug to give. It's hard to give it at the FDA-approved dose. And so, I think that when these drugs are approved, they’re going to be your first-line option for patients.

Activity and tolerability of the selective RET inhibitors pralsetinib and selpercatinib; preliminary results from the Phase III ADAURA trial evaluating adjuvant osimertinib

DR LOVE: So Ben, this is, when we asked the faculty, what’s your first-line treatment for a patient with a RET rearrangement, but we said the patient has a high PD-L1 level, everybody says — of course we said reimbursement issues aside — so everybody says one of the new RET inhibitors, they actually say selpercatinib, except Govindan who would use actually chemo plus pembrolizumab — carbo/pem/pem.

And it gets into the idea of the paradigm, kind of the paradigm that came out of EGFR with I-I-PASS. And then I think there were studies like that in ALK, where they proved it was better than chemo. And then, all of a sudden, I saw you all going, well, if there’s a response rate over 50, 60%, let’s bring it in first line.

So using this as an example, Ben, can you comment a little bit about that paradigm of first-line therapy of metastatic disease, in a targetable situation where you don’t have randomized trial comparing it, for example, to chemo/IO?

DR LEVY: Yes. I think you’re right. I think we all kind of use some loose rules in the absence of first-line data, to suggest that we should use our best drug first. I think that's what you’re seeing here and that all of us feel that. Oftentimes, you only get one chance to get it right, number 1. And number 2 is, these drugs elicit meaningful activity that improves quality of life, which is often underrepresented in the datasets, that we favor using these drugs first.

I don’t know what the metric is to say, well, this study — this drug was evaluated in second line and elicit response rate of 40 or 50 or 60%, thus it should move frontline in the absence of front-line data. I think we all gain experiences with these drugs and realize that they have a favorable toxicity profile and patients feel better on them, and they elicit meaningful responses. And we want to move them, in the absence of front-line data, to the frontline. And I think Dr Govindan’s answer is actually the more academic based on evidence answer. WE have most data with this drug in the refractory setting. I think we’ll learn more and more about it in the treatment-naïve setting. But I think this is medicine as an art and not a science, where we’ve got a good drug and we want to move it up front. And I think this is where we’re heading with precision oncology.

DR LOVE: So John, speaking of targeted therapy, we had a press release that went out, I guess in the last week or 2, about the adjuvant osimertinib trial, the ADAURA study. I think this is the first major Phase III, randomized trial of targeted therapy in lung cancer that's going to report. Which is, reportedly, closed early because of efficacy. It looks like it's going to be presented at ASCO in their virtual meeting, so we’re going to be seeing the data soon.

John, any comments about some of the things that you’re going to be looking to in the dataset? As soon as I saw that press release, I arranged an interview with Tom Lynch, incidentally, who I’ve been wanting to talk to anyhow because I was curious what happened when he got to Seattle. But in any event, I’m just kind of curious, one of the things I was talking with him about is the same thing we were just talking about, are we going to have to do adjuvant trials? We know there are some out there, ALCHEMIST, etc.

But Ben, are we going to have to do adjuvant trials on all these things? Or if ADAURA is really positive, do you think RET and everything else is going to come into the adjuvant setting, maybe after chemo?

DR HEYMACH: I think this is going to be a hugely important results of the field, actually for a couple of reasons. So right now we don’t have any targeted therapies that are approved in the adjuvant study, in the adjuvant setting. And so, even though we know that drugs like osimertinib are far, far more effective than chemotherapy, we’re not using osimertinib in Stage III or Stage I and II resectable disease.

So having the first one approved, assuming the ADAURA study is positive and osimertinib beats placebo in the adjuvant setting, which I’d be surprised if it doesn’t, that's really going to change the game because it means for early stage disease, molecular profiling will have to become part of our standard paradigm. Right now, you can cut out an early stage non-small cell and you can say I don’t care what the molecular profile is. I don’t care what the PD-L1 is because the only approved adjuvant regimen right now is chemotherapy. Well, we’re going to have evolution of the field in a couple of ways. So, if ADAURA is positive, well then getting molecular profiling will become part of it.

Now, there’s also the ALCHEMIST study, but that's going to take a long time to read out with ALK, as well.

And then in the near future, we’re going to have immunotherapy studies reading out as well, that will determine if you’re not a driver, should you be getting immunotherapy, either in the neoadjuvant setting or the adjuvant setting?

So I’m very excited for these results to come out. And I think it’ll be a huge benefit for patients. I think it's going to open the gates essentially, or the dam will break, where now we’ll be molecular profiling all the early stage patients and we’ll find out who’s got ALK and who’s got RET and so forth. They’ll be similar trials coming along. And I think a couple of years down the road, we’re likely to see these neoadjuvant studies reading out as well. And I think a few years from now, for early stage, we’re going to be using immunotherapy and we’re going to be using TKIs, where appropriate.

So I think it's going to be a huge gain for patients.

Management of metastatic NSCLC in the era of COVID-19; preliminary results from the TERAVOLT registry on the impact of COVID-19 on the care of patients with lung cancer

DR LOVE: So I just want to finish out and ask you all if you have any comments or thoughts about the issue of COVID and lung cancer. And, Melissa, we saw this really interesting virtual session that AACR did last week. They had a whole session on COVID. There was this incredible presentation that Leora was very much involved with that.

We’ve been talking with everybody we can about how management algorithms now are being altered, depending upon where people are, of just patients without COVID. And we can talk a long time about trying to avoid infusions, don’t be bring people in, telemedicine, etc. But the thing that I thought was so interesting about this effort that Leora was involved with, is they were looking at people with lung cancer and COVID. And I think it's just the beginning of looking at that. And I don’t even know how many patients there might be out there in peoples’ practices.

But Leora, maybe you can just start a little bit and summarize what your take was on some of the more important data that came out of that, and how you think you ought to be affecting peoples’ practices right now?

DR HORN: Yes. So it's a little hard because this was for patients who were symptomatic with COVID and lung cancer. And so, what we don’t know, and I think we’ll get a lot of data on in the future, is the asymptomatic cases — because we’re now screening everybody at the start of chemo to make sure that they don’t have COVID. But in the initial dataset, and we will be presenting an update at ASCO which shows a lot more information than what we had at AACR.

We had symptoms at presentation were very similar to what you would with lung cancer. So, just mere cough, fever. The shocking information that came out was that for the majority of patients who were admitted who, if they required ICU, they were not transferred to ICUs. The reasons being — we’re still sorting some of that out, but some of it was their health systems that were overwhelmed and could not accommodate a patient with Stage IV lung cancer who had a terminal illness despite the better prognostic information we have about lung cancer. And so, they were having to triage these patients.

We also, unfortunately, saw a 34.6% mortality for lung cancer patients who were admitted to the hospital. And in our initial dataset, the only patients who were at decreased risk were those patients who were potentially on a TKI, they would less likely be admitted. But we didn’t find a discriminator in terms of type of chemotherapy or type of comorbidity for admission or death from COVID.

We’re now up to almost 500 cases in the database. We’re starting to get a lot US sites on board. I think we’re up to 26 countries and over 200 cancer centers. And so, that's changed. We actually have found treatment regimens that do appear to be associated with increased morbidity and mortality from COVID and some interesting prognostic features that we’re going to be presenting at ASCO.

DR LOVE: So one of the things that I heard in that first presentation, and we’ll be really looking forward to the second one, is it sounded like the IOs weren’t a major issue one way or the other. It seemed like that was the case.

Ben, any thoughts about your speculations about how IOs effect the biologic environment, and how that relates to some to somebody with COVID?

DR LEVY: Yes, this is purely speculative, but IOs may alter or dysregulate the immune system and that may lead to a lower threshold of either getting the virus or having complications. I have not seen this in my practice. Thankfully, most of the patients we’ve diagnosed with COVID with lung cancer, have been doing okay. I think it even begs the question with chemo. I mean chemo certainly causes neutropenia, but is neutropenia really relevant for viral infections?

We had a conversation about this on our Command Center call at our institution. I think, scientifically, yes, it could, but it's more lymphopenia that may, at least scientifically, be aligned with susceptibility to the virus. I mean I think the parting shot here is just that we are at a new environment where we're trying to learn best guess, not based on any scientific evidence, how we should alter our treatment paradigm for these patients.

