Lung Cancer Update, Issue 2, 2017 (Video Program)
Lung Cancer Update, Issue 2, 2017 (Video Program)
Proceedings from video interviews with Drs Matthew Gubens and Suresh S Ramalingam on optimal strategies and emerging concepts in the management of lung cancer.
2.75 AMA PRA Category 1 Credits™
To complete the Post-test and Educational Assessment and Credit Form, click here. A statement of CME credit will be issued for scores of 80% or better. Your statement of credit will be mailed to you within 3 weeks or may be printed online.
Case: A 79-year-old man and former smoker with KRAS mutation-positive metastatic adenocarcinoma of the lung receives second-line treatment with nivolumab DR GUBENS: Yes. This is one of my patients, 79-year-old gentleman, pretty heavy smoker before, who had Stage IB cancer, had it resected, and 2 years later on surveillance he was found to have pleural metastatic disease, so a pleural nodule, some pleural fluid. We gave him carbo and pemetrexed for 2 cycles, progressed right through it. Nivo had just come on board, nivolumab. So we started nivo at that point. Fifteen doses, fantastic response, probably 80% reduction in the nodules. The fluid stayed away. But then kind of between screening — between surveillance scans, he had really rapid onset of exertional dyspnea. He called and said, “I can’t even make it to the shower.” So we brought him into clinic. His 02 in clinic was about 91%. And so, of course, we got STAT imaging that day, and very clear pneumonitis, classic, every lobe involved, a very clear pattern. So we stopped nivolumab. We started steroids, kind of had a debate about whether to admit him to the hospital. But he didn’t need supplemental oxygen. He was at 91%. He had good follow-up. So we just started steroids at about 1 mg/kg. And within a couple of weeks, he had very clear improvement in the symptoms and, over the course of about 2 or 3 months, we had a very gradual taper, saw resolution on imaging. And then it’s always that question: Do you rechallenge? Do you stay off drug? We chose to keep him off and just watch him closely, got scans every 6 to 8 weeks. Unfortunately, after about 6 months of treatment holiday he had very clear progression, newly, in the liver. At that point had a long conversation about the risks and benefits of maybe rechallenging. He hadn’t failed nivo, it was just the toxicity. Now, in the interim atezolizumab had come on board, so it kind of offered 2, at least, theoretical benefits. One was a logistical benefit, every 3 weeks instead of every 2 weeks. That’s nice for a patient who lives far away. And also, this kind of maybe more theoretical one, that maybe, just maybe, the PD-L1 inhibitors may have a titch less toxicity. I’m not solely convinced about that, but I do think it was a reasonable thing to say, “Hey, it’s worth a try, a different agent.” And we had a long conversation that the recidivism rate is real, kind of estimated maybe a 50% chance we’d see pneumonitis, but figuring that, even though that’s a risky proposition, if we’re very much on it, we hear any change in symptoms, we could certainly stop and readdress. He actually tolerated it beautifully, but right after 3 doses, after 9 weeks, he had further progression in the liver and so unfortunately had to switch over to chemo, which he did not respond well to, just as he hadn’t in the first place. And we put him on hospice and unfortunately kind of were not able to get him any more treatment. DR LOVE: How long has he been on hospice now? DR GUBENS: This is fairly recent. He’s been on about 2 months. He’s kind of had slow but steady decline in his performance status, not especially dyspneic actually but just really the fatigue/malaise and anorexia that we kind of think is just further progression. DR LOVE: And what’s his current state of mind? DR GUBENS: He’s at peace with it. Certainly he was — it’s kind of all this excitement about immunotherapy that we share with our patients. And he was the one of the one’s who had just this beautiful response. So we had a lot of hope for that second rechallenge. He wasn’t scared of the toxicity. He trusted us to be able to follow him closely and nip it in the bud if it came. But I think it was obviously disappointing to not get that second response. DR LOVE: I want to go back through a few aspects of his case, but just to follow up a little bit in terms of his current situation, I am going to guess that at some point he had some type of multiplex or NGS testing. DR GUBENS: He did. We didn’t do it for his original Stage IB disease, but, of course, on metastatic presentation we did go NGS. We did detect a KRAS mutation. So we talked a lot about that. We didn’t have any active trials in the first line. And by the time he kind of blew through first line, we had nivolumab to offer. And he lives far away from our center. We talked about clinical trials in the immunotherapy space, but just single-agent nivo was available and, I think, very appropriate for his care. Incidence, monitoring and management of immune checkpoint inhibitor-associated pneumonitis DR LOVE: Going back to your decision to try the atezolizumab, which I thought was really interesting when I saw this case, because when the anti-PD-L1 agents came out, you heard this thing about maybe there’s going to be less pulmonary toxicity. My understanding is, the anti-PD-L1 agents, like atezolizumab, gets the ligand for PD-L1, so it doesn’t get the ligand for PD-L2. Tell me if my simplified explanation is right. DR GUBENS: Exactly. DR LOVE: But yet the anti-PD-1 agents that hit the receptor, therefore, essentially suppressing PD-L1 and 2? DR GUBENS: Exactly. DR LOVE: And what was the thinking about why it might have less pulmonary toxicity? And it kind of sounds like you think it really didn’t pan out. DR GUBENS: I think you have to look at the data we’ve got in front of us. We’re never going to see a head-to-head trial, certainly nowhere in the near future. I think that numerically there’s maybe a couple of percentage points’ difference, but you always have to say across-trial comparisons are somewhat suspect. I know there was a meta-analysis, again, that amplified that 1% to 2% difference. And again, it’s just hard to say when the response rates track so well, when the other immune-related AEs track fairly well across nivo, pembro, atezo. I can’t put a lot of stock in it, but I do admit, in this case where I admit there’s a rechallenge risk and I think some of those recidivism rates look like 50% or so. I’ll take any little marginal benefit I can. But yes, that idea that only hitting the ligand, it’s just a much more selective inhibition there and kind of avoids some of the more general effects of the PD-L1/PD-L2 axis. DR LOVE: We’ll get into the issue of first-line therapy, but I am curious how you approach the issue of second-line therapy right now. Typically, that theoretically would be a patient who has a lower PD-L1 assay. So they’re not going to get it first line — or a checkpoint inhibitor first line. They get chemo. Maybe they get bev if they’re nonsquamous. We should mention, this man was an adenocarcinoma. DR GUBENS: Right. DR LOVE: But then they progress, and second-line therapy. How do you go about choosing the checkpoint inhibitor there? DR GUBENS: I have a lot of equipoise among the 3 agents we talked about, pembro, nivo, atezo. Again, I’m not convinced by any real significant efficacy or toxicity differences. So a lot of it really plays into logistics. A lot of our patients are coming from a distance. And absent any clear benefit for one or the other, three 3-week dosing sure is easier for patients. I’m a little paternalistic. The occasional patient that I’m nervous about, maybe their performance status is a bit marginal, I still kind of reach for nivo because I know someone’s going to lay eyes on them every 2 weeks. But in general for a relatively fit good PS patient, I’d typically use atezo or pembro, mostly for logistics. And pembro, of course, has the PD-L1 requirement of 1% or higher, so if they happen to be a nonexpresser or if we don’t have the data, atezo is right available out of the box. DR LOVE: So this 79-year-old man, you say, was a heavy smoker in the past. Did he have chronic lung disease? DR GUBENS: He had just a touch of emphysema, didn’t have clinical COPD but had radiographic evidence of emphysema but really an excellent 02 status and actually a good performance status, dancing and singing when I first met him. DR LOVE: Wow! Interesting. So it sounds like in this case it was straightforward that he had pneumonitis. I’m curious how often you run into this situation, specifically with lung cancer, when you’re really not sure. You’ve got a patient with prior lung disease, maybe cardiac disease. You have tumor in the lung. Do you find situations where you’re really not sure what’s going on and have to stop the checkpoint inhibitor? DR GUBENS: Most definitely. And I think that our radiologists have gotten so much savvier about trying to call these differences. I’ve seen lymphangitic disease called so much more frequently in the last 2 years than I have the 10 years prior. So I think it winds up being a discussion. And I completely acknowledge that point. This case happened to be fairly standard, or radiographic appearance was really inexpugnable. But when there’s that question of preexisting interstitial lung disease, whether that’s smoking associated or other or kind of a lymphangitic picture that’s hard to discern or the patients who have had radiation in the past where there’s kind of a radiation pneumonitis piece to it, I think it’s challenging. So really, as with a lot of these immune-related AEs, I bring in my pulmonologist. I’m talking to my ID docs. And in this case it was so straightforward, we didn’t even admit to the hospital. But I have a very low threshold to bring them in, ask my pulmonologist to do a bronchoscopy. Again, it’s not always a satisfying entity and sometimes we have to kind of look at it a bit empirically and say, “It’s risky enough. Let’s just stop the drug and put them on steroids” and kind of see how it plays out. DR LOVE: So obviously there’s a lot of individualization involved, but I’m just kind of curious, your gestalt when you start a patient on a PD-1 agent like nivolumab, in this case, in your own mind, putting it all together in terms of data and your own clinical experience, what’s your best estimate for the risk of pneumonitis, for something like this to happen? DR GUBENS: I’d say kind of the 5% proportion that you kind of see, plus or minus a couple of percent, is very consistent with what we’ve seen in our practice. And again, I think there probably is a mild correlation with previous radiation. We were very involved in one of the clinical trial groups, and they were very conservative to say, “Anyone who had had,” let’s say, “40 Gray of radiation was prohibited from seeing a PD-1 inhibitor within 6 months.” In real practice, of course, that’s not always practical. Patient may have had chemo/rads for Stage III disease and has a rapid recurrence. When I see that kind of patient, when there’s not as much clinical data, I’m very honest with the patient. I say, “We’re going to follow you very closely. There’s at least a theoretical risk of some radiation pneumonitis or maybe pneumonitis that is more likely in the setting of that recent heavy dose of radiation.” So I think it’s something to bear in mind. PACIFIC: A Phase III trial of durvalumab after chemoradiation therapy for Stage III non-small cell lung cancer (NSCLC) DR LOVE: Actually, not to get too far off topic, but I was going to ask you: There’s a press release out there about a trial with durvalumab after chemoradiation with locally advanced that supposedly — and I think — did you participate in that study? DR GUBENS: I didn’t, no. But yes, that’s exciting news. The PACIFIC trial I’m looking forward to seeing more than a press release. DR LOVE: Right. But there would be the situation you just described. DR GUBENS: Right. DR LOVE: The patient’s just gotten radiation. Do we know from that trial or any other similar trials globally kind of what happens when you give a checkpoint inhibitor right after chemoradiation? DR GUBENS: Again, I think that just kind of — the early data suggests that it probably is safe. And it’s just something you have to watch. But yes, I think that’s a really good counterpoint to say, again, the company we worked with was very conservative about that, again, before we knew. But now, seeing the PACIFIC trial, at least in press release form, hinting at no great huge safety signals, it may well be safer than we thought it was. DR LOVE: It’s going to be interesting — maybe, who knows? — that’s going to become standard of care in the next year or two. It wouldn’t be, like, a total shock, I don’t think. DR GUBENS: It wouldn’t, though I think the bar in that Stage III setting is a little different than what we’re looking at in metastatic. I think with any adjuvant trial, even in these high-risk Stage III-ish patients, I think we have to be very mindful of are we just buying some PFS that we would have bought at the metastatic recurrence, or are we actually genuinely buying more cure rate? I think it’s an interesting question that we’ve got to keep in mind as we evaluate these data as they come in. Treatment options for patients with locally advanced NSCLC and rapid disease progression DR LOVE: It also ties into a question I’ve kind of had — and I’m not really sure exactly what people are doing in their practice — which is, you have a patient who either gets adjuvant therapy or chemoradiation for locally advanced disease and then has a quick relapse. Like maybe, say, within 6 months they’ve had adjuvant chemo or chemoradiation. When you treat them at that point, do you follow a first-line algorithm or a second line? DR GUBENS: It depends on how strenuously I treated them in the chemoradiation setting. And I tend not to use, say — let’s say in a nonsquamous patient, where pemetrexed really is my favored metastatic regimen. For a patient who is getting chemo/rads for locally advanced disease, I often actually don’t use a pemetrexed-based regimen. So I tend to use carboplatin/etoposide or carboplatin/paclitaxel. So sometimes I’m more tempted to say, “Let’s give