DR LOVE: So getting back to you, Leora, because I think you’re most familiar with these data that are coming out. For practical purposes right now, if you had a patient who tested positive for COVID that was asymptomatic, how would that affect — and putting aside social distancing and all that stuff that I know you’re considering to start with, but would it materially affect the way you would approach adjuvant therapy, chemoradiation therapy, therapy for metastatic disease?

DR HORN: So if they haven’t started therapy and they test positive, we’re delaying their therapy and then retesting them and starting therapy at that point. You worry a little bit with patients with locally advanced disease, are they going to become metastatic? But honestly, if delaying treatment 2 or 3 weeks, and they become metastatic, they were metastatic to begin with.

And if patients are asymptomatic, but have a test for one reason or another, there are some patients who are just going to get tested because they’re curious and you can in Nashville right now, if they test positive, we’re just putting them in a separate area of the infusion room, but we’re continuing them on whatever therapy they’re on.

Genomic profiling for patients with metastatic NSCLC; role of liquid biopsy

DR GOVINDAN: I will favor using tissue-based testing, using NGS, where you can get not all these days, DNA sequencing in the cells, but, also, RNA sequencing as well as the mutation burden, multiplex analysis, including PD-L1 testing and MMR mismatch deficiency.

So it would be nicer to get a comprehensive analysis of the driver mutations, along with other important pieces of information. That will be ideal.

But in reality, a lot of our patients need quick answers. They want to move on. We can do liquid NGS or liquid biopsy, but the trouble is that in about 30% of the time he could have an actionable mutation that may not be revealed in the blood-based assays, particularly for those patients with the low volume disease, lack of significant visceral involvement. So, we have to be mindful of that.

If you do both at the same time, clearly you are increasing the cost of these testings, so you have to mindful of that as well.

The other approach is to serially, PD-L1 testing, EGFR, ALK, ROS, RET. And then if they are negative, do a larger multiplex tissue-based testing.

So I think it is an area where we don’t have 1 correct answer, in my view. It depends on the context. It depends on the burden of the disease and how sick patients are.

DR LOVE: So Hoss, getting back to this issue of the sicker patient who really don’t want to wait on. One of the things that I’ve heard, and I was curious of what your thoughts are about it, is in a patient like that to maybe start the chemo, but hold off on the IO until you get back mutation status, including EGFR, with the concern that if you start an IO, even if the PD-L1 level is high, and then you find out the patient is EGFR, you’re not going to want to give a targeted therapy after an IO. Any thoughts about that?

DR BORGHAEI: Sure. That is something that actually I’m doing in my clinic if I really do have to start somebody on the treatment and I don’t have the molecular data, my preference is to start with chemo, for reasons that you just exactly said. And that's as a result of the newer information that has emerged over the past couple of years, suggesting that if you do use IO and then want to switch to a targeted therapy, you might run into additional toxicities.

Not all TKIs seem to do this. Obviously, there is again variation in that. But the ones that we use commonly in lung cancer, at least in my experience, such as osimertinib or alectinib or drugs like that, do seem to be associated with a higher risk of additional toxicities. And I’ve actually have had a couple of patients, unfortunately, who, when I wasn’t doing that, ended up having severe toxicity. So, I have adjusted my practice to suggest it — so, start the platinum doublet chemo while the molecular testing is pending and then adjust the treatment, based on the results of the molecular testing.

DR LOVE: I’m curious what you saw in your 2 patients. Initially, the report seemed to focus on pneumonitis, and then I started to hear about other things.

DR BORGHAEI: I think there are different toxicities that can happen. In my case, I had 1 person with severe rash, was really difficult to manage. And then I had a patient who developed Grade IV hepatotoxicity. It was again, very scary, leading to admission and multiple other offshoot of all of that.

So in my case, it wasn’t pneumonitis, but it was the rash and the hepatitis.

DR LOVE: Julie, any comments on this? And any thoughts about how long theoretically you’d want to wait before starting targeted therapy after an IO?

DR BRAHMER: I think from an IO standpoint, to start a TKI, I mean theoretically it's going to take a long time to get out of your system. So, some folks would say 3 months with a half-life is anywhere between 2 to 3 weeks for just the antibody, and the actual immune effects can last for months. But in some cases, there’s just no way you can wait, and you have to start the next drug.

Ideally, it would be nice to at least wait through at least 1 or 2 half-lives, but again, you cannot guarantee that that heightened toxicity or immune response, or potential cross toxicity could not — it potentially would happen almost at any point.

DR LOVE: Eddie?

DR GARON: I think 1 of the big questions that I would have in terms of the speed of turnaround is, we’ve actually set our system up where we get an EGFR, ALK and ROS1 test very early. I think that it is reasonable, in my mind, to divided tests into ones that you need to have an immediate answer on and ones where, if you don’t have an immediate answer you don’t think you’ve harmed the patient by not knowing that result upfront.

DR LOVE: So that approach is kind of what I thought was going to — it made sense, until I started to hear about all these other targetable mutations, Govindan, that seemed to respond very well. And I see people using first-line therapy, for example, BRAF.

So can you really say now that if you, put aside ALK, ROS1 and EGFR, that you really have eliminated the first-line targeted therapies? What are you using first line?

DR GOVINDAN: So obviously, if you have a targetable mutation, I tend to go with, obviously, targeted therapies, definitely for EGFR, ALK, ROS and RET.

BRAF, I typically use this in the second-line setting, but I can see the point where you’re seeing case or using it in the front-line setting.

But, back to the issue that you brought up with Eddie, Neil. Can we get the test earlier, sooner, especially the important ones like EGFR, ALK, and ROS? I think 1 way to do that is to do FISH based assay for the fusion, and then get an EGFR by PCR-based testing. And then, if they are negative, then you can send for NGS. That's the current model used in many places.

The other approach is to get the NGS results done sooner. And I think I won’t be surprised if, in a couple of years, we can get the turnaround time shortened to about a week or so. As it is, now we can get a blood-based test, NGS results — limited panel though — within 3 to 4 days. And the tissue-based testing, we could potentially get this in about 7 to 8 days. And technically, it's very, very feasible. So that may be the way to go, to be honest with you, because of tissue problems.

DR BRAHMER: Yeah, that's what we’ve started to move towards getting a liquid-based test as soon as they walk in the door — as soon as the patient walks in the door, while the tissue-based testing is — for us, it still typically takes 14 days. So, in order to try to not miss those targetable changes, we’ve been using the liquid-based assays while the tissue is being processed and the NGS is — and we’re waiting for the NGS on the tissue.

DR BORGHAEI: And that's exactly what we’re doing, too, for the same reason. I mean a liquid panel comes back a lot of faster. The tissue, in-house, takes about 12 to 14 days. So, by having the blood-based assay, and having most of the actionable mutations, at least you have some idea as to where you’re going. I mean you do have to keep in mind that not all these tests are 100% accurate, so there’s always a false-negative with the liquid biopsy and with the tumor. But at least in 70, 80% of cases, you have some information to start talking to the patient about what treatment options you have for them.

DR LOVE: So Eddie, another thing we asked about was the issue of testing in squamous cell. In the past, I kind of really wonder whether it was even worth doing. And then I heard people starting to say, well, you can find targetable mutations. And then I was like, is that really squamous or something with an inadequate tissue, et cetera?

If you know that you have a squamous cell cancer, Eddie, what do you think should be done in terms of genomic analysis?

DR GARON: Sure, the guidelines in this setting recommend that you do not necessarily routinely get a broad panel of tests. Now, that has generally been what I have done as well. On the other hand, if you have patients that give you any reason for suspicion that they may have targetable mutations, it's felt that that significantly raises the likelihood that they would have a targetable mutation.

And, in my own practice, for instance I have had several people, Asian women with squamous cell cancers, who have never been smokers, and almost all of those patients have turned out to have an EGFR mutation. So, that's obviously anecdotal. But the current guidelines would suggest that a patient who is a non-smoker, patient who has features that are not consistent with squamous cell carcinoma should undergo testing.