you the best metastatic chemo regimen, has an opportunity for maintenance use.” A patient who kind of saw that or really blew through quickly, which suggests they’re just really going to be chemorefractory, then I’m more inclined to say, “That was your first line of chemo, and now let’s go straight to the immunotherapy approaches.” That’s very — I think it’s very case based and situational, but I think even the difference between 2 months and 6 months makes a difference in my mind, in that setting. Pseudoprogression in patients receiving immune checkpoint inhibitors DR LOVE: Now, this patient had a great response. Incidentally, how quickly was it before clinically you felt he was responding? DR GUBENS: Scan number 1. And I usually get my scan after 8 or 9 weeks, kind of depending on the cycle length. But he was one of these folks on the spaghetti curves who had a response straightaway and sustained that through the course of his nivolumab. DR LOVE: Was he symptomatic? And did his symptoms improve? DR GUBENS: I think there was an element of dyspnea. And there was also kind of — again, kind of fatigue, malaise, anorexia that improved beautifully, really even in the first month of treatment. DR LOVE: And is that typically what you see in terms of time to response with checkpoint inhibitors? DR GUBENS: I typically see the responses by staging scan number 1 or 2. Very rarely do I see this kind of — I’ve only seen pseudoprogression once in my entire practice. For all the words spilled on paper about pseudoprogression in lung cancer, I have seen it once. It’s kind of a unique patient who had kind of really small pulmonary nodules bilaterally, so I think he was kind of a special case. But really for the most part, I think in my practice I’m seeing the response straightaway. DR LOVE: You brought up the issue of pseudoprogression, and that comes up in every cancer where checkpoint inhibitors are used. The rule of thumb that, kind of, I think we first heard from the melanoma people a few years ago was that if the patient’s feeling the same or better and it’s an imaging thing, just keep going. Is that the way you do it? DR GUBENS: I think that’s fair. Again, obviously, the devil’s in the details. But if patients have a very modest increase in the size of their tumors and they’re feeling great or even improved, I’ll leave them on. But I think the problem is, we’re so hopeful about these agents we have to remember, at least as single agents in the second line, our response rate’s 20%. So I think some of it winds up being hopeful thinking. So if a patient has any clinical decline or really kind of multiple new nodules that aren’t lymph nodes that could be reactive, I really say, “We should be moving on to the next thing or thinking about a trial” at that point. DR LOVE: So this man, he actually had a clinical CR? DR GUBENS: No. He had a PR. CCR is very vanishingly rare. I’ve had one or two in my whole practice. But he probably had about a 60% to 70% reduction. DR LOVE: So he had a very good response. He got 15 treatments. It sounds like he was kind of cruising out into the good part of the curve that we like to see people on. Again, from your point of view globally in a situation like this, second-line therapy, for example, what’s your global estimate of the risk that the patient’s going to have a really useful, prolonged response or prolonged stable disease or without having to change therapy? DR GUBENS: Yes. I’d say our experience has been kind of what’s been seen in the trials, about 20% authentic response rate with another, let’s say, 10% or so of folks who maybe don’t meet RECIST criteria for a partial response but who have meaningful stability in the face of what had been growing tumor. I’ve definitely had a subset of those. DR LOVE: One other, kind of, detail this case I was curious about is, I noticed initially when he was treated for metastatic disease, again adenocarcinoma, he was given carbo/pem. So just a real simple question: How about bev? DR GUBENS: I actually still use bevacizumab in a significant proportion of my patients. I didn’t get into it in the case presentation, but he had a fairly long history of hypertension requiring a couple of medications to manage, so kind of this phenotype of patient where I’m a little more concerned about the downsides of bev versus the upsides. For younger, fitter patients who aren’t vasculopaths, I do tend to add bev in that setting. I still believe there’s a subset of patients who get real benefit. But this is that patient who, I think, had enough comorbidities that I figured that the marginal benefit was going to be lower. Immune checkpoint inhibitor-associated side effects DR LOVE: So again getting back to your global take on checkpoint inhibitors in lung cancer, you threw out the figure of 5% for pneumonitis. What other complications do you see with these agents that are the most common? I’m specifically curious about colitis and endocrine problems. DR GUBENS: Mm-hmm. Certainly thyroid function is very common. I think we probably have low double digits, maybe 10% to 15% in our experience, which I think replicates what we see in the trials. Unfortunately, it’s something that’s common enough and followable enough that we can get thyroid function tests on a regular basis. It’s baked into our protocols. And so it’s something that we’re usually able to identify quickly and handle and usually not have to suspend treatment for. The rarer endocrinopathies, let’s say, diabetes, we see that a handful of times, definitely 1% or less. Some of the other endocrinopathies we see. And I think that we’ve just become very, very quick when we see any derangement to be sharing these patients with our endocrinologists and kind of trying our best to keep patients on protocol or on treatment if we can manage symptoms, like with hyper/hypothyroidism, but otherwise to really manage aggressively. I’d say the other things we’re seeing, maybe even more than I expected, rash, sometimes just bothersome, but sometimes actually severe enough that we are sending folks for dermatology referral much more quickly. Fortunately, we have kind of forged really good relationships with our dermatologists in the oncology setting for side effects of other targeted therapies and treatment. So we already have a good relationship. But a lot of folks we are having to refer over for management of rash that can be problematic. I’d say we also — and this is more of a nuisance issue, but I’ve seen it more than I think I’ve heard it reported. I have so many patients who have pruritus, and sometimes just annoying and we can manage it with diphenhydramine or what have you, but sometimes severe enough, affecting quality of life and sleep enough that we entertain, at least, the idea of treatment hold. And then the other one that I think has been interesting, a lot of migratory arthralgias and such, sometimes very classic, even kind of swollen joints and discomfort, but often more subjective, saying, “Today my left hip hurts. Tomorrow my hands hurt,” and we just kind of follow those over time. And I’ve made a couple of referrals to Rheumatology to help out. Sometimes what I wind up doing or having to do is brief holds or brief pulses of steroid to kind of help assess that, help address that. Use of immune checkpoint inhibitor therapy for patients with preexisting autoimmune disorders DR LOVE: So I’m curious. You would think — I don't know, I thought maybe it was going to be more of a problem, but I don’t hear too many cases. Have you had any patients with prior autoimmune disorders who you’ve then wanted to give a checkpoint inhibitor to, and what do you do? DR GUBENS: Yes. That’s a tough question. And again, it’s an area where I hope we learn more in the real-world practice of these drugs, because all those patients, appropriately, were excluded from the trials. Often I think you can kind of stratify those patients a couple of ways. Folks who have kind of long histories but maybe burnt-out autoimmune diseases or things that are in very good control over years I’m more comfortable with — psoriasis that doesn’t require systemic treatment. We’ve treated several patients in that setting and haven’t seen recurrence. What makes me nervous are folks — I didn’t get back to your question about colitis. I don’t see colitis de novo more — I’ve seen it once or twice. But I admit to a lot of nervousness with folks who have baseline ulcerative colitis or some such. And sometimes, the risk-benefit ratio — they’re out of other options. We try it with a lot of caution, but those patients are the ones that make me quite nervous. Case: A 51-year-old woman and previous smoker with metastatic adenocarcinoma of the lung and no actionable mutations responds to pembrolizumab and carboplatin/pemetrexed on the Phase II KEYNOTE-021 trial DR LOVE: So let’s go on and hear about your second case, I think, really gets to the question of the day in lung cancer — let’s hear about your 51-year-old lady. DR GUBENS: Yes. So we have a 51-year-old woman who had metastatic lung adenocarcinoma. She had a big right hilar mass that was actually the source of her symptoms but on staging had very clear bilateral lung nodules, not other distant sites of disease, though. She’d had a mild smoking history, 10-pack years a couple of decades ago. She actually wound up on one of our trials. She wound up getting on-protocol carboplatin/pemetrexed and pembrolizumab, the regimen that’s now, of course, approved that we’ll talk about in a second. But she had an excellent response. She, like our patient before, had a response by scan number 1 after, I think it was, 9 weeks, had 80% reduction in the size of her disease. And so she continued on carbo/pem and pembrolizumab on protocol and just had a very good sustained benefit scan after scan. And per protocol — it was baked into the protocol that pembrolizumab comes off after 2 years. So she hit her 2-year anniversary. That came off. We actually had a long conversation about whether to continue pemetrexed, which obviously is a maintenance drug, as well. For various reasons, we decided to give her a break from that as well. And now she’s 4 months out, off all treatment and with a sustained response, still early going, but it kind of presents some of the interesting questions about how to look at these chemo/immunotherapy combinations. DR LOVE: So she was on the randomized study, correct? This is the randomized Phase II that she was on? DR GUBENS: She was actually on an earlier part of the study that wasn’t randomized. Approach to first-line therapy for patients with newly diagnosed, mildly symptomatic adenocarcinoma of the lung and a high PD-L1 tumor proportion score (TPS) DR LOVE: Oh. That's interesting. So at that point she was on a trial using the combination that eventually, as you said, in a randomized Phase II led to the FDA approval. I’m kind of curious. Now that we have the trial done, if you saw her again today, how would you manage her? And what was her TPS level? DR GUBENS: So that was the part of the study that was so early on, we actually don’t know. It wasn’t performed as standard of care. It was collected on trial but not reported back to us. So, of course, I’m burning with curiosity. When I present the case at case conference, it’s always the first question. And one day we’ll probably kind of collect that, but that’s what’s — DR LOVE: So again, maybe just think about her — you can remind me, like, what her clinical status was, whether she was symptomatic and maybe I could ask you if she was under 50%/over 50%, how do you think you’d approach her today? DR GUBENS: Excellent question. So she was relatively symptomatic, good performance status, kind of 1, but verging on 0. Really her main complaint, the presenting complaint that got her diagnosed was a cough that was presumably from this right hilar mass and obstruction associated with that, but very functional and otherwise not too symptomatic. So I think it’s an excellent question. It’s been a long conversation with any patients starting therapy these days. I think right now if she had a PD-L1 or TPS of over 50%, I still go with the KEYNOTE-024 data and say, “Pembrolizumab as a single agent has an excellent response rate, excellent chance of durability and, of course, very modest, manageable toxicity.” So I tend to still recommend pembro as a single agent. Therapeutic options for patients with a PD-L1 TPS of 1% to 49% DR GUBENS: I think the more interesting area is that 49% and below. And I have a very long conversation for folks like this to say, “Carbo/pem, plus or minus bev, has been the standard for quite some time,” that there’s this randomized Phase II trial that suggests a better response rate, a better PFS. But I also express my concern to say that, “What we don’t know” — and I think what all of us are really waiting for — “is there an overall survival benefit?” And I pull out the history books. And I say, “Look. Any trial of 3 chemo drugs always beat a doublet in terms of PFS and response rate but never managed to marshal an OS response.” Now, I totally admit that immunotherapy is a completely different mechanism. But before I stack my treatments and kind of use all my agents up front and not know what I’m going to next, I want to be a little circumspect. So I admit that this deficiency in the data of not knowing if there’s a survival benefit compared to just PFS and response rate — and I talk it out with my patients. And I admit, some of my patients kind of buy the argument that we’re hoping that by sequencing, if I can get some period of time with chemo alone, maybe with bevacizumab, which I can’t offer with the combination here, maybe in the next line of therapy I can offer immunotherapy plus X, immunotherapy plus combos, immunotherapy plus vaccine, options that may do a better job of immunotherapy use in the second line. So my long-winded way of saying it’s a discussion. And I have to admit that I think changing practice on the basis of a 123-patient study gives me a little pause. DR LOVE: You’re expressing the viewpoint that our survey work suggests, around two thirds of investigators. DR GUBENS: Oh. Okay. DR LOVE: There’s about a third who are pretty pro the triplet. But specifically, I think, in a situation where there’s not an urgent need for a response, more typical kind of situation, kind of, that’s the way we’re seeing it. Again, what about the more symptomatic patient? You lean more towards a triplet there? DR GUBENS: I admit that’s a situation where I kind of want to throw the book at somebody and get the response I need, and I think that’s a really valid point. And that’s where I admit my resolve wilts a little bit. I say, “I want a response. And if you’re going to be the one who gets an immunotherapy or a chemo or both response, I need it now.” And so I’m willing to kind of deploy the triplet. DR LOVE: You referred to the issue of maintenance therapy when the carbo is stopped. And I believe that the trial, or at least the randomized Phase II trial, allowed you to, I guess, do what you wanted. But I’m not really sure exactly what people are doing. Theoretically, they could do what you did, which is keep both the pembro and the pem going. But I guess you could pick one, also. DR GUBENS: Yes. I admit that’s always — it’s always a nice problem to have, just when we’re using carbo and pemetrexed anyway, that rare patient that does make it out to a year and a half and 2 years with a beautiful response. I admit that my bias is usually to leave patients on pemetrexed if it’s being well tolerated. But those are patients — as great a drug as pemetrexed is toxicity and benefit wise, people start accruing their peripheral edema. I sometimes see, kind of, creeping up of creatinine that I have to attribute to that because there aren’t any other factors. So I’m very open to the idea, if there’s mounting toxicity, to give people a pemetrexed maintenance break. But I admit, that situation where I have to make a call after 2 years is pretty rare. Choosing between pembrolizumab alone or in combination with carboplatin/pemetrexed as first-line therapy for metastatic nonsquamous NSCLC DR LOVE: So I guess one more thing I’ll just throw out there, that I — kind of thinking about this situation, do you add in a checkpoint inhibitor to chemotherapy? And, I think, even if you don’t see a survival benefit to the whole trial, I kind of get the feeling that docs and patients are looking for a home run. Maybe it’s not really a home run, but a prolonged, unmaintained response. I mean, this lady’s able to come off treatment, for example. And saying to yourself, “The only way that’s going to happen is if they get a checkpoint inhibitor. Granted, they could get it second line. But that’s really the only thing that’s really going to change this in a really big way.” What do you think about that logic? DR GUBENS: I think that that’s a really fair point. I can argue this out of both sides of my mouth, and I have. And I think that the reassuring thing is — and I’m sure this was in the minds of our regulators at the FDA — the good news is, we have a fully accrued Phase III trial with a lot more patients than 123. So we’ll be able to really answer both the OS question but also get some granularity on the subgroups. So obviously subgroup analysis is fraught, but when we’re trying to look at the PD-L1 statuses, for example, and divide up these 123 patients, if you look at the data, the response rate for the less than 1% is higher than 1 to 49, which is less than greater than 50%. So there’s kind of this instability of the results that, I think, is born of small numbers, but again, when we have a greater number of patients, we can kind of tease out the greater than 50 percents. Are we really adding anything to pembro as a single agent? If we’re not, I’d love to give a single agent and save my chemo for the second line. Sequence has a role here. But again, I admit there’s a bit of a to-be-continued element to this. And we’re going to have Phase III data, hopefully as soon as the next few months. Single-agent pembrolizumab as first-line therapy for patients with advanced NSCLC and a high PD-L1 TPS DR LOVE: And, incidentally, our survey data does show for sure that when the TPS is over 50% — at least the 25 investigators we asked recently — people, they go with pembro alone in general. DR GUBENS: Completely agree. Completely agree. And I think — yes. I think that’s it. Those are the investigators you polled. What I see in my second-opinion clinic is that I think the uptake in the community is very high. And there’s an element, (A), of just trying to get trying to get as much response rate and PFS from the beginning, the logic we talked about before. One of my colleagues kind of makes the joke it’s also kind of a set it and forget it. There’s no PD-L1 selection. It’s a very easy regimen to give. And even toxicity wise, we were happy to see that there was a little bit of augmented toxicity, but it’s manageable. Carbo/pem/pembro is a manageable regimen for the vast majority of patients. So it’s a relatively easy combo to give and not to have to think too much about. Immune checkpoint inhibitor-associated cellulitis and conjunctivitis DR LOVE: Now this lady, her main problem was rash? DR GUBENS: She would have kind of a recurrent leg cellulitis back and forth. So it wasn’t the traditional kind of immunotherapy rash you might expect, and not something we typically see with pemetrexed. But it’s someone who never had cellulitis who then had it 3 or 4 times, requiring oral antibiotics, always resolved. But I think we have to conjecture it had something to do with the pembro or the combination. She also had recurrent conjunctivitis, which was not, kind of, the immune-related uveitis we think of with more concern but again, something she’d never had before, incidentally hasn’t had in the 4 months off therapy. So it makes me think there was some element of compound toxicity there. DR LOVE: That's interesting. Are there autoimmune conjunctivitises? DR GUBENS: Not that I’m aware of. Again, I think of uveitis, which I’ve only seen a couple of times. DR LOVE: Right. DR GUBENS: But again, it was something that she’s had frequently on trial that I just have to chalk up to something we were doing to her. Decision-making about the discontinuation of immune checkpoint inhibitor therapy after a prolonged response DR LOVE: One final question, which is the issue of stopping checkpoint therapy. Now, this lady was on a trial. I don't know how often in lung cancer you get into a situation where a patient’s really doing well for beyond a year, 2 years. Have you stopped checkpoint inhibitors? And how does that go when you talk about it to the patient? DR GUBENS: What an interesting discussion to have, again, because we really are in a data-free zone. Unlike chemo, where I think keeping that brake pedal down makes sense and it’s kind of verified empirically, I think there’s a case to be made mechanistically that once you’ve gotten your response that the T cells can do their job. And so I appreciate the fact that some of the studies are trying to say, “Let’s stop at 1 year,” “Stop at 2 years” but also work in the rechallenge piece to the trial as well. A lot of the trials had that opportunity to stop trial either at 2 years or with a CR but write into the protocol that a patient has the opportunity to restart. Anecdotally, I’ve seen both. Certainly, patients have been happy with their response. They’re scared to go off. And I completely understand that. So if they’re tolerating it well, I’ve kind of let them stay on. Other patients who maybe have had a good response who don’t have to go off for a bad immune-related AE but have just noticed some of the small things accruing I’ve been very comfortable to have a conversation after a year or two and say, “Why don’t we trial off and just do imaging on an every 2- or 3-month basis and just try to catch the recurrence?” And at least once or twice, I’ve been able to treat at the point of recurrence and get that response back. A couple of others I didn’t. So I think we’re in a data-free zone, and I certainly hope that we can get some more data to support these decisions going forward. MYSTIC trial: Lack of progression-free survival (PFS) benefit with durvalumab/tremelimumab versus platinum-based chemotherapy for previously untreated metastatic NSCLC DR LOVE: Any other studies out there that you’re really excited to see, maybe things that are going to be coming out in the next couple of years? DR GUBENS: So I know this’ll probably be public by the time this comes out, but I think we’re all interested to see the MYSTIC trial, which we know the press release came out. This is the study that looked at people with a TPS of 25% or higher. And what we saw was the durva plus tremelimumab, a PD-1 and a CT — excuse me — a PD-L1 and a CTLA-4 did not beat chemo in that patient population. And I think it maybe gives us pause. We’ve been excited about combos, just because we’ve seen such great responses in melanoma, for example, to the CTLA-4 plus PD-1. And we’ve seen some early phase data, especially out of Matt Hellman’s group, showing some really good Phase I responses. But now as we look at these bigger Phase III trials roll in, it’ll be interesting to see how the details play out. And the fact that MYSTIC was negative, I think, was a little bit of a warning show across the bow. And again, we have to see — was that all PFS? Was there an OS kind of difference? And again, all the devil will be in the details that we don’t have as of today. But as those data come out, it’ll be very interesting to see how that trial comes out, how the other PD-1/CTLA-4 combinations come out as well. It’s nice to know there are different companies with different agents looking at this. We have a couple of shots on goal to replicate these data or see how they differ. DR LOVE: Obviously, there’s been a lot of attention to that initial press release or data release. And I guess the question is, I mean, it’s one thing if the combination doesn’t do anything. But is this, in some way, suggesting that durvalumab did not provide the same kind of outcomes that, for example, you see with pembro? Of course, that was seen with the nivo in the first-line situation also. DR GUBENS: And that’s a really good point. And again, we talked before that I have fairly good equipoise among the 3 drugs that are approved in lung cancer in the second line, but they are different drugs. And besides that, the biomarkers are very different. The PD-L1 status does not equal PD-L1 status antibody to antibody. And I think that’s going to become an important thing to really suss out as well. But I think you make a good point, that (A), durvalumab may just be a different drug. Secondly, tremelimumab is certainly a different drug. In the trials of nivolumab and ipilimumab, we see a lot of differences in the subgroups, even by dose of ipi and dose of nivo. So it’s not quite as straightforward, I think. And some of the — again, the devil will be in the details about the dosing schedules and the drugs themselves, along with the biomarkers. So there’s a lot of variables at play that we haven’t had to reckon with as single agents. But as we look at combos, it may become more important. Nivolumab/ipilimumab as second-line therapy for small cell lung cancer (SCLC) DR LOVE: In trying to summarize some of the input I’ve gotten from investigators over the last year or so it sounds like the combination of anti-PD-1 or PD-L1 with anti-CTLA-4 in just non-small cell is not necessarily, at this point, that overwhelmingly excited. I don’t hear people talking about doing it off study. But I hear a different story with small cell. So is that kind of your take? DR GUBENS: It is, but maybe that’s more a function of small cell having so many fewer options. Certainly we’re excited about some targeted therapies, like rova-T coming to the fore. But when your best bet is platinum/etoposide, that we’ve been using for 20 years, and a second-line agent, topotecan, that I don’t know anyone who gives on a regular basis, I think the threshold for calling an interesting development is much lower. And I think that’s reflected in the NCCN Guidelines as well. DR LOVE: So specifically, how do you use checkpoint inhibitors, if you do, outside a trial setting in small cell? DR GUBENS: In small cell, it depends on the patient in front of me. I’m always wary of the toxicities in these patients, who are generally sicker even, than our non-small cell. So nivo/ipi, nivolumab/ipilimumab, is not to be trifled with. There’s real toxicity. And we’ve seen, for example, some limbic dysregulation that’s really devastating. That said, if I have a very fit small cell patient and I’ve used up my first line of chemo, I will have a conversation. And I’m willing to offer nivo/ipi, just because again, because it’s on the NCCN, I can get access to it. I always prefer a trial, of course, but these patients are — it’s harder to let them travel. It’s harder to accrue trials. I’m willing to try nivo/ipi in those folks. And I’ve had some limited success with it. DR LOVE: And are you willing to try it in non-small cell? DR GUBENS: I’m not. (A), I don’t think that the data is set — again, there are enough competing alternatives in that setting. And secondly, it’s not on the guidelines. It’s not insurance approved. And I would be reluctant to do it off trial. I think there’s some kind of hints of benefit. Again, they have a Phase III trial that’s accrued that we’ll see results out of soon. So I think it’s a stay-tuned kind of situation. I know there were some cases of fatal encephalopathy seen, particularly in the small cell patients. So that gives me pause. When you see Grade 5 toxicity with a nivo/ipi combination — again, in a small number, but maybe more than we expect from non-small cell — it gives me pause. I talk about it with my patients, and certainly I’m looking out for that. DR LOVE: So encephalopathy, I’m not sure I’ve heard about that with checkpoint inhibitors. DR GUBENS: Yes. In the small cell trials of nivo plus ipi, there were a couple of cases of fatal encephalopathy that give us pause. DR LOVE: Wow! Interesting. Huh. Potential benefit of immune-directed therapies for patients with preexisting paraneoplastic syndromes DR GUBENS: And maybe there’s something more — we see more paraneoplastic syndromes with small cell, so there may be something at play there — DR LOVE: Oh, yes. That’s true. DR GUBENS: — that we might expect some excess toxicity. DR LOVE: Yes. Actually, I was thinking to myself, “What