But there is a counterargument as well, that if the more data we get perhaps we’ll be able to learn more about targetable approaches in squamous cell carcinoma. But right now, in the patients with a significant smoking history, I have not necessarily been testing.

Therapeutic approach for patients with NSCLC with no targetable mutations

DR LOVE: So I want to move on now and talk about management of patients without targetable lesions, wild-type disease, starting out talking about first-line therapy. And, Govindan, we’ve been asking these kinds of questions ever since checkpoint inhibitors came on the scene, which actually wasn’t that long ago;

Govindan, you see that there are some faculty members, even with a high PD-L1 level, are inclined, in general, including, well you, or at least in terms of consideration, will consider using chemotherapy. WE hear a lot about looking at the symptom status of the patient, the age of the patient. How do you approach these patients, Govindan?

DR GOVINDAN: So I think, in general, I feel — I think there are 2 options for patients with PD-PD-L1 50% or more: single-agent pembro or combination of chemo plus pembro or by pembro/pemetrexed maintenance.

I think the issue of which 1 is better is unsettled at the present time, the way I understand. So, in general, if patients are elderly, they have toxicities, comorbidities, renal dysfunction, not ideal candidates for chemotherapy, I think, in my opinion, those are ideally suited for single-agent pembrolizumab. All other patients I give combination chemotherapy with immunotherapy, regardless of PD-L1 expression. That's our approach.

DR LOVE: So, Hoss, I’m curious, are there situations where you use other combinations, particularly involving a taxane? And the 1 situation I wonder about is patients with compromised renal function. But, in general, how do you decide upon the chemo, Hoss?

DR BORGHAEI: So, histology, obviously, comes into effect, if you’re talking about the non-squams. Obviously, we’d like to use pemetrexed-based regimens, just because of the tolerability issues. But you’re absolutely right. If somebody has a compromised renal function, then the use of pemetrexed is not possible. I think if you go by guidelines, if the creatinine clearance is less than 42, 43, you’re not supposed to use pemetrexed.

So in that setting, my preference has been to use paclitaxel, although nab paclitaxel also has gained a lot of popularity among some of my colleagues here. But, in general, I’ve preferred to use a carboplatin/paclitaxel and pembro approach if I cannot use a pemetrexed-based approach.

DR LOVE: Julie, you can reflect a little bit on your thoughts about this question? And what other situations, for example, does it matter to you how high — how far above 50% the PD-L1 is? Do you treat somebody differently or approach someone differently if they’re in 90% range, for example?

DR BRAHMER: There is data out of the groups in Boston that showed the higher the PD-PD-L1, and the fact that the majority of the response on, at least in some of their data, is in those patients with extremely high PD-L1. But for myself in general practice, if you’re above 50 and I feel that you can tolerate single-agent immunotherapy and I don’t need a significant response rate, I will use single-agent pembrolizumab.

DR LOVE: Any data on time to response, Julie, in terms of — when you add a checkpoint inhibitor, or a checkpoint inhibitor alone versus combined with chemotherapy? I guess a lot of people think that chemotherapy works fast, or faster. What do we know about speed of response with IO?

DR BRAHMER: If you look at the data from single-agent IO versus chemotherapy, the time to response is the same, it's within the first 2 cycles. So I think you can bring more patients into having a response by adding the chemotherapy to the immunotherapy, but I think the time to response, at least based on single-agent IO data, it's very similar to chemotherapy.

DR LOVE: So Eddie, I think one of the most common, if not the most common question I got from these general medical oncologists, related to duration of therapy. It's really amazing to here general oncologists say, I’ve got these people that are out 2 years with metastatic — just that, starting a sentence like that is impressive enough as it is. But, first of all, Eddie, what do we know about — we’re seeing more and more data and longer-term follow-up of these patients, about 5-year survival, the possibility of cure. And how do you approach — I see that, in general, you’re using indefinite therapy.

Others are stopping. What are your thoughts about that, Eddie?

DR GARON: Yes, so 1 thing is that, unlike our colleagues in melanoma, a complete clinical response is very rare in non-small cell lung cancer. So, although the question is fair, we rarely see this. What we generally see is a partial response radiographically. And my initial interest in this area was with the KEYNOTE 001 study. That study treated forever, and, in fact, most of my clinics, I still see somebody who’s from the 001 study. We actually rolled them over and now have switched them — now it will be less true, because we’ve switched them to the every 6-week dosing very recently.

But the main thing that gives me pause is the CheckMate 153 study. It is very difficult to evaluate duration of response. And the CheckMate 153 study, to me, is maybe 1 of the best datasets you’ll ever get on this. Patients who were doing well were randomized at 1 year to continue beyond that year or to stop after a year. And there was a very clear difference, with respect at least to progression-free survival, in favor of continuing. And, in fact, the people who progressed after it stopped and then resumed, generally didn’t do very well. And so, the pembrolizumab program has generally made 2 years their cut point. And some people will say well, 1 year wasn’t enough, 2 years is enough. But to me, this natural extrapolation of indefinite being better than 1 year is not, 2 years must be the right number.

DR LOVE: So Hoss, Eddie referred to the recent approval of pembrolizumab at a 6-week interval. And I’m curious — when I saw that I was kind of thinking, particularly with COVID going on, but just in general, I sort of figured everybody was just going to do it? Is that the case? Are you switching over or thinking about it?

DR BORGHAEI: We’ll switch some of our patients, but that is pretty much in response to the COVID situation and the fact that we don’t want to have patients around, traveling, sitting in waiting room areas, sitting the infusion room areas. So I think the approval was well received because we are dealing with all of these issues related to COVID.

Do I like every 6 weeks? I mean it's more convenient for the patients. The data seems to suggest the toxicity and efficacy is really no different than the other, the traditional every 3-week administration of the drug. So, it does make sense for patients to come in less often.

Do I think that we would have gone to this as quickly as we have if it weren’t for COVID? I think some of us would have been probably a little bit more hesitant to switch all of our patients to this. Case in point, I have somebody that I’m a little bit nervous, about and I like to see the patient the patient every 3 weeks, just because she’s dealing with issues that I’m not sure are really toxicities or related to something else.

So in that patient I might not be as comfortable saying, okay, come back in 6 weeks. I might want to see that patient in 3 weeks. But in majority of the ones who are a year, year and a half on treatment and doing well and are reliable patients in terms of calling if there is something going on, I think switching to a 6-week regimen actually makes sense.

DR LOVE: It's interesting, Julie, because I kind of figured everybody was just going to be like, okay, let’s do it. And Hoss’ answer too, made me think that not everybody’s all that — yes, COVID, and all that — but I feel like there’s some kind of hesitation, maybe people want to wait or whatever. Do you feel that, Julie, or you’re ready to go?

DR BRAHMER: I think for those patients who are just starting out, I don’t feel comfortable doing every 6 weeks, just in the fact that if patients are going to get toxicities, it's going to tend to be in the first couple of months. So, I would, even I’m doing every 6-week infusions of pembrolizumab, I would want to at least have some type of early- or mid-cycle toxicity assessment. Now, we’re all getting very good at doing videoconferencing, but trying to figure out different ways to be able to catch toxicities earlier while patients are remote, that would be helpful. But that’s part of the reason why, at least upfront, I’d feel more comfortable with more often infusions, since I’m going to bring those patients back anyway at 3 weeks.

But, if someone’s doing well for a long period time, even just beyond 6 months, I would feel comfortable going to every 6-week infusions. And I think that really is the advantage to that, from a patient standpoint.

DR LOVE: So another issue that I’ve heard questions about, Eddie, as it relates to maintenance, and the question of do you need to continue both drugs? And maybe you can comment — and, of course I’m talking about carbo/pem and pem. Can you talk a little bit about how you approach that decision?

DR GARON: Sure. It's interesting. There is data from flatiron databases now that say a large percentage of physicians are not giving the pemetrexed in the maintenance.

My feeling is that you conduct a study to answer a question. And there was 1 regimen that was tested against another regimen and that regimen showed almost unprecedented superiority with a hazard ratio for survival at the initial readout of — there was a hazard ratio of less than 0.5, which was really incredible.