happens if you give a checkpoint inhibitor to somebody with a paraneoplastic syndrome?” DR GUBENS: I think that’s a really good question. I haven’t done it yet, but I think that’s a really fair question in terms of this is already disease that has kind of — small cell is a disease where if you have a paraneoplastic syndrome that’s already engaged the immune system somehow, you got to give that healthy respect. DR LOVE: Yes. I mean, if you think about it, if you don’t have an antitumor effect but you do have an anti-immune effect, that maybe is not going to be too good. DR GUBENS: Exactly. Exactly. Case: A 77-year-old woman and never smoker with ALK-rearranged metastatic adenocarcinoma of the lung experiences disease progression on third-line alectinib DR LOVE: So let’s get into targeted therapy a little bit and start out with your 77-year-old woman. DR GUBENS: Yes. So she had a really interesting presentation. A 77-year-old lifelong nonsmoker, a urology nurse, actually, her presentation was actually acute right-eye blurriness. So goes to her optometrist, gets referred to ophthalmology, and they find a choroidal met. And they did staging at that point. There was a large left lower-lobe mass, some supraclavicular lymphadenopathy, osseous mets. At initial presentation — this was a few years ago — we got ALK FISH, which was actually negative. And EGFR and ROS1 were negative. So we actually started carboplatin and pemetrexed, but as we often do when our first quick pass of mutations are negative, we sent for a next-generation sequencing, which takes another 3 or 4 weeks. And so while she was on carbo/pemetrexed, our NGS panel actually showed an EML4 ALK rearrangement. So interesting that it was not caught on FISH, but we did see it on NGS. She was having a great benefit from the chemo, so we decided to play it out. We said, “Let’s” — we had already seen our first restaging. She was tolerating it well. So we actually left her on. She got a full 6 cycles of carbo/pem, actually got maintenance pemetrexed for 16 full cycles and then finally progressed. And now, hey, we had crizotinib in our back pocket and gave her that. DR LOVE: Hold on. Before you continue with the case, because there’s so many interesting things about it, obviously, the testing issue — I mean, I’ve heard stories like this before. And we try to publicize them. I think I’ve heard this with EGFR also, where your standard one-off tests negative. You pick it up on NGS, and the patient has a response to targeted therapy. I mean, there’s really a story there. Did this patient have IHC? DR GUBENS: She did not. We weren’t doing that at the time on a routine basis. Again, this idea that IHC is a great diagnostic and may even be superior to ALK FISH, there was a nice paper in CCR just this month that suggested for FNA and small biopsies, ALK IHC outperforms ALK FISH. And the nice thing is, it uses less tissue. So when you’re trying to be parsimonious anyway, with very limited tissue, I think it’s a fantastic alternative. DR LOVE: Do we know about — obviously here you have a situation that was FISH-negative/NGS-positive. What about IHC-negative versus NGS? What about FISH versus IHC? Are you still primarily using FISH? DR GUBENS: FISH is still our first pass. So in our shop, what we do is, we kind of do a rapid assay to get us EGFR, ALK, ROS1 and BRAF right out the gate, because that’s something I’m going to act on in the short term. PD-L1 as well, of course, to help decide on first-line therapy. If those are pan-negative or PD-L1 is under 50%, then I go ahead and start my standard chemotherapy with or without bevacizumab, as appropriate, but then I take the time to send for the next-generation sequencing, hoping, (A), maybe to find actionable mutations that are kind of on the second tier, things that I might be able to reach for off-label targeted therapies or clinical trials for, but also, as this case exemplifies, to kind of check our work. As we see there’s not total overlap or concordance. And occasionally we do pick up some EGFR, ALK or ROS1 alterations on the NGS that weren’t readily accessible on the rapid point mutation assays or FISH. Resolution of choroidal metastases with crizotinib DR LOVE: I’m also curious about the choroidal met. Did you actually look at it yourself? DR GUBENS: I’ll admit I didn’t. I’m not a very good internist anymore, I’m afraid. DR LOVE: Probably don’t know how to use an ophthal — DR GUBENS: I have beautiful pictures my ophthalmologist put in the medical record for me. Really was impressive. But I admit, I did not pull out my scope. DR LOVE: Actually, one of the first things I ever did in medical education, we did this video where we showed choroidal mets. I was just kind of curious. Did it get better? Did the ophthalmologist say it got better? DR GUBENS: Oh, completely. It was basically a CR by his assessment. It was remarkable. DR LOVE: So it went away with chemo. DR GUBENS: I should clarify it. For the carbo/pemetrexed, there was a nice reduction and her vision actually went back to normal, didn’t quite resolve on serial ophthalmologic exams. But on crizotinib, there was actually a clinical CR. The ophthalmologist said, “I don’t see it anymore.” It was really kind of a neat case that way. DR LOVE: Except, probably — did she have any of the visual things that you see with crizotinib, incidentally? DR GUBENS: That’s actually — I don't know off the top of my head. But I always warn patients about the little floaters on the peripheral vision. And I don’t remember if she had them or not. DR LOVE: I just wonder, if she’s going to the ophthalmologist anyhow, if he picked up on that, because I’m always curious about what that is and why people have it. But in any event, the other thing that this case brings up that we’ve kind of heard whiffs of over the years is the particular sensitivity to pemetrexed to ALK disease. Is that really a reality? I mean, this lady did great with pemetrexed and carbo. DR GUBENS: Yes. I know ever since Colorado published those data, I have to say, in my practice, these are my long-term pemetrexed responders. It’s really kind of gratifying. And so again, I think there’s some conjecture, there’s some TS, thymidylate synthase, differences in ALK patients. I don't know mechanistically why that would be, but I definitely clinically have seen these really nice pemetrexed maintenance responses. And that’s why I felt comfortable leaving her on chemo, saying, “I have an ALK inhibitor in my back pocket. You’re tolerating it well. Let’s ride this wave to its completion.” DR LOVE: Right. So she does really well then on the chemotherapy. And then she has disease progression. You put her on crizotinib. She has a great response to crizotinib. Any tolerability issues at all with the crizotinib? DR GUBENS: She tolerated it beautifully, actually. A lot of patients do have at least some mild GI symptoms that we wind up having to manage, more often in the first couple of months or so. But she actually tolerated it beautifully. She went back to the gym even, as a 77-year-old, really fit and a patient who did very well on crizotinib. Testing for specific gene mutations to direct up-front targeted therapy for adenocarcinoma of the lung DR LOVE: So I’ve been trying to figure out — our whole group trying to figure out what the current algorithm is for management of non-small cell, very biomarker driven and particularly how TPS is fitting in. And you mentioned something that we saw and we have been seeing, which is, what do you need to rule out to start initial therapy? And your list was EGFR, ALK, ROS1 and BRAF. And I would say my list, based on what I’m hearing from investigators, would include MET exon 14, although I know there’s not an FDA indication. But do you consider — maybe you can just comment a little bit about MET exon 14 and whether you — kind of what I’m starting to get the feeling for is, there are, like, 4 or 5 of these alterations that a lot of investigators will treat first line with targeted therapy, so you really need to know it up front. And then there’s others, it seems, maybe HER2 and RET that maybe aren’t that critical to get up front. Is that kind of globally — and where does MET exon 14 fit in? DR GUBENS: It’s interesting how we kind of see it. I almost see it as people who are at bat, people who are on deck. And NCCN has that nice page that looks at the — I think they call it the “emerging targets,” which I think is very useful, useful for thinking about it, also useful for insurance authorization, to be honest. I show them that page and, at least in my California practice, I always get approval for those targeted agents even though they’re off label. I admit I think that it’s hard to say when something should get promoted to first line. BRAF, as soon as it got the FDA label I kind of made that a first-line choice. I had done one or two off label before that and got it approved. Again, the MET exon 14 data are fantastic for crizotinib. I think my concern is, (A), kind of insurance approvability. I know I can get it in the second line. And, I think, across all these targetable mutations, I think we’ve always seen that chemo followed by the targeted agent or vice versa generally has the same overall survival. So I think it’s reasonable to wait for the second line and not kind of spend the time with an appeal and a peer to peer to get the approval, to say, “I’m going to get it second line. And oh, by the way, if I get a year out of chemo, maybe it bubbles up to FDA approval” or there’s a trial of an even better MET inhibitor that’s purposely designed. So that’s kind of how I think about it. But I agree with you that the 4 on top, EGFR, ALK, ROS1 and BRAF. I do want that MET, RET, HER2 and the others, but I’m willing to let that be my second tier that I get as I’ve started chemo. If I can make one other interjection — this comes up a lot in second-opinion practice, and also just because of the timing of the results — PD-L1 is IHC, so it comes back quickly, often before even targeted mutations come back and certainly before NGS comes back. So what I’ve noticed — and I’ve certainly heard colleagues mention, maybe your panel has noticed — it seems that especially ALK patients often have a high TPS. It’s really interesting. And, of course, when that comes back first, I think everyone rejoices. They want to get started on pembrolizumab. But if you have this phenotype of a nonsmoker, young patient, you haven’t gotten EGFR and ALK back, it’s not the patient who, so to speak, ought to have a high TPS, the smoker or the inflamed tumor. And I think really the results have been pretty consistent that ALK and EGFR underperform in the second line, other patients on PD-1 inhibitors, even TPS for TPS level. And in the first-line trials, of course, EGFR and ALK were excluded. So I think it’s really important, as excited as we are to get that PD-L1 result back and to get started on treatment if I’m looking at a patient in front of me that I have a high pretest probability of EGFR or ALK or ROS1, I wait for that mutation result because I think that for some mechanistic reason that none of us truly understand that high PD-L1 happens, and it doesn’t really reflect or predict for better PD-1 inhibitor response. DR LOVE: And again, I’ve heard that consistently. And this reluctance or delay in using checkpoint inhibitors in patients with driver mutations. And again, it kind of seems like the cancers that tend to be treated up front are the ones that people are pulling back from the most, so EGFR, ALK, et cetera — ie, that when they do progress on targeted therapy, I’m hearing a lot of investigators saying, even if they have a high TPS — like, as you say, for some reason, sometimes these ALK patients have it — “We’re not giving a checkpoint inhibitor with chemo at the beginning. Maybe later, but not that early.” Is that the way you’re doing it? DR GUBENS: Exactly, because again, as we talked about, let’s say an ALK patient has high PD-L1. First of all, we’ve got great agents. We’ll talk about alectinib and some of the great data coming there. Chemo is going to serve them well, too. We talked about pemetrexed maintenance. I have great options in my pocket. And as exciting and sexy as immunotherapy is, my expectation of response rate and duration of benefit is shorter for these patients. So I’d rather give them these better agents up front. But it’s interesting. In this milieu, where patients are hearing about immunotherapy on the TV, that that’s what they reach for, even though I can say, “Actually, I’ve got this drug in my pocket that I think will work better, longer, more safely right up front.” Therapeutic options for patients with ALK-rearranged advanced NSCLC and disease progression on crizotinib DR LOVE: So this lady gets on crizotinib. She does great on it. Then what? DR GUBENS: So interestingly, she progressed after 15 months. And she literally progressed the month that alectinib came on the market for crizotinib failure. So her timing was impeccable. She went on alectinib, which she tolerated also quite well, had about 10 months of disease response and really good quality of life. And then on restaging scans actually had development of bilateral adnexal masses, kind of an unusual site of metastasis, and nowhere else. So odd enough — whenever there’s kind of an oligometastatic situation, I’m always kind of interested in making sure and actually offering a local therapy. We wound up sending her to our gyn-onc colleagues. And they did elect to resect those masses, which did reveal ALK-positive metastatic lung adenocarcinoma. But, unfortunately, about a month later she had much more widespread abdominal progression as well as some lymph nodes growing in the chest. So really had realized that alectinib had come to its end and it was time to move on to the next line of therapy. DR LOVE: That’s really interesting, though, about the adnexal mets. That’s unusual. So what happened next? And where is this lady now? DR GUBENS: Right. So I think we’re in this interesting time where we know there’s so many ALK inhibitors on deck. This progression happened shortly before brigatinib was approved, which is, I think, a reasonable agent to use. But we kind of felt like there was bulky enough disease, we wanted a response. We knew she was