I’ve tended to stick very closely to the regimen in my own practice. The 1 place where I think you could make a strong argument to say, particularly dropping the pemetrexed, would be if you went into this trying to decide in someone, for instance, with high PD-L1 expression, that you were on the fence between adding the chemo or not adding the chemo, based on your feeling that you needed to get a quick response. I can certainly imagine under those circumstances dropping the pemetrexed during maintenance, saying I just wanted that chemotherapy upfront in order to take my best chance at controlling disease at the beginning.

But outside of that, particularly in patients with low or no PD-L1 expression, the data with single-agent PD-1 inhibitors in general hasn’t been as great as we want, certainly. And so, I’m not confident we can achieve the same things by just continuing the pembrolizumab in that setting.

Case (Dr Deutsch): A woman in her late 60s with metastatic squamous cell carcinoma of the lung and a high PD-L1 tumor proportion score (TPS) of 60% initially treated with pembrolizumab monotherapy receives carboplatin/paclitaxel and maintenance pembrolizumab upon disease progression

DR LOVE: So why don’t we get into a case here.

This is a 67-year old woman who’s from the practice of Dr Margaret Deutsch, who’s in North Carolina in Raleigh. And this patient presented with squamous cell, adrenal met and had a high PD-L1 level.

And got pembrolizumab and had an objective response. Did well. And then had disease progression in the chest. And basically, the patient got chemoradiation therapy and was restarted on maintenance pembrolizumab.

Here’s Dr Deutsch.

DR DEUTSCH: So in this patient who did well on initial immunotherapy for almost 2 years, she had a local relapse in disease and was treated locally with chemotherapy — combined chemoradiation therapy. And I put her back on pembrolizumab as maintenance therapy. So, my question is, what’s the utility or is — is there utility in going back to pembrolizumab therapy after a — a limited area of recurrent that can be treated by other modality?

And then the second question I had is, once she progresses on disease, because inevitably she’s going to progress on the pembrolizumab and may not have a solitary area of recurrence that can be treated by another modality, then what’s the next systemic therapy that's appropriate for her? Would that pac — docetaxel and ramucirumab? Or would people utilize carboplatin and gemcitabine? So I guess the next systemic therapy option would — would be of interest to know what the experts would think.

And Govindan, Dr Deutsch wants to know is this something you would have done? And, also, the whole issue of isolated progression on checkpoint inhibitors when you use local therapy and keep going? What kind of local therapy? So Govindan, can you comment, please?

DR GOVINDAN: So we would have answered this question very differently, in my opinion, 15 years ago, but now I’m not so sure on thoughts that are evolving, at least mine. This would have been a really unconventional approach to give somebody definitive chemoradiation in the setting of a metastatic disease. But if you think about eradicating some clones, especially with the use of premium therapy, here, if it's adrenal mets alone with right upper lobe mass is 1 thing, but it looks like this person may have had bone metastases as well. So that really raises a really complicated question. These patients probably have far more extensive disease than what we appreciate, even when the PET scan 2 years later shows nice response.

So I would have probably given radiation alone rather than chemoradiation if this person had bony metastases and widespread disease. On the other hand, in May 2017, this patient had oligometastatic disease with right upper lobe and adrenal metastasis. The adrenal metastasis resolved. No nodal involvement. You could make it even more complicated by raising the question, should we have done surgery, especially in the absence of mediastinal lymph node involvement.

And there have been a number of patients published, all small, from various institutions describing benefit of surgery in these situations, mostly in metachronous rather than synchronous renal metastasis. But I think you could do various different ways.

So to answer your question, in this patient I would have probably just given radiation alone and then perhaps resume pembrolizumab, hoping that the almost never seen abscopal response will kick in, following radiation treatment in this situation.

DR LOVE: Hoss, can you comment a little bit more in this issue of oligomets and how you approach in this patient having an adrenal met? Do you take those out? Do you radiate them? We had a case the other day of a patient who developed a node in the porta hepatis at the end of durvalumab consolidation after chemoradiation. They radiated that.

DR BORGHAEI: Sure. Again, and this is an evolving area. There have been a number of studies as you all know, that have looked at this oligometastatic disease and how to manage it. The definition of oligometastatic disease is — some words under discussion, is it 3 sites? Is it 5 sites? There are trials now, looking at more than 5 sites for local therapy. And by local therapy we often mean radiation, formally the stereotactic or otherwise.

So, in our practice, probably this patient would have gone to the OR. Probably the adrenal gland would have been either surgically removed or by radiation, depending on the performance status of the patient or how fit the patient is. And same thing probably, the lung followed by some sort of a pseudo-adjuvant treatment. I call it pseudo-adjuvant just because, in this case, it's really metastatic disease with oligo met.

On the other hand, if you have somebody on immunotherapy, much like what we do for patients who are on targeted therapy now, they have fantastic response and there are 1 or 2 sites that are misbehaving, then approaching them with stereotactic radiation I think would make sense. And, again, in the targeted therapy world, we have pretty good evidence that PFS, too, can be pretty significant if you control the disease elsewhere with the stereotactic approach.

So again, I think the same kind of principles might apply here for patients who are receiving immunotherapy. In that particular case, I agree with Govindan, I probably wouldn’t have done concurrent chemoradiation. I would have probably just gone with the radiation in that case. But the strategy seemed to have worked for that patient, so it's good.

Case (Dr Picton): A man in his mid-80s and previous smoker receives carboplatin/paclitaxel/pembrolizumab as first-line therapy for metastatic squamous cell carcinoma of the lung

DR LOVE: So Eddie, here’s another case.

This is from Dr Maria Picton, also from North Carolina, and this is an older patient, an 86-year-old man, a previous smoker, who actually had Stage IIA squamous cell of the lung that was removed 3 years ago, but then came back with metastatic disease.

The patient got carbo/paclitaxel and pembrolizumab. Did well. Actually, had a very good response to that. But, unfortunately, now recently has developed progressive disease, although the patient is clinically stable. And Dr Picton right now is trying to decide what do. This is an older patient, a little bit frail.

Here’s Dr Picton.

DR PICTON: I’d like to know what would be the next step in treatment. I know that he already is progressing on immunotherapy, so the next step would be chemotherapy. Eighty-six years old, and he's tough. That's my question. I think that I would expose him to carboplatin paclitaxel again versus single agent vinorelbine or gemcitabine.

DR LOVE: Can you comment, Eddie, first of all, on the case itself? And then, what you might be thinking about right now?

DR GARON: Sure. So one thing that I think is interesting about this case is that, particularly since the approval of pembrolizumab as monotherapy in patients with any PD-L1 expression, there’s been a push towards putting patients on more monotherapy, particularly if they are older. But it sounds like this is a case of an 86-year-old, who the practitioner felt would be able to tolerate chemoimmunotherapy, and actually did quite well with it. And I think that's important to note.

I think that in terms of the duration, in general there’s always a thought if somebody — if you stop the pembrolizumab and then they have disease worsen, should you go back to it? In this case, the time period is so short that it sounds like it may just be incidental that the pembrolizumab was stopped and this increased.

The only thing that I might say is that prior to the immunotherapy era, we generally didn’t go back to platinum-based chemotherapy after a patient had once received platinum-based chemotherapy and I would be cautious of doing that. I think that this post-chemoimmunotherapy situation is a particularly problematic one where we don’t have good options.

So single-agent gemcitabine seems to me to be a reasonable treatment option, but, also, to recognize that in an 86-year-old who now has progressed on this regimen, that this is a difficult clinical scenario.

DR LOVE: So Hoss, Dr Picton had another question that I thought was very interesting, which is, what about afatinib? So this patient does not have an EGFR mutation, but I think afatinib is actually approved in this situation.

DR BORGHAEI: I will not use afatinib in a situation like this. I think either a clinical trial of some sort if the patient is in good shape, or single-agent gemcitabine is what I would go with. But I would not use afatinib.

DR LOVE: Govindan, any situations where you do use afatinib? There was some data that came out trying to link up the benefit to some type of HER2 abnormality. Do you ever use afatinib in wild-type disease like this?

DR GOVINDAN: Not really. No. Only in the context of EGFR mutation.

I was going to say that I agree with Eddie’s point, that I would advise against the use of carboplatin second time around. And also, the combination of gemcitabine and carboplatin, even in a fit 45-, 50-year-old can cause significant thrombocytopenia. And I think in an 86-year-old gentleman, I would say we ought to be cautious.

The one drug, which I think has been overlooked a lot and is excellent in this setting, is vinorelbine. It got a bad rap because of the one randomized study that compared docetaxel versus dealer’s choice, ifosfamide or vinorelbine that showed low response rate vinorelbine. But vinorelbine as a single agent is an excellent drug in non-small cell lung cancer, particularly in the squamous cell type. So, I tend to use vinorelbine, single agent, quite a bit.

Selection of second-line therapy for patients with metastatic nonsquamous NSCLC; benefits and risks with the addition of ramucirumab to docetaxel

DR LOVE: So I want to bring in more about the issue of second-line therapy, and Julie, this was an interesting set here that I thought. So we asked about second-line therapy in a patient with metastatic non-small squamous. And we asked about a patient who got first-line treatment with carbo/pem and bev and then maintenance pembro/bev, who had a low TPS score. And you can see, everybody is using docetaxel. Most people are adding ramucirumab. But I thought it was also interesting to see what people would od in patients who got single-agent pembrolizumab because of a high TPS score. And it's interesting, there are a couple of faculty who want to add in bev.

So Julie, can you talk about these 2 situations in terms of second-line therapy — you’re 1 of the people who uses ramucirumab — can you talk a little bit about your rationale for that?

DR BRAHMER: Sure. So after first-line chemotherapy and immunotherapy, then we would typically recommend second-line chemotherapy. And that second-line chemotherapy hosed is a docetaxel-based regimen.

Adding ramucirumab will increase the response rate and does increase PFS and OS, but it does come at a cost of increased side effects. And certainly, in a lot of our patients, we’re even worried whether or not they can tolerate docetaxel. So I think that really comes into play when we’re sitting in front of a patient and trying to decide, do we use the doublet of docetaxel/ramucirumab or a single-agent docetaxel?

Honestly, in someone who is 65, they most likely would be able to tolerate the combination, but it's truly a shared decision making. Some patients, where I’m questionable, I will start with the single-agent docetaxel and add in ramucirumab in the second cycle if they're tolerating docetaxel as a single agent.

Now, in those patients who start off with single-agent IO, then once they progress, then the next thing to do is to go into first-line chemotherapy, platinum combination therapy. Adding bevacizumab again will increase your response rate, but it will also increase toxicity. And so, you have to weight that in a particular patient situation. But I’d feel completely comfortable doing a platinum combination in this patient’s situation.

DR LOVE: Hoss, going back to the situation where the patient’s going to get docetaxel plus/minus ramucirumab, as Julie was alluding to, a lot the concerns that we hear relate to the docetaxel. And for general medical oncologists, they’re used to using ramucirumab with paclitaxel in gastric cancer; it's standard second-line therapy, much better tolerated.

Any thoughts about whether or not you can substitute another taxane or you have to stick with docetaxel? And any variations on the dosing with docetaxel that you utilize to try to decrease side effects?

DR BORGHAEI: So let me answer the last question first. So we do use a lot of weekly docetaxel, particularly for our older patient population. And in that setting, if I do it on a weekly basis, I add the ramucirumab to the first week, and then continue with small, subsequent dosing of docetaxel on days 8 and 15. So from time to time we’ve used kind of a strategy.

As far as the combination with pacritinib is concerned, again going back to what Eddie was referring to, we have a study. The results of the study suggest that if you do this particular combination you get the results that you see on the study. And, therefore, we try to stick to what was investigated. Now, I think the study was done at the time with docetaxel and ram, where a lot of patients were receiving paclitaxel as front-line therapy, therefore, it made sense to use docetaxel at the time of recurrence, which is what we have here. Do I think biologically that would be enough of a difference between paclitaxel and docetaxel to make the combination with ram less effective with paclitaxel? No. I think it will be a reasonable option. However, if you have a trial that already shows you that docetaxel can work well, give you the response rate that you have, then I see little reason, in terms of wanting to switch to paclitaxel.

Perspective on the use of bevacizumab or ramucirumab in combination with an immune checkpoint inhibitor and chemotherapy for metastatic NSCLC

DR LOVE: Govindan, I’m kind of curious, you see two of the faculty here who, in the second situation, would bring in bev. And it brings, in general, the issue of bev, an issue that was raised by a couple of the docs is the fact that bev, from their point of view, provides a lot of benefit to patients with effusions, particularly persistent pleural effusions. And so, the question that we heard was, if you have a patient like that, is there a way to integrate bev into first-line therapy with a checkpoint inhibitor, with chemotherapy?

And is that something you would ever consider?

DR GOVINDAN: Not for squamous, Neil. I think the concerns are real with bleeding, as was demonstrated in the earlier study years ago.

In the nonsquamous, I think, legitimately, one can consider that. But we have prospective data to suggest that the combination of the 4-drug regimen was better than the control arm of chemotherapy with bevacizumab alone. But the challenge is with 4 drugs, the toxicities get to be a little prohibitive. So I do not embrace that regimen.

Regarding the comment about the pleural effusion, that's an interesting one. In fact, the VEGF used to be called VPF, vascular permeability factor, that I think there is some suggestion that this could decrease the effusion. And, in fact, the brain tumor doctors use it all the time to decrease the edema with single-agent bevacizumab. But I’ve not been so persuaded to do it. Today, we can do PleurX^TM and take care of the effusions reasonably well. But if that’s a reason to use the 4-drug regimen, which we have some prospective data for, and the patient is a candidate for that, I would have no objection. But I don’t —

DR LOVE: So you’re talking about the IMpower regimen —

DR GOVINDAN: Correct.

DR LOVE: — that utilizes atezolizumab, which is approved — right. But getting back, also to this issue, Eddie, of using antiangiogenic in squamous and non-squamous, another question we got is, why is that you can use ramucirumab in squamous cell, but you can’t use bev? And I know you have the clinical data that supports that. But I’m just kind of curious, Eddie, whether you have any hypotheses, whether you think there actually is a difference or maybe it's just the way the trials were done? Any thoughts about this issue of risk of bleeding with ramucirumab in squamous cell, for example?

DR GARON: So I actually think there is a difference. If you look at those early studies of bevacizumab in squamous cell carcinoma, it was a very significant number of patients with hemoptysis and that is of concern. Now, there is 1 question whether, with learning over time, people knew, on the ramucirumab study, to keep patients off. But at least if you assume that the selection was similar, they clearly are different.

I think in terms of mechanism, it's very difficult. People have argued that since really the ligand is the target for bevacizumab, but in ramucirumab it's the VEGF receptor, maybe that could have something to do with it. I think in the end, we don’t know. And whether or not the exclusion of central lesions in the docetaxel/ramucirumab study is responsible for this, I think is hard to know.

But the one thing I think we can safely say is, that the results for pulmonary hemorrhage that were seen between studies of bevacizumab and ramucirumab really were different. It was a Phase II study of bevacizumab where they were excluded. And again, it is possible that it really is just the issue that in the ramucirumab study that they were not including patients with central tumors. But the difference between the studies is real.

DR LOVE: So Govindan, the other thing that's interesting to me about that IMpower study is it does combine an IO and an antiangiogenic, so in this case atezolizumab and bevacizumab. That combination just took over as first-line setting of hepatocellular cancer. And we see other situations where antiangiogenics and IO have been very effective, particularly TKIs. So, for example, lenvatinib/pembrolizumab, for example in endometrial cancer.

Any comment, Govindan, in general, about the synergy between antiangiogenics and checkpoint inhibitors and whether that's a strategy that you think can be further exploited?

DR GOVINDAN: That's a great point, Neil. I think 1 interesting hypothesis is that the VEGF inhibition needs to decrease in myeloid-derived suppressor cells and may result the tumor microenvironment in a more favorable way. So, there is some interesting science behind this and there have been studies conducted, ongoing and completed, that patients who progressed on a checkpoint inhibitor were then given a combination of a checkpoint inhibitor with a multikinase inhibitor. And, anecdotally, we have seen some dramatic responses in a few patients in lung cancer with that combination, having progressed on a checkpoint inhibitor before. But, after seeing data from a variety of cancers, it seems to be quite dramatic in renal cell and hepatocellular cancer. It's unclear whether this strategy will pan out long-term in non-small cell lung cancer.