chemosensitive. We did decide to start chemotherapy, but pending, what I’m doing is actually sending that adnexal mass pathology to see if we can identify any ALK-resistant mutations. Now, I know this hasn’t kind of broken through to standard of care, but as we think about sequencing the ALK inhibitors and when I need a response, I kind of almost want to make sure that one of the resistance mechanisms is addressed by, say, brigatinib or, in the future, maybe lorlatinib. But I’m actually hoping that I can get a chemo response and give me time to choose the next line of therapy. But I think this idea is interesting, where right now alectinib first line, on progression on alectinib there are other drugs on the market, but can we actually do precision medicine to figure out which of these bevy of ALK inhibitors is right? One of our colleagues calls it a “Sanford guide.” From a residency, we used the Sanford guide of antimicrobial resistance. You kind of go down the chart, look across and choose your agent. I think we can all envision a day when you do a repeat biopsy, or even liquid biopsy, figure out how ALK has morphed and choose the drug on the list that hits that best. Lorlatinib hits G1202R, which is kind of the dominant resistance mechanism, but there are many others, and so maybe actually purposely choosing drugs as appropriate. For her I think the science is too far ahead of her. So I did choose to do chemo next. But I am kind of profiling that tumor and seeing if it might inform which ALK TKI I use next. DR LOVE: That’s really fascinating. We’re going to talk in a second about plasma testing with EGFR. What about plasma testing with ALK and these resistance mechanisms? DR GUBENS: Yes. It’s coming along. I think that — again, it’s just amazing how quickly this technology has progressed. I still have not been as impressed by the fidelity of ALK catches on the liquid biopsies. So I think month by month we’re seeing better results. I still prefer for my ALK patients, for the rearrangement patients, I should say, both ALK and ROS, I still prefer tissue biopsy, but I am seeing some of the panels come through with concordant ALK initial mutations and ALK resistance mutations. So it’s kind of exciting. Clinical utility of FDA-approved (alectinib, brigatinib, ceritinib) and investigational (lorlatinib) ALK inhibitors in advanced NSCLC DR LOVE: So, of course, we saw the ALEX data at ASCO. Seems like alectinib should be on its way towards being standard first-line therapy. Is that your take? DR GUBENS: No question. And I would fight for it. I know it’s not FDA approved yet, but the size of that benefit, the tolerability and, above all, the CNS control is really so compelling that I don’t think it’s appropriate to start someone on anything other than alectinib at this point. DR LOVE: But then, as you say, the question is, what’s going to be next? And, of course, there’s a lot of variables. You mentioned resistance, maybe brain mets. But, I mean, just globally, what do we know about responses? Brigatinib, at least it’s new. Do you see responses? For example, have they been reported to, say, patients who had either alectinib or crizotinib, do you see responses to brigatinib? DR GUBENS: That’s a really good question. I think the breadth of data that got brigatinib approved really was just in the crizotinib failure, initial line of therapy. And I think we hear anecdotes of folks getting responses even after subsequent TKI progression. I think the more compelling data really are coming from lorlatinib. And it was so exciting to see a response rate after 1 TKI, 2 TKIs, even 3 TKIs and still after 3 TKIs, lorlatinib has — I think it was a 30% to 40% response rate. So I’m not sure brigatinib has been able to show us that second-, third-, fourth-line data. So that’s why I was a little circumspect in this patient. Maybe I don’t want to throw that at her now. I need a response, because she has bulky disease. But in the future, again, if I can be guided by resistance mutations or by data on the ASCO screen that shows me progression in unselected patients after 3 or 4 lines, then I’m more convinced. DR LOVE: What about ceritinib? I mean, it’s another potent agent. Can you describe a situation where you might use it? Would you use that after alectinib? Can you describe a situation where you use the drug? DR GUBENS: Right. I mean, ceritinib has a role. I think it’s always a drug development game. They were first to the field but unfortunately hadn’t kind of refined their way of giving it. So it was a toxic drug to give at full dose. At 450 mg, I think it actually is very tolerable. We’ve had a couple of patients on it. Again, it’s one that I would reach for, even in this woman, I think, after failure of alectinib. It wouldn’t be unreasonable. I don’t think we have data to say brigatinib or ceritinib would be the right next choice. But again, what I’m looking for, I hope, over time is this kind of idea that I can look for the actual resistance mutation and act on it. I admit that all those graphs we see on the screen are still mostly preclinical in nature and not patients who have this mutation who get this response rate in a human. But I think it’s the route we have to take, just the way we did with it with T790M and EGFR. But I think there’s going to be more multiplicity of resistance mechanisms and certainly more drugs to offer. But ceritinib still plays a role. And I think it just — I still would put it well after alectinib at this point. Case: A 74-year-old man and never smoker with EGFR exon 19 mutation-positive adenocarcinoma of the lung and multifocal brain metastases receives osimertinib DR LOVE: So let’s finish out and talk about your 74-year-old man. DR GUBENS: Right. So I had this 74-year-old gentleman, a lifelong nonsmoker. He presented with a large, right upper-lobe mass and extensive bilateral pulmonary nodules. And so he was actually symptomatic because of cough and some mild chest pain. It turned out on screening he had 26 subcentimeter brain metastases, completely asymptomatic but really an impressive presentation. So we sent him to our radiation group. And I expected them to offer him whole brain radiation as we did the workup for his mets. They actually — and this is way off the reservation, I admit. But our Gamma Knife® group is fairly aggressive. And, I think, because of the very tiny size of the brain mets, they did wind up offering him Gamma Knife, which he tolerated quite well, so stereotactic radiosurgery. DR LOVE: So I’m just trying to think about doing — so if you do that on 26 lesions, that’s a lot different than whole brain? DR GUBENS: That’s a really good question. They really were tiny. And to be honest, I think that there was a little bit of a delay to his mutation analysis. And I think — this was a couple of years ago. I think now I would have felt more comfortable in this asymptomatic patient if I knew he had an EGFR mutation and these almost miliary but multifocal brain mets. I would have let him get the EGFR TKI for a period of time, maybe just 4 weeks or 6 weeks, again assuming asymptomatic brain mets because we see such a nice brain response. So I admit that I think I might have handled that a little differently if I had seen him before the radiation docs had seen him, to be honest. So I definitely want to stress, a Gamma Knife to 26 lesions not standard of care, shouldn’t be standard of care. But he actually tolerated it beautifully. DR LOVE: Wow! Interesting. So this paradigm — and the key, I think, from what we’ve heard — is what you mentioned, asymptomatic. So no symptoms in the brain or related to the brain mets, particularly, as you say, with smaller lesions. So a lot of people, the majority will do what you just said, systemic EGFR TKI. DR GUBENS: Especially in this case, where again, the Gamma Knife was exception. This is a patient who, by all rights, probably should have seen whole brain radiation, or that would have been the traditional recommendation and still would be for the nonmutated patient. So I think we are so loathe to give whole brain because of that potential for longer-term cognitive function, that when we have a CNS active drug, we will go to great lengths to try to forestall the whole brain. And again, in this case I think we probably could have. And we did, fortunately, but again, it was kind of an odd progression of events. DR LOVE: So this paradigm, though, of using targeted therapy in people with asymptomatic brain mets I’ve heard people talk about freely in patients with EGFR and ALK, but what about ROS1? What about BRAF? What about MET exon 14? Again, small lesions, asymptomatic, are you going to try targeted therapy? DR GUBENS: I think that it really depends on the patient. I think when we’re talking about 1 or 2 lesions that are asymptomatic, we have the luxury of time and follow-up. And we can really see if these targeted therapies work straight out of the gate. And if not, then we have the opportunity to use our stereotactic radiosurgery. But I think yes, for all the ones you mentioned, though, that’s a very reasonable approach. And again — but that somewhat depends on symptoms and size. But, for example, some of these targeted therapy trials have been bold enough to allow untreated asymptomatic brain mets on trial, ALEX included. So they get a lot of credit for proving that that’s a very reasonable treatment paradigm. DR LOVE: Would you consider the same paradigm if you were, let’s say, using a checkpoint inhibitor first line in somebody with TPS over 50%, again, asymptomatic? DR GUBENS: That’s a vexing question. And I think that we just don’t have enough CNS data at this point for the checkpoint inhibitors that I would be willing to wait. It’s a conversation. And I think for the motivated patient who I trust I can do follow-up, we could think about it. I know there’s some early data out of that Yale study that says there’s a modest response rate in untreated brain mets. And that’s across a melanoma and a lung population. But that starts making me a little bit more nervous. But if it helps me forestall whole brain, if it lets me do stereotactic to a small number of lesions, it may be a reasonable strategy. But I admit that for the checkpoint inhibitors, I haven’t been as willing to do that. First-line erlotinib versus afatinib or gefitinib for EGFR-mutated metastatic NSCLC DR LOVE: So this patient, in addition to the brain mets, as you mentioned, had a lot of disease inside the chest. The patient had an exon 19 deletion. I see that you decided to use erlotinib. Just a simple basic question: Why erlotinib as opposed to afatinib, and, for that matter, gefitinib? DR GUBENS: Right. So certainly it’s something we discuss all the time. Especially for this gentleman who’s a 74-year-old, I’m even more mindful of the toxicity profile differences between the drugs. Afatinib clearly has a bit of PFS benefit. We still really haven’t seen an OS benefit. And so it’s a conversation I have. And I offered him afatinib and talked about the data we have before us. But I think for most of my patients, I’m kind of leaning them toward erlotinib. Gefitinib I also haven’t really been too enthusiastic about. Since we’ve used erlotinib, so much and my staff are very facile with it. I think we feel very comfortable in our ability to pull back on dose and manage toxicity even in our kind of older or frailer patients that I don’t feel the need to reach for gefitinib first line, though it’s a very reasonable option, of course. But again, I just haven’t found many patients who find the toxicity benefit palatable or the PFS benefit palatable enough to justify the additional toxicity. DR LOVE: So it looks like this patient did really well on erlotinib. DR GUBENS: Did beautifully well. DR LOVE: What happened with the tumor? DR GUBENS: Twenty-two months, so really better than the median, really did quite well with very nice lung and mediastinal response. But then after 22 months, he did have progression. We biopsied a subcarinal node that had grown. This was kind of before liquid biopsy was as widely available. I would be very comfortable at this point offering a liquid biopsy. But we did do a biopsy that revealed a T790M resistance mutation. DR LOVE: Incidentally, how did he do on the erlotinib for the 22 months, tolerability wise? DR GUBENS: Kind of as you’d expect, mild rash, worse in the first month, that we managed, but then very mild and tolerable thereafter. He actually didn’t have any diarrhea, which was nice to see, really tolerated it quite well and never actually had to have a dose reduction. Identification of T790M EGFR resistance mutations in NSCLC DR LOVE: So I’m kind of curious. I’ll make a couple of statements. Tell me if you agree or not. I think you’re going to agree that in a situation of a patient progressing on an EGFR TKI, regardless of whether they’ve responded or not — although maybe — I’d be curious to know what you think about somebody who progresses right through it — but in general, T790 testing, standard of care? DR GUBENS: Completely. I should say that my practice now — maybe we were going to get to this later — I generally, when patients have a fairly indolent progression, which is often the T790M clinical pattern, my approach has generally been to offer the liquid biopsy and schedule the patient back in 2 weeks, because I usually have almost certainly a response by then. If it’s negative, then at that point I schedule a biopsy, if it’s at all feasible. I think that it’s a situation where we can trust a positive liquid biopsy result, but a negative result, especially when we don’t even have the EGFR initial mutation seen on the liquid biopsy, kind of meaning the uptake wasn’t good enough, I really think this patient would warrant a physical biopsy as well. For patients who are progressing a little faster, it doesn’t mean it isn’t T790M, but then I kind of move at a little more speed. So I might do the liquid biopsy and schedule the biopsy in the same week, 2 shots on goal. Within a week or two, I have both results back and I don’t run into the issue of discordance or negative results in a patient who needs to change therapy more rapidly. DR LOVE: In a patient who’s clinically stable, do you see any rationale for looking for T790 earlier? I don't know — I’m not even