DR LOVE: So John Heymach explained to us why EGFR mutated lung cancer biologically, Eddie, in his eyes, is similar to renal cell, which I never heard before, somehow through the HIF mechanism. But I am just curious, Eddie, what your thoughts are about using further immunotherapy, as Govindan was just citing, for example, a situation like this, a patient who, after having a response, progresses on immunotherapy. Any strategies for follow up? Govindan, was talking about adding an antiangiogenic. Obviously, people think about ipi/nivo, although I don’t know there’s much data to support that.

Any other immune strategies you can think about are being tested beyond first line, Eddie?

DR GARON: There clearly are many such strategies being tested. I think, in terms of outside of a trial, I have not been using immune strategies. I think that 1 that is questioned, and I think nobody knows, is whether, for instance in a patient who would have received front-line pembrolizumab, should you continue an immune checkpoint inhibitor when they get the chemotherapy?

I don’t know of data on ipilimumab and nivolumab in a salvage setting. I know that we did present data on durvalumab and tremelimumab in patients who had progressed on a PD-1 or PD-L1 inhibitor and the response rate in that setting was 5%. So, that is a strategy that I tend to be somewhat skeptical of until I see data that would show that it really induces a significant response rate.

DR LOVE: Julie, your former colleague, Chuck Drake, who then moved to Columbia, used to tell me about these other checkpoints, TIM, LAG

Anything there, in terms of, again either synergy upfront or used in the relapsed setting? Any trials? Any trial data that has you interested?

DR BRAHMER: I think, so far, the TIM-3 data, at least in lung cancer in the metastatic setting, has had some small numbers of response, but relatively low. And the same for LAG-3, though each one of us probably has a patient where that combination seems to have worked, but I think trying to figure out why.

I think looking at some of the bispecific combinations in the near future may be helpful, taking a checkpoint blockade and then bringing the T-cell to the tumor. There are various different newer agents coming out that are quite interesting. But, again, most of these are in Phase I studies.

bispecific is basically having a PD-1 or PD-L1 on 1 side of the antibody and another on the other side, it's either CTLA4 or 1 of those other checkpoints.

DR LOVE: Wow, that's interesting.

Case (Dr Martins): A man in his early 70s and previous smoker with metastatic nonsquamous NSCLC develops diarrhea with bleeding after receiving first-line pembrolizumab

DR LOVE: So I want to talk a little bit moving to — about toxicity associated with immunotherapy. And we had a couple of cases of colitis, just to get started, that I want your take and we can move into other things, but both of these cases I think are interesting.

This is a patient of Dr Joseph Martins from Tyler, Texas. He had a patient who did well on pembrolizumab, had a partial response.

Was doing well. And came in, interestingly, Hoss, with mainly bleeding. I mean some diarrhea, but much more bleeding. He told me that he would not have really had to work the patient up or send for colonoscopy that, just based on the diarrhea, but it was more bleeding that he was concerned about. So, he was worked up and the pembrolizumab was discontinued and the patient did well.

So here’s Dr Martins and his questions.

DR MARTINS: He was not hospitalized for it. He had severe diarrhea, not severe, severe. Again, he didn't need IV fluids or anything. His main symptom honestly was bleeding. If it had just been diarrhea, I probably wouldn't have even sent him for colonoscopy. But because he was having bright red blood per rectum, I decided to send them to GI.

DR MARTINS: The diarrhea lasted maybe two to three weeks. It was bad enough that he also did not want to go back on immune therapy. He didn't want to risk having that come back again. So, he felt bad. He managed avoid no hospitalization, no IV fluids. Little bit of bright red blood, but he didn't feel good.

DR MARTINS: So the big question I have is, say he had a great response to therapy, and had you know, let's call it mild to moderate inflammatory bowel disease, would you re challenge them? One of my fears is that you retreat them and they have even worse toxicity than the time before. So when do you re challenge and when we do re challenge, how often do you, are you burned even more so?

I think that's an important question for us practicing doctors to know. Because, unlike lung toxicity, I'm not scared of the diarrhea, but I certainly don't want somebody to end up with a worse problem than before.

I’m just kind of curious what you’ve seen, Hoss, in terms of GI toxicity, particularly colitis? And have you seen bright red blood, for example?

DR BORGHAEI: Very rarely have I seen bright red blood, but the original description of a colitis is diarrhea with bleeding. So I think that it's not surprising to get a few patients that actually do have this sort of a presentation. Although, I don’t think I’ve seen bleeding to the extent that was described in this particular case, but I think it's possible.

So most of the cases that I have seen, in terms of GI toxicity, involved massive diarrhea requiring hospitalization, IV hydration. There’ve been a few patients that we’ve been able to manage as an outpatient. I’ve had a case of a very severe gastritis that we ended up attributing to the use of immunotherapy, biopsy-proven. So those are the kind of cases that I have seen when it comes to GI toxicity and they're usually manageable with the high-dose steroids and discontinuation of the drug.

DR LOVE: And you can see, Govindan, what happened with this patient.

Screening and clinical care of patients with lung cancer during the COVID-19 pandemic

DR LOVE: But of the things I thought was interesting was when the patient got bronchitis, the doc wrote in that they ruled out COVID. And I’m curious, Govindan, right now, how workups of patients with pneumonitis or pulmonary problems is being affected by the COVID crisis? Are there situations where you’re not sure of what to do?

DR GOVINDAN: So absolutely. We have a clearly defined policy, I’m sure the others have too — patients, if they are suspected to have COVID related, based on the radiographic appearances or clinical symptoms, they’re put in outpatient clinic. We have a special area for them to be screened. They are sent home if they are not sick. And then we do 2 tests, nasopharyngeal swabs, 24 hours apart, before we resume any medical therapy. If they are sick, there are designated COVID units where they get admitted. And these are negative-pressure rooms and that's where they're taken care of. But our rule is 2 nasal tests, 24 hours apart, before we can resume therapy.

We don’t screen routinely everybody before chemotherapy. The point is that you can be negative today truly, and then you could chemo, and then develop — pick up COVID infections. So I’m not sure the negative tests would be that reassuring. We ask people to take the precautions that we all take these days. And that's how we approach that.

DR LOVE: Julie, one of these docs was telling me that they had a patient who felt perfectly well, but when they went for screening CAT scan, the radiologist read it as suggestive of COVID. The patient completely asymptomatic. The patient is on a checkpoint inhibitor, so they didn’t know — they were thinking it was maybe pneumonitis. Is there a difference in imaging — I should think there should be a big difference, but I don’t know, maybe not. Is there a difference in imaging COVID versus IO toxicity?

DR BRAHMER: According to our radiologists, there are some slight differences, but it I think it really, truly depends on the situation and there can be some similar changes, ground glass opacities that are scattered throughout. You can see in both. So I think for a lot of our patients right now, if they come in, have changes on CT scan and otherwise asymptomatic, we are screening for COVID.

DR LOVE: Eddie, we’ve been talking about the challenges of diagnosing patients who were on IOs with lung cancer and had pulmonary problems because you have the issue of potentially COPD, cardiovascular disease. You have the tumor. And now we’re adding in another theoretical consideration, which is COVID.

Can you talk a little bit about how you approach patients on IOs who have pulmonary problems?

DR GARON: Yes, I think this has always been a concern — lung cancer patients are different, particularly with respect to their pulmonary issues. And even on a clinical level, we’ve had —I’ve had issues where the front desk didn’t want to let a patient in to the clinic because they were coughing severely. And, of course, this is not the only patient that I’ve ever seen who has lung cancer who had a significant cough. And, fortunately, we were able to work with everyone and give people a level of comfort, to bring the patient in and be seen and their problem clearly was lung cancer, as would be expected.

I think that it is always difficult to know what’s going on from immunologic – excuse me — from radiologic perspective. And I think that there are some changes that will be a determinant. I think the 1 thing that I would say about your question to Julie, about the COVID appearances on scans is, we’re still learning a lot about COVID-19. And we know there are all those asymptomatic carriers and patients who aren’t expressing — really showing any symptoms, we don’t really know what their CT scans look like. And so, I think it's hard to tell a radiologist that would be seeing a patient, saying this looks like COVID-19 to me, my bias would be to test that patient and see. I think it's something we don’t have a great handle on.

DR BRAHMER: Yeah, before COVID, everything was called pneumonitis and now everything’s being called COVID. But I think, honestly, in someone on IO you don’t know and so you do have — we do routinely, if someone has changes, or symptoms that we really think is probably pneumonitis, we’re screening for COVID because of the possibility.

DR LOVE: It also brings up another question I heard form a couple of these docs, which is, who do you test? Should you test every patient who’s about to be given chemo or systemic therapy? What are you doing right now at Hopkins, Julie?

DR BRAHMER: Right now, for systemic therapy we are not testing. We’re testing based on symptoms or exposure. But I can tell you, our Radiation Oncology department is considering testing all patients that are about ready to start on radiation therapy. And, again, what happens in the middle of the treatment, do you test them every day? Do you test them on a weekly basis? I think we just don’t know how best to handle these patients. But from a testing standpoint, in that situation you’re really more concerned about that patient exposing others, including other healthcare workers as well.

DR LOVE: Govindan?

DR GOVINDAN: It's an interesting area because it is changing. It is going to change in the coming months even more. It also varies from different parts of this, from 1 area, another part of the country. We have tested 998 patients, close to 1000 patients at WashU, with cancer, for 1 reason or the other. We have tested them for COVID. And 63 turned out to be positive and only 11 so far, unfortunately, died. But again, that's 11 out of 63 is a big number, sadly.

All the radiation oncologists brought up the same issue. When I asked them, are we going to test every day, then they backed off. So I think people have to be cognizant of their practical challenges.

Case (Dr Picton): A woman in her early 50s with metastatic squamous NSCLC with a BRAF tumor mutation receives carboplatin/pemetrexed/pembrolizumab followed by maintenance pemetrexed/pembrolizumab and develops immune-related colitis

DR LOVE: So I want to show you another case.

I think this is a very striking case. I was actually shocked when I heard this case, so I’m kind of curious what you think about it. So actually the case was sent in with the idea of, as I mentioned, a common question we’re hearing is how long to keep therapy going? And this is a patient who actually presented with very extensive adenocarcinoma of the lung, pericardial effusion, brain mets.

Admitted. Got a pericardial window. TPS of 10%. Gets carbo/pem and pembro, also gets Gamma Knife^® to the brain. And gets restaging. Is having a response. And currently, the patient’s doing well. And I think on maintenance therapy

But when I looked back at the case, I see that she had colitis, it's a young patient, 52-years old. So Dr Picton says, what happened was this lady started to get diarrhea and instead of calling us and coming into us, she went to her local small hospital in North Carolina and was admitted; was there for a week. Is on pembrolizumab.

And got discharged from the hospital on antibiotics and antifungals. And she saw the patient in clinic, and promptly started on corticosteroids. And she resolved slowly over 2 months. And I said, well, what happened there? And she said, well, we have hospitalists who take care of our patients when we get admitted and this hospital used hospitalists, and nobody thought about the fact that this could be related to pembrolizumab or to the IO.

Here’s Dr Picton.

DR PICTON: She kept having diarrhea. She went to the hospital, she was hospitalized for about a week. I did not know anything about it because it was an outside hospital. She came to see me and our gastroenterologists in my practice and he calls me, "Did you know that your patient was in the hospital with diarrhea.”

It's not until she comes back to the clinic that I say, "Hey, this is related to your immunotherapy. Antibiotics are not going to...” She was getting metronidazole and ciprofloxacin. That's not going to cut it. So I added the prednisone there, three weeks later and that's when she started getting better.

DR LOVE: And I was just curious, I’ll start with you, Eddie, about this issue — in the past people have told us about ER, people going to the ER and then not realizing that they’re on immunotherapy and the implications. But she brought up this whole issue of the hospitalists, which I wasn’t that tuned into. And actually, I think I was talking to your colleague, Ben Levy, Julie, and he was saying you guys use a lot of hospitalists. It’s not just small, rural hospitals using hospitalists.

So just curious, Eddie, any comment about this issue of people outside of oncology who end up managing these patients?

DR GARON: So first of all, hospitalists are obviously a significant part of the medical care infrastructure and particularly in this era of COVID-19, maybe an important one. I have limited my hospital rounding to some extent, feeling that if it's not a month where I’m assigned to the hospital, it's probably a good idea to limit the number of exposures.

It's interesting, when all of these drugs were research drugs and our patients came into the ER or outside hospitals, there was an immediate assumption that any problem they had was related to the research drug. And sometimes it's a little interesting, that as soon as a drug becomes approved, now it becomes part of their standard of care. And I think sometimes it does lower the threshold for concern as opposed to when it was a research drug.

I think that there are many practitioners that are familiar. But there are still many practitioners who don’t have a great, really just conceptual understanding of what the different toxicity profiles would be from cytotoxic chemotherapy to targeted therapies to immunotherapies. And I think that the best solution to that is clearly going to be communication. We don’t need every physician and nurse across the country to be experts in immuno-oncology.

On the other hand, we do want them to have at least some understanding of conceptually, what the different side effect profiles might be or comfort asking us if they don’t know.

DR LOVE: And another issue, and Julie, I loved your paper in the JCO, 70-page paper on toxicity. There had been a bunch of other attempts to really do comprehensive reviews of toxicity. But 1 of the things that I think was most helpful about that, and other similar efforts, was the grade the — standardize the approach based on grade of toxicity. So, moving away from what the toxicity is to more what the grade is and then using that as a benchmark of what you do.

Can you kind of go through that and how that has evolved?

DR BRAHMER: It really evolved out of clinical trials using the standard CTCAE criteria, and using that as a basis of what to do. And really, that's helped us figure out at least for those toxicities that we don’t think are permanent such as non-endocrine toxicities, using the grades to help manage. Not all patients need to go on high dose steroids. And not all patients do have to stop immunotherapy. And it really depends on the grade and the particular toxicity.

And so, that's evolved all the way from the original Phase I trials, now into Phase III trials. And there are even trials looking at specifically how best to treat these toxicities, all the way from using vedolizumab upfront rather than infliximab in those patients being treated with corticosteroids for colitis. So there’s a trial on that. And then there’s also a trial looking at steroid-refractory pneumonitis as well.

Management of metastatic NSCLC with targetable EGFR mutations

DR LOVE: So I want to move on and talk a little bit about managing patients with EGFR tumor mutations. One thing, Hoss, that I thought was interesting, is we asked the audience — or we asked the faculty about the issue of combining an antiangiogenic with an EGFR TKI. And you reviewed that in the lecture that you did as part of this program. When we asked the faculty whether or not they’ve actually done that strategy, most of them have either done it or would do it.

But any thoughts, Hoss, about the basis for this strategy of combining an antiangiogenic and whether or not you think it's appropriate. You say that you’ve already done that.

DR BORGHAEI: I’ve done it in specific cases, in case that I remember using an antiangiogenic agent. First of all, they were mostly with the first- or second-generation TKIs and they were patients who did not have as good of a response to the initial TKI. And, based on some of the studies that were published previously, having an antiangiogenic agent, I thought we might be able to rescue that patient population and get a better response out of the TKI.

In the case of patients who are on osimertinib, I have to say recently I have not had the occasion of adding an antiangiogenic agent. But we all have seen several studies now, again with first-generation drugs in combination with bevacizumab, suggesting better clinical efficacy.

The data with osimertinib is still cooking. There are studies again looking at that. But it does appear there is at least a little bit of an additive effect, I wouldn’t say it's necessarily synergistic, but definitely some additive effect in terms of improvement in the clinical efficacy.

Clearly, the downside of using an antiangiogenic agent with a TKI is that now you’re taking a drug, in this case let’s say, osimertinib. That's very well tolerated, a once-a-day pill. Nobody has to come to the office for anything. And now you’re adding a drug that's, at least at this point for the most part and IV agent, and now all of a sudden patients are going to have to come to the clinic, they’re hooked up to an IV. And then you’re upping the ante a little bit in terms of the toxicities, dealing with the traditional toxicities associated with the antiangiogenic agents, including hypertension, proteinuria and risk of thromboembolic events and things of that nature.