sure what data we have about when T790 occurs. And how long is it from the time you can detect it to the time you have clinical progression? What about that preclinical progression space? DR GUBENS: I think it’s fascinating, because we know that this evolution is part of what we’re addressing. But I think in the clinical practice, I really would push against the idea of checking T790M on a regular basis. I really wait for some kind of clinical progression. Utility of osimertinib for patients with advanced NSCLC who acquire the T790M EGFR mutation DR GUBENS: Now, of course, we’ll talk about the adoption of osimertinib, maybe in the first line. And so that’ll change the whole scope of things. But I think the correlative science is fascinating. And maybe we will come to a day when the moment we detect T790M in blood, that’s the time to start. I know that there’s kind of an interesting study design looking at that. A really interesting idea of doing serial liquid biopsies and, basically, the patient — have 1 arm have patients switch to osimertinib at the point of T790M detection in the blood, even short of radiographic progression, versus only doing it at radiographic presentation. And I think it’s a fascinating idea. But for now, I’m not looking for T790M until there’s a clinical progression event. DR LOVE: Although maybe by the time most people watch or hear this interview, it’s all going to be, as you said, a moot point, because we’re about to see — DR GUBENS: It might well be. DR LOVE: — the result of the — what is it? — FLAURA trial, looking at osimertinib first line. The preliminary suggest it was, quote, positive. So we’ll see what happens. But what happened with this patient was really incredible. Maybe you can talk about it. DR GUBENS: Right. So given the T790M resistance mutation, we went ahead and started osimertinib as a standard of care. Again, tolerated it very nicely, had a very nice reduction on the very first scan and had several months of benefit. And then — DR LOVE: Could I just — I’m just kind of curious before you go further, you said he tolerated it well. I’m just kind of curious, as you were seeing him in clinic, how was his quality of life like on the osimertinib compared to erlotinib? DR GUBENS: It was a little better. He noticed a clear benefit in terms of energy. And he had a little bit of drop-off of energy with his progression event. But when he was a couple of months into osimertinib with a nice response, he could say, “I haven’t felt this good since before I was diagnosed.” So it was really gratifying. Very sold on osimertinib, I can tell you that. BLOOM: Activity of osimertinib in patients with leptomeningeal disease from EGFR-mutated advanced NSCLC in a Phase I study DR LOVE: So then what happened next is really fascinating. DR GUBENS: Right. So it’s interesting. He was an avid hiker, avid tennis player. And he noticed that he had this very subtle weakness, something that I would have never detected on exam. But he was just — he was a little slower catching his serves. He would actually have a lot of discomfort climbing down a hill. And so he was actually referred to a neurologist by his primary doc. The neurologist did detect some loss of vibratory sense, very subtle findings, but said, “We know your history of brain mets. We know you have lung cancer. Gubens, should we do a more comprehensive study?” So he said, “Let’s get an MRI of the spine.” Lo and behold, despite hiking and playing tennis, but with these subtle findings had intradural disease and some intrathecal deposit — some leptomeningeal deposits, which were very concerning. We actually did an LP. And an LP revealed positive cytology. DR LOVE: Wow! DR GUBENS: So he’s a gentleman who has been on erlotinib for 22 months, osimertinib for 14 months and he has leptomeningeal disease, but he’s not that sick a patient that we kind of think of, that diagnosis that strikes fears in our hearts. Here’s a guy who is fully functional, more physically active than I am and yet he had clear neurologic deficits by a neurologist’s exam and radiographic findings. And so that was right after the Bloom data had come out at ASCO in 2016, which had shown that osimertinib at 80 and at 160 had, actually, really nice leptomeningeal benefit. So we said, “We’ll try to get the double dose.” And with that double dose, on his subsequent MRI he had slight reduction in those deposits we had seen on the first MRI and some of the symptoms he reported had improved. So now I totally grant his leptomeningeal presentation is way off the map. This was indolent, which we certainly don’t associate that adjective to leptomeningeal disease. But he had a clinical benefit with this double dosing. Activity of osimertinib in patients with T790M mutation-positive advanced NSCLC and brain metastases DR LOVE: You mentioned the Bloom study. But what do we know in general at this point about osimertinib in CNS disease, particularly compared to erlotinib/afatinib? DR GUBENS: Certainly has great intraparenchymal benefit. We’re seeing really nice brain benefit at regular dose. I think over the years we’ve kind of toyed around with different applications of erlotinib, which has good penetration but not great penetration. You talk about pulsatile dosing and the like. But osimertinib at full dose has had really nice CNS response rates. So that’s very gratifying. The leptomeningeal piece, I think, was really inspired on the part of the drug developer to say, “We’re confident enough in our CSF penetration, let’s actually prospectively study some of these very potentially sick patients.” And again, small study size. I think it was on the order of 30 to 35 patients. But of the patients who got this osimertinib for a new diagnosis of leptomeningeal disease, a majority of those folks who got to that 12-week scan, which already is a good time frame for leptomeningeal, we’re seeing radiographic and clinical benefit. Mechanism of action and efficacy of the antibody-drug conjugate rovalpituzumab tesirine (Rova-T) in DLL3-expressing recurrent SCLC DR LOVE: So one final question. You mentioned earlier when we were talking about small cell, rova-T. Could you just take a minute and explain — because there are not too many antibody-drug conjugates out there right now in clinical oncology. And I’m curious what we know. What is rova-T, and what do we know about it? DR GUBENS: Yes. So delta-like ligand 3, DLL3, has been identified as overexpressed on a lot of neuroendocrine tumors, especially small cell lung cancer, and some really nifty work of synthesizing, basically, an antibody to the DLL3. And yoking it to a toxic payload, an ADC, has resulted in some really nice response rates and with a gradient of benefit where people whose small cell lung cancer has high DLL3, better response rate, better durability. So this really neat drug development has really gone nice and quickly for a patient population who has had, again, really nothing since platinum doublet for the last 20 years. And I think it’s really gratifying to see this design happening so fast. And I know there are concurrent trials in the first line, in the second-line setting and hoping for registration in the near future. Really exciting stuff. DR LOVE: Have you yourself used the drug in a trial? DR GUBENS: I have not used it in a small cell patient. We actually at our center have a separate trial, again looking at the rest of the neuroendocrine tumors. So there’s small cell, of course, that are very high-grade, high DLL3 expressers. But the company, appropriately, is looking at other swaths of the population, so looking at folks who have, say, atypical carcinoid or large cell neuroendocrine of the lung, things that again don’t have great options. But this is at least promising but very early. I don’t think that we have really any data reported out for that yet. Case: An 83-year-old woman and never smoker with EGFR exon 19 mutation-positive advanced adenocarcinoma of the lung acquires a T790M mutation DR RAMALINGAM: Sure. So this is a patient I continue to see. I just saw her 2 weeks ago in my clinic. She’s 83 years old. And about 4 years ago, she was diagnosed with advanced-stage non-small cell lung cancer. It was an adenocarcinoma. She had presented with an upper-lobe mass in the right lung with multiple bilateral pulmonary nodules. And when we conducted molecular testing, she had an exon 19 EGFR mutation. So we treated her with erlotinib, 150 mg per day, which she tolerated well and achieved a partial response and did really well. And we saw her every 3 months and, for the first year and a half, everything was fine. And then toward the second half of year 2, we started seeing some of the pulmonary nodules start increasing in size. Initially they were minor increases and, therefore, we decided to keep her on erlotinib. But a few months later, there was full-blown progression. These nodules were doubling and tripling in size. And there were new nodules, and she was beginning to experience cough and some dyspnea. So at this point we tested her plasma as the first step in our algorithm for T790 mutation. And she indeed harbored the T790 mutation. So we started her on osimertinib, 80 mg per day, about 8 months ago. And within the first few weeks, she started noticing improvement in her cough. She was feeling better overall. And the scan showed significant reduction in the size of her nodules. Some of them were gone. And she has maintained her PR. She’s fully active, fully functional. In her full-time work, which she does, she’s an artist. And she paints and goes to shows and exhibits her paintings. And she’s able to do all of those things quite well. And she was seen with a scan 2 weeks ago and is looking good. DR LOVE: So just to go back through a few points in her case, beginning with her first diagnosis when she was put on erlotinib, she was 83 years old. I guess it sounds like she was in very good condition. DR RAMALINGAM: She had a good performance status to start with. And she had minimal comorbidities. She was a never smoker. And she wasn’t somebody who’s been in a hospital in a long time, until she presented with the lung cancer. Selection among available EGFR tyrosine kinase inhibitors (TKIs) as first-line therapy for EGFR mutation-positive NSCLC DR LOVE: Where do you stand today in terms of selection of a first-line TKI? We have 3 available. And does it matter what type of EGFR mutation it is in making that determination? DR RAMALINGAM: So the 2 common hot spots in EGFR, exon 19 and 21. These are the patients where EGFR TKIs work really well. Erlotinib and gefitinib are the first-generation agents. And afatinib is the second-generation agent. And these are all approved based on Phase III trials that showed that they were better than chemo. But what’s exciting now is, we have started seeing results of trials that compared one agent to the other. And there are 2 trials we have data from. One compared afatinib to gefitinib. And in that trial, there was a modest PFS benefit, but afatinib also had more adverse events. There was also a study that showed perhaps if you had exon 19 mutation, afatinib may be a better drug for you than if you had an exon 21 mutation. There was a trial reported that compared another second-generation EGFR drug, called dacomitinib. And that was compared to gefitinib. And that trial showed a significant improvement in PFS, 14½ months versus 10 months, favoring use of second-generation drug over first generation. So at least based on 1 Phase IIB and 1 Phase III trial, I think it’s fair to say that the second-generation drugs seem to have a higher efficacy compared to first-generation drugs, like erlotinib or gefitinib. But they also come with higher toxicity. So if you have an older patient, such as my 83-year-old lady, I perhaps would still stick with the first-generation drug, like erlotinib or gefitinib. But if I had a younger patient, I might be more inclined to put them on afatinib and dacomitinib if and when it becomes available. DR LOVE: Do you see that same tolerability issue — and particularly in terms of diarrhea — with dacomitinib that you see with afatinib? DR RAMALINGAM: They seem to be very similar in terms of adverse event profile. DR LOVE: What’s your clinical experience with afatinib? Do you use preemptive antidiarrheals, for example? DR RAMALINGAM: At the approved dose of 40 mg a day, I do see more diarrhea and skin toxicity. Teaching the patients about how to manage this, proactive antidiarrheals, proactive skin prophylaxis, are important. Very often I end up reducing the dose to 30 mg per day, which seems to be a much better tolerated dose. Tolerability of osimertinib compared to erlotinib DR LOVE: So I’m curious with your 83-year-old lady. How did she tolerate the erlotinib compared to the osimertinib? DR RAMALINGAM: Most of the patients that I’ve treated with osimertinib in the second line after they have received erlotinib in the first line tell me that this is a lot easier compared to what they went through, first-line therapy. Now, that could be a result of 2 factors. One is possibly attributed to the fact that osimertinib is more selective to the mutant receptor and has less of an effect on wild-type receptor. So the toxicity profile is better. The second is perhaps the skin has already been treated with an EGFR TKI and it’s been primed and perhaps less sensitive when you expose a second agent. DR LOVE: How about with this lady? DR RAMALINGAM: This lady actually did much better. With the first course of erlotinib that she received, skin was the main issue. She had some Grade 1/Grade 2 skin effects initially, which were much better as the treatment went on. With osimertinib, she had practically no skin toxicity or GI. FLAURA study results: Osimertinib versus erlotinib or gefitinib as first-line therapy for EGFR-mutated advanced NSCLC DR LOVE: So, of course, osimertinib came into practice in patients with T790-mutant disease. What do we know about it, though, in the patient who presents up front generally without T790 mutation? And where are things now with the so-called FLAURA trial that’s comparing osimertinib to erlotinib up front? DR RAMALINGAM: Sure. I can go over that, Neil. We presented data looking at 60 patients with non-small cell lung cancer with EGFR mutation that received