So doing it in special cases I think makes sense. But if we are to use this, I think we need more definite data to say that a combination like this is better than single-agent TKI by a lot more to justify the additional cost and toxicity and inconvenience to the patient.

I think it's a decent strategy. I think, when you think about the fact that we’re not curing patients with EGFR mutations with simple osimertinib at this point, that progression eventually happens. If there are strategies that can delay the progression, improve, again the clinical efficacy without too much toxicity, I think it's a reasonable approach.

DR LOVE: So Eddie, Hoss reviewed some of the ongoing trials that are looking at this, both looking at bevacizumab and ramucirumab. Just kind of curious what your thoughts are about it. The one situation I’ve heard people talk about that Hoss was alluding to, is the patient who has a good response from an EGFR TKI, for example, osimertinib, and then has disease progression — the idea of adding in something like bevacizumab. Is that something you would ever do or consider, Eddie?

DR GARON: It's something that I’ve done, although, in all honesty, it collaboration with other physicians who have chosen to do it. When I have used this approach, it has not been in a salvage type setting, but really as a front-line approach. And I think that you do see very robust progression-free survival. We do not have overall survival data on this approach yet. But you now have 3 large studies showing strong progression-free survival advantages.

The other thing that is of note is that these patients, at least from the recent trial looking at ramucirumab, appeared to have T790M develop at a similar rate as would be anticipated with just single-agent TKI. So, if, for instance, a patient were to start erlotinib plus a VEGF TKI, the toxicity of the erlotinib is generally more difficult than the osimertinib, but, nonetheless, that patient potentially could benefit from osimertinib at a time of progression, particularly those who have a T790M.

So I think it's the sort of thing that is a reasonable discussion to have with the patient.

DR LOVE: So that leads into one of the big questions that comes up with these patients, which is what to do when they have progression after first-line therapy. And Julie, 1 of the questions we asked is, should you get more tissue or blood for an assay? And it looks like everybody is doing that. I’m kind of curious, Julie, what do you usually do if you don’t find anything specifically targetable at that point?

DR BRAHMER: Then I move on to chemotherapy. And I really feel most comfortable with the chemotherapy plus bev plus atezo regimen.

In these patients, my gut is that these patients need a taxane and do seem to need the anti-VEGF and this is the only trial, particularly if the patient wants to use IO. This is the only trial that included patients with EGFR mutant disease.

DR LOVE: So Eddie, this introduces a question that's been heavily debated the last few years, not just in patients with EGFR tumor mutations, but other patients, ALK, etc, which is where does an IO fit in? Where does PD-L1 testing fit in with these patients? And you can see, that at least in terms of EGFR, the faculty is going to think about it, whether it's on first relapse or beyond. A lot of them will think about it right after targeted therapy. But you have Dr Horn saying, she doesn’t want to use IOs, period. Which is a little bit unusual response.

But Eddie, can you summarize what you say when you’re making rounds with your fellows about what the question is here about using checkpoint inhibitors and how you deal with it clinically?

DR GARON: I think what we have to explain to them is that the data to date would indicate that amongst EGFR mutation-positive patients, a checkpoint inhibitor is not a great therapy. Now, the question is, whether or not adding it to some other therapy can significantly improve that, and I think that's still debated.

I have used the 4-drug regimen of atezolizumab, carboplatin/paclitaxel and bevacizumab. However, I recognize the limitations of that data. The numbers were small, and it was not considered sufficient for the FDA to approve it; it was not part of the design of the study to assess that group of patients for approval.

So what I would say is, that when I talk to people, I don’t have a tremendously high hope for single-agent immunotherapy in those patients. It's something that they can try if they have exhausted other options. I have tried to incorporate it along with chemotherapy, with varying results.

DR LOVE: So Hoss, What about other targetable — ALK, BRAF — do you view them all the same, Hoss? BRAF, do you see more smokers, do you treat them differently, as it relates to checkpoint inhibitors, for example?

DR BORGHAEI: Yes, I do treat BRAFs a little bit differently than some of the other targetable mutations. So the EGFR and the ALK patient population, I pretty much treat the same. As was just suggested, I don’t think there is evidence that single-agent immunotherapy is active in that patient population. So if I were to use it, I would use it combination with chemotherapy, like it was presented here.

BRAF is a little bit different. I think there is some data to suggest that maybe they would respond better to immunotherapy approaches. I think that is a little bit of a separate, or a different patient population, as far as I’m concerned.

For the other ones, like the NTRK and all of that, I think again, we’re learning a lot more about these mutations with the drugs that are coming up. And the majority of the patients are going on, for instance, some of the other trials with NTRK, have had exposure to chemo and immunotherapy. So I think it's a little bit of a different patient population there also. But for the biggest chunk of patients, the EGFR and the ALK, the strategy that I have listed here, chemo with immunotherapy, is something that I would definitely do and not single-agent immunotherapy.

Case (Dr Deutsch): A woman in her late 40s and never smoker with metastatic adenocarcinoma of the lung with an EGFR tumor mutation develops a T790M mutation 2 years after treatment with erlotinib and bevacizumab on a clinical trial

DR LOVE: So let’s talk about another topic we also discussed with the other panel…immunotherapy toxicity and here’s a case to get started.

This is a patient who was diagnosed in 2015. Actually, went on a trial — we were talking before about this idea, antiangiogenic. So, this patient went on a trial of bev and erlotinib at that point.

Interestingly, the bev had to be discontinued because of proteinuria, but the patient was on it for 2 years, which is pretty good response. And then, remember the old days of T790 — I took the Boards last year and T790 was still on it. I just wanted to see what was on the Boards. They’re kind of behind the times. But anyhow, so I guess this was in the last wave of T790, got osimertinib and had disease control, and then finally progressed. And is currently responding to carbo/pem and pem.

In her fourth cycle of carbo/pem and pem, had to be admitted to the hospital for leg weakness and reticular leg pain. There were no new lesions in the bone, et cetera. And the diagnosis by the neurologist was transverse myelitis. And the thinking was that it was related to the pembrolizumab, which was discontinued. She had a little bit of response with corticosteroids, but did poorly and ended up dying.

Here’s Dr Deutsch.

DR DEUTSCH: So my question was, how — how common are neurologic complications from immunotherapy? I’ve seen optic neuritis. I’ve seen this transverse myelitis. I’ve seen PRES in my patient population. And my question would be, is there any way to predict for people who are going to be more likely to get immune-mediated toxicities? I mean obviously the people you treat who have underlying immune diseases are at risk, but what about the general population? And is there any way to predict for those people who are going to have toxicity? Because, unfortunately, the neurologic toxicity can be pretty profound and serious.

I’m just curious, Julie, if you comment on the spectrum of neurologic problems that you see with IOs and whether or not they respond to things like corticosteroids?

DR BRAHMER: I mean the spectrum of neurologic diseases is associated with IO is so widespread, all the way from a generalized encephalitis, where the patient is potentially even comatose and nonresponsive to just a single nerve root issue. And so, the presentation is so varied, that having a neurologist evaluate these patients with you is so important. And the steroid response is typical, but sometimes here, depending on the neurologic disease, you may have to consider plasmapheresis. You may need to consider IVIG. There’s also been some improvement in some of these patients with rituximab, if you think that it's a B-cell mediated toxicity.

And so, again, this, particularly when you’re given not so pemetrexed/carboplatin and pembro, but if you have someone who’s getting paclitaxel, carbo and pembro and coming in with peripheral neuropathy that seems to be out of proportion to what you typically think with paclitaxel then you have to think about potential immune-related toxicities, as well.

DR LOVE: Hoss, any comments on your experience with immune-related neurotoxicity?

DR BORGHAEI: I’ve seen 1 case of cerebritis, which is basically what Julie was just describing, a patient who came in and was basically comatose. So, fortunately responding to high-dose steroids very nicely, and obviously, after that, we had to discontinue immunotherapy. But these are very scary. I’ve got to tell you this patient was very sick. And for all practical purposes, looked like she might have a massive stroke. And, of course, imaging studies didn’t show anything like that. And we did LPs and a whole workup and started her on steroids, but with response. So that is the one case of neurotoxicity that sticks in my mind particularly.