osimertinib as front-line therapy. And we reported that in those patients, the response rates were about 77%. And the median PFS was approximately 19½ months. But those exciting results led to the first-line trial, FLAURA, which compares osimertinib 80 mg per day to either gefitinib or erlotinib. This is a Phase III trial. It accrued more than 500 patients. And we saw from a press release that it met its primary endpoint. We’ll have far more information on the response rate/PFS in this Phase III trial and so forth. Antitumor activity and no evidence of acquired EGFR T790M mutation after disease progression with osimertinib as first-line therapy for EGFR-mutated advanced NSCLC on the Phase I/II AURA trial DR LOVE: From what you’ve seen so far, it kind of reminds me a little bit about the questions that came up about alectinib versus crizotinib with ALK, which is, why is it better? Is it just a better targeted therapy, or is it something to do with better effect in the brain? What do we know about that in terms of EGFR and osimertinib versus, say, erlotinib? DR RAMALINGAM: Sure. So when you treat patients with erlotinib/gefitinib or afatinib, nearly half of them will develop resistance mediated by T790. Now, what we can say is, when you treat them front line with osimertinib, in the Phase I experience that we are publishing in JCO, not a single patient developed T790 as a resistance mechanism. So you are shutting down entirely that pathway as a potential mechanism of resistance. And as a result of that, you see a median PFS that was reported in that trial of approximately 20 months. So when you have a patient with EGFR now, the current approach is, you give them erlotinib or gefitinib or afatinib. You have a median PFS of about 10 to 11 months. And half of them will develop T790. And you give them osimertinib, and you get another median PFS of about 10 months to 11 months. So the case has been put forward, by giving both of these in a sequential approach, you might get to 20 to 22 months. But the fact that is not accounted for is, what happens to the other 50% that does not develop T790? And they don’t get a chance to have a better outcome. So this is shifting the biology of the disease away from T790 and having a strategy that delays and avoids resistance, and this is what all of our strategies with targeted therapies are. How can we delay resistance? How can we avoid resistance? We also know that osimertinib has activity in the brain, but in my view, a lot of the effect may be from shutting down T790 and perhaps some to the brain effects of it. DR LOVE: Globally, are brain mets more of a problem in terms of resistance, particularly in the first-line setting, with ALK as opposed to EGFR? DR RAMALINGAM: Yes. The ALK-positive patients tend to have brain mets far more commonly than we see with EGFR mutation. DR LOVE: So is it your take — because this issue came up again with ALK disease and alectinib — that indirectly — is it going to turn out that by giving osimertinib up front, I guess, people have talked about the time to progression compared to going with, say, erlotinib and then, as you say, the patients with T790, osimertinib? Globally, do you think there’ll be a greater global benefit to patients by starting with osimertinib? DR RAMALINGAM: I think if the Phase III trial results show that the type of results we saw in the Phase I experience are held up, when you have a median PFS of approximately 19 to 20 months, it would be hard not to use that in the first-line therapy. Another factor is tolerability. We know that osimertinib is easier for patients to tolerate compared to either erlotinib, gefitinib or afatinib. So I do feel that there will be a shift in paradigm if these numbers hold up. Plasma testing for T790M mutations DR LOVE: And I wanted to ask you, also, this lady was diagnosed with T790 disease based on a plasma assay, as a lot of people are nowadays. And I’m kind of curious what you would have done or how you would have thought it through if she had been T790-negative. And particularly I’m interested — and I don't know if this applies to her — in the situation where it’s going to be difficult to get tissue. Is there a way to make this determination without — say somebody has some bone mets or something, the patient’s older, et cetera. Is there a way to just rule out T790 purely using plasma? DR RAMALINGAM: The plasma is a good initial step that we use. And I order them. The current assays are quite sensitive for T790. And the sensitivity of these assays range anywhere from 50% to 85% across studies. So if you did a plasma test and you detected it, that’s enough evidence to proceed with therapy with osimertinib. But if the plasma test was negative, I would still have tried to get a biopsy of one of the pulmonary nodules for this patient, because there is a 15% to 30% chance you missed that T790 on the plasma. Now, if you had no disease to biopsy and the patient only had bone-only disease, that is a bit of a problem, because if the plasma is negative in those patients, you are left with a bone biopsy, which does not provide you with material that’s conducive for genetic testing. I would still try to test it. I wouldn’t completely shy away from testing it before I make my treatment decision. DR LOVE: So, I mean, I guess — I don't know if you could actually access it, but theoretically you could do empiric osimertinib. But the other thing I’ve heard about — I think it was Heather Wakelee, actually, in an interview for this series — is in the T790 patients, to see in the plasma if the mutation is detected. And that if it’s not there, the basic EGFR mutation, then you’re less confident. If you see it there, more confident. I hadn’t heard that before. Have you heard that? DR RAMALINGAM: Yes, I have. I heard that theory. In other words, if your EGFR mutation is present, then you are detecting something coming out of the tumor. And when you don’t see T790, it’s more likely to be a true negative as opposed to a plasma test that does not detect any mutation at all. And again, I think that’s perhaps one possible way to get to the answer, but these mutations are present in much smaller frequency of tumor cells and, therefore, it’s not a foolproof way to say whether you have a right-positive or a right-negative or a false-negative. Design of the Adjuvant Lung Cancer Enrichment Marker Identification and Sequencing Trials (ALCHEMIST) for patients with early-stage disease DR LOVE: So this also brings up the issue of the potential use of TKIs in people with targeted therapy in the adjuvant setting. I thought maybe you could comment a little bit on the ALCHEMIST study that you’re so involved with. It kind of seems like osimertinib not only is — maybe — seems like it’s going to be more effective to start with. But also to take a TKI for a year or two might be a lot easier, if it’s osimertinib. But where are we right now in terms of trials looking at this question? And are we looking at osimertinib in the adjuvant setting? DR RAMALINGAM: Yes. The ALCHEMIST study is for patients with early-stage non-small cell lung cancer, Stage IB to IIIA. These patients all undergo surgery. And they all, if they are candidates for adjuvant chemo they get 3 to 4 cycles of platinum-based chemotherapy. And while they are getting chemo, they are screened for EGFR and ALK. If they are EGFR-positive for mutation, then after they complete chemo, they are randomized to erlotinib or observation. And they get it for 2 years. And if they are ALK-positive, they get crizotinib or observation. And if they are negative for both EGFR and ALK, we now have a third arm, which is nivo versus observation. And the nivo is given for 1 year. So that’s the framework of ALCHEMIST. The goal is to screen about a thousand patients, so we have about 350 EGFR, about 350 ALK-positive patients and about 700 patients on the immunotherapy arm. DR LOVE: So the targeted therapy and the checkpoint inhibitor are after adjuvant chemo? DR RAMALINGAM: That is correct. So the standard approach is, you give adjuvant chemo and then you randomize them to one of the targeted or immunotherapy approaches. But if a patient, say, is declining adjuvant chemotherapy or the patient is not a candidate, you can then directly go to the next step of the trial, which is the randomization. ADJUVANT (CTONG 1104): Initial results of a Phase III trial evaluating adjuvant gefitinib or vinorelbine/platinum for EGFR mutation-positive Stage II to IIIA NSCLC DR LOVE: Any comment about the trial presented at ASCO looking at adjuvant gefitinib versus chemotherapy? I thought that was really an interesting study. DR RAMALINGAM: Yes. That study from the Chinese Cooperative Group by Dr Wu showed that gefitinib had a much better disease-free survival compared to chemotherapy. Now, it’s a relatively small study. And one of the main caveats is, for patients randomized to the chemotherapy group, nearly 25% of them did not get any chemo at all. So that basically dilutes the results that you see, because it was not a comparison of chemo versus gefitinib for all patients. It’s chemo versus gefitinib for three fourths of the patients and gefitinib versus nothing for the remainder of the patients. So our approach is, chemo is proven in this setting at this point. Those data from the Chinese Cooperative Group, while interesting, are not sufficient to displace chemo out of the treatment algorithm for our patients with EGFR mutations. DR LOVE: Although, I mean, it is kind of strange to think about the idea of receiving chemo knowing that you can take an oral therapy that’s going to give you a higher PFS. I guess you have to believe in the survival data. DR RAMALINGAM: So the study was relatively early in its readout. The median follow-ups are relatively slow. We don’t have any information on overall survival. And this was a very early, top-of-the-line look at the top-line results. We anticipate getting more data. I agree with you. If we see good results with gefitinib alone, which is an oral agent that patients can take at home, and if it results in a dent in the overall survival even better than chemotherapy, then we would absolutely adopt that. DR LOVE: So again, getting back to your 83-year-old lady, if she were clearly T790-negative — now she sounds like she’s starting to get symptoms from the disease. How would you have treated her — would it have changed based on what her PD-L1 level or her TPS numbers were? DR RAMALINGAM: Sure. So our approach for T790-negative patients has been to do a platinum doublet combination. In this case, a carboplatin/pemetrexed regimen would be the one that I would have used. Now, if her PD-L1 expression was really high, whether an immune checkpoint inhibitor would be effective in this space is a question for which we don’t have a clear answer. So my approach would be I would still use the platinum doublet and then go to checkpoint as the next line of therapy for her in that situation. Case: A 63-year-old woman with ALK-rearranged metastatic NSCLC who had to discontinue crizotinib/immune checkpoint inhibitor therapy on a clinical trial subsequently receives alectinib DR LOVE: Why don’t we move on and hear about your 63-year-old lady? DR RAMALINGAM: Sure. So this is another 63-year-old patient. She was diagnosed with Stage IV adenocarcinoma based on presentation with pleural effusion, which resulted in shortness of breath. And subsequently the fluid was tapped, and she was found to have a malignant effusion. She also had bilateral nodules. And testing revealed ALK positivity. So she was seen in clinic. And we were going to start her on first-line crizotinib. We also had a clinical trial with crizotinib and an immune checkpoint inhibitor. And she opted to enroll in the clinical trial. And she started the treatment, but within a few weeks she developed elevation of her liver enzymes. And that resulted in treatment interruption. And we had to treat her with steroids and discontinue the checkpoint inhibitor, as required by the protocol. And then we just kept her on crizotinib for — we reintroduced the crizotinib after a few days. And sure enough, her liver enzymes started going up again with crizotinib exposure. So at that point we decided that continuing on with crizotinib is no longer likely to be safe, so we stopped it. And we switched her to alectinib, but before we did that we got a scan, which was to serve as the baseline for alectinib. And, interestingly, that showed a good response to treatment. And now she has been on alectinib, which she tolerated really well. Her liver has not been affected by this therapy. She has a partial response that’s been maintained. She’s now, I would say, about 8 months into treatment and is doing really well. Overview of the activity of available ALK inhibitors DR LOVE: Can you talk a little bit about where we are in terms of choosing agents for ALK disease? We saw some data, first-line data, at ASCO looking at alectinib. We’d seen some the year before. And we have a new drug approved, brigatinib. Can you talk about kind of where these data and this new agent fits in? DR RAMALINGAM: Sure. So our treatment paradigm, until recently, was crizotinib as the first-line treatment. And then when patients developed resistance, we had 2 options. We could go to ceritinib or alectinib, which are both potent ALK inhibitors, second-generation ALK inhibitors. Ceritinib actually is now approved in the first-line setting as well. And that was based on a Phase III trial that compared it to chemo. And the median PFS with ceritinib was over 16 months. And with crizotinib, we would expect a median PFS of about 10 months. So ceritinib is an approved option. And then alectinib, which is also an effective agent, was compared to crizotinib in a Phase III trial. Now, there are 2 Phase III trials — one done in Japan, one done rest of the world — that compared alectinib with crizotinib. And both these trials resulted in very impressive hazard ratios for progression-free survival. In the Japanese trial, the hazard ratio was 0.34, favoring alectinib. And the median PFS data was over 2 years. And in the ALEX trial, which is in the Western population, the hazard ratio for PFS was 0.46. So based on these results and the good tolerability of alectinib, I think that alectinib is going to move to the first-line setting. It is not FDA approved for the first-line indication, but I think based on these compelling data that we’ve seen from 2 trials, it’s going to be there in the first-line space. Efficacy and tolerability of brigatinib DR RAMALINGAM: Brigatinib is another interesting agent. And it’s approved in the second-line space after patients develop resistance to crizotinib. And what stood out to me about brigatinib is, the median PFS in the second-line space with brigatinib in these Phase II studies was approximately 15 months, which is way higher than what we’ve seen with all the other agents in the second-line space. And brigatinib also has activity in the brain, which alectinib and ceritinib do as well, but when you look at the median PFS results, it really stands out. Now, we're not allowed to compare across clinical trials, but we do have to make decisions based on data that we have in front of us. And I feel that brigatinib is a good agent in the second-line space. So it’s given at a dose of 90 mg. Give it for 1 week. The main adverse event is this pulmonary toxicity that can happen in about 5% of the patients. That declares within the first week. And if you don’t see any toxicity week 2, you go up to 180 mg of that drug. So that’s the approved regimen with brigatinib. Second-line therapeutic options for patients with ALK-rearranged NSCLC DR LOVE: So I’m curious what your second-line therapy is for a patient whose first-line therapy is crizotinib and alectinib, and even ceritinib. What would be your second-line therapy for all those patients? DR RAMALINGAM: So if a patient got first-line treatment with crizotinib, until now, we’ve been using alectinib or ceritinib. But since the approval of brigatinib, I’ve started using that agent as well. Now, as the landscape shifts, when patients get first-line alectinib therapy, the question is, what’s the appropriate second-line agent? We have some anecdotal reports of one drug, such as ceritinib or brigatinib, but we don’t know a whole lot. But what is promising in this space is a drug called lorlatinib, which, in a Phase II study, has shown activity in patients who develop resistance to 1 or 2 prior lines of ALK inhibitors. So I anticipate lorlatinib could be another agent. For now, for lack of anything better, I would try brigatinib in these patients, because brigatinib does have better activity against some of the resistance mutations that you see in ALK-positive patients that develop acquired resistance. DR LOVE: What do you see in terms of tolerability, particularly with ceritinib? I know there were concerns about GI toxicity, about the dose. How did that play out? DR RAMALINGAM: So ceritinib is approved at the dose of 750 mg once a day. And at that dose, we do see more GI toxicity. So I often, when I start ceritinib, start at a dose of 600 mg per day and do appropriate GI prophylaxis. And by giving it with food, you actually improve bioavailability and also reduce toxicity. So there’s some data that when you give ceritinib at 450 mg and have the patient take it with food, it’s much better tolerated. So 600 as a starting dose, if needed dose reduction to 450 is the approach we use in our clinic. Alectinib versus crizotinib for ALK-rearranged NSCLC DR LOVE: So I am curious. You gave an explanation in terms of osimertinib up front in terms of the T790. What about, for example, alectinib versus crizotinib? I’ve heard the term “alectinib is a better ALK inhibitor.” I’m not sure exactly where that comes from or how it plays out clinically. Again, there’s the issue of brain mets. Why is it that it seems to be better in the first line? DR RAMALINGAM: The explanation goes back to crizotinib was originally being developed as a MET inhibitor. And it also has ALK activity. So it’s not a selective ALK inhibitor. Now, once crizotinib was approved, the development of the drugs that are very specific and potent against ALK were developed. And that’s where ceritinib and alectinib come in. So they do have a more potent inhibition. If you look at preclinical models, just at the IHC 50, which is the amount of drug that’s needed to kill 50% of the tumor cells, alectinib and ceritinib and brigatinib do much better than crizotinib. So that sort of a potent ALK inhibitory effect could explain why they do better. And also, the fact that these agents have much better brain activity compared to crizotinib could account for the fact that you are suppressing development of brain metastases for a longer period with these agents. Case: A 64-year-old woman with concurrent ER/PR-negative, HER2-positive breast cancer and BRAF V600E-mutated adenocarcinoma of the lung DR LOVE: So I really want to hear about the 64-year-old lady. What an incredible case. DR RAMALINGAM: Yes. I mean, she’s a remarkable lady who comes 5 hours from Atlanta to me. And when she presented, she actually had developed the respiratory difficulty, significant dyspnea, worsening cough and found out to have significant involvement with her right lung and bilateral diffuse pulmonary infiltrates. And a workup resulted in the diagnosis of lung adenocarcinoma. But unfortunately for her, at the same time workup also showed a nodule in her right breast. And that was biopsied, and that was breast cancer. So she had 2 cancers. The good news was, she had good functional status. She didn’t have any comorbid medical conditions. So she was seeing me and one of my breast cancer colleagues and the breast cancer surgeon. And the team decided to give her 4 cycles of treatment with carboplatin and paclitaxel. Because she was clinically sick, we didn’t want to wait for any molecular testing to come back. We thought carboplatin/paclitaxel would affect both breast cancer and lung cancer, get her lung cancer under a little bit of control and then do the surgery for breast cancer. DR LOVE: What was the ER and HER2 status on the breast cancer? DR RAMALINGAM: She was ER/PR-negative for breast cancer. DR LOVE: And how about HER2? DR RAMALINGAM: Her HER2 was positive. DR LOVE: Wow! Interesting. So you gave her chemo — DR RAMALINGAM: We give her 4 cycles of chemo for lung cancer. DR LOVE: I mean — okay. DR RAMALINGAM: And she showed stabilization and some improvement in her symptoms, but radiographically there was not much of an improvement. So stable disease would be the best at that point. And she did go on to have surgery. So we had to interrupt treatment while she had the surgery. And came out of surgery. Fortunately for her, the breast cancer was Stage I. There was no nodal metastases, which we knew ahead of time that she did not have extensive breast disease and it was a small lesion. So the good news was, the breast cancer was out. But the bad news was, it had been 3 to 4 months since her last chemo. And she started getting worse in her — she had developed a pleural effusion. She had lymphangitic carcinomatosis involving the lungs. She had required a lot of oxygen. And by then, we had the results of her lung cancer molecular testing available. And we knew she had a BRAF V600E mutation. This was almost 14 months ago. And at that time, there was no approved therapy for a BRAF mutation. So we had to get the combination of dabrafenib and trametinib through the company as part of their patient support program. And the dramatic part is, within a few weeks of starting the treatment, she came off oxygen. Her breathing resolved — I mean, breathing status improved considerably. She was able to ambulate, go out of her house. And when we saw the scans 2 months later, we were truly amazed at the extent of response. She tolerated the treatment well. And now she is continuing to lead a normal life and continues to be on therapy, doing extremely well. Dabrafenib/trametinib for BRAF V600E mutation-positive metastatic NSCLC DR LOVE: So first of all, I’m sure she’s really glad she’s making that 5-hour trip. Sounds like it really made a difference. I’m not sure how many people in this country would be getting a BRAF mutation, although it certainly seems like it ought to be done, based on not just this but many other cases. I’m curious — of course, this combination has been used in melanoma for a while. It does have — we’ve heard about those toxicities for a while. What did you see in her? And what have you seen in general with this combination in lung cancer? DR RAMALINGAM: So the Phase II study that looked at this combination has recently been published. It showed that the response rate was quite impressive, close to 65%. And the median PFS was approximately 10 months, suggesting it’s in the same ballpark as what you would get with an EGFR-mutant patient with erlotinib or one of the TKIs. The tolerability for this combination is not a walk in the park, as one of the PIs told me from the study. Fortunately for my patient, we did not encounter any toxicities. But GI toxicities, fatigue, some rare incidents of ocular toxicity are all seen with the combination approach. So if a patient who’s on this combination develops eye symptoms, it’s important to get them to see an ophthalmologist right away. DR LOVE: What about fever? DR RAMALINGAM: Fever is also something that’s been noted in a subset of patients. She did not experience any of those things. And I think, given the extent of how bad her lung cancer was, it’s possible that she doesn’t even appreciate some of these symptoms. But fever is reported with the combination. DR LOVE: Really interesting story. I’m kind of curious. If she were to present to you today up front and you knew she had a BRAF mutation — of course, now you know she responds better. But just in general, how do you approach the symptomatic patient who has a BRAF mutation up front? DR RAMALINGAM: My approach would be to give them targeted therapy, the combination of dabrafenib and trametinib in the first-line setting, and use chemo in the second-line setting for those patients up front. So testing for BRAF is part of our panel now. And we test patients for EGFR, ALK, ROS1 and BRAF at a minimum in the front line. DR LOVE: And I’m sure that applies to patients with EGFR and ALK in terms of, again, the sick patient. But what do we know, though, about the time to response and the chance of early response, chemo versus targeted therapy? DR RAMALINGAM: In patients with these driver mutations, the responses happen very early. In fact, when you go back and look at the early studies of erlotinib and gefitinib, when you look at patient-reported outcomes of how early they can feel symptomatic benefit, you see in 1 to 2 weeks these patients start noticing improvement in their symptoms and overall sense of well-being. So I would say that these responses do occur quite early. In 2 to 3 weeks we see clinical improvement in patients with driver mutation who are given targeted therapies. So I wouldn’t hold back on targeted therapies under the notion that chemo, perhaps, may work sooner. Case: A 63-year-old man with metastatic squamous cell NSCLC and Type 2 diabetes receives carboplatin/nab paclitaxel followed by second-line atezolizumab DR LOVE: So I want to finish out in terms of the cases with your 63-year-old man. DR RAMALINGAM: Yes. So this is a patient who had comorbid conditions. He has Type 2 diabetes. He had presented with shortness of breath, an episode of pneumonia. He was hospitalized. And during that time, they found that he had a lung mass. It was biopsied. He had Stage IV disease based on metastatic disease to liver and bones. His histology was squamous cell. And his performance status, I would say, after he had recovered from the pneumonia, was 1. And, therefore, he was a candidate for combination chemotherapy, but when we looked at what are the options for him, we decided to go with carboplatin and nab paclitaxel for 2 reasons. One, of course, in squamous histology, this combination has resulted in a slightly higher response rate compared to carboplatin and paclitaxel. Number 2, with his Type 2 diabetes, not exposing him to steroids would be a good thing, because with nab paclitaxel we do not need steroid premedications. So that was the reason we decided to do it. He was also living not too far from our center, so he could come every week and get this treatment. So we started him on carboplatin and nab paclitaxel. He did experience a reduction in size of the tumor. He did not meet the criteria for a PR but definitely felt that it helped him. His improvement in symptoms were seen after the first cycle of therapy. So we treated him for 4 cycles. And then once we reached the point of maximal response, we kept him on nab paclitaxel maintenance for a few more months. And now the maintenance part is not something that’s approved or proven, but we felt that given his underlying disease burden, giving this as a maintenance would be beneficial. And so that was the rationale for this regimen for him. DR LOVE: And what’s his current situation? DR RAMALINGAM: So he did well on maintenance for another 5 months, and then he progressed. And more recently we’ve started him on a checkpoint inhibitor. DR LOVE: Which one? And what’s happened so far? DR RAMALINGAM: So he’s been started on atezolizumab. He has received 2 cycles of therapy. His symptoms are stable. We went back and looked at his PD-L1 expression, and his expression was approximately 20%. DR LOVE: Any tolerability issues with atezolizumab? DR RAMALINGAM: No. He’s actually tolerated it quite well. And he’s coming up for his scan in the next few months for the next one. Second-line immunotherapy options for patients with metastatic squamous cell NSCLC DR LOVE: Can you talk about how you generally approach second-line therapy of squamous in patients with low TPS scores and how you choose a checkpoint inhibitor and why you chose this, atezo, in this patient? DR RAMALINGAM: So in the second-line setting, if a patient has not received first-line checkpoint inhibition — obviously if you had a squamous patient with a greater than 50% expression, they would get first-line pembrolizumab. But let’s say this patient’s PD-L1 status was not at that point or they were treated with chemotherapy, for whatever reason, in the first-line setting. For those patients we, as you know, have 3 approved options: pembro, nivolumab and atezolizumab. Pembrolizumab is approved only if their PD-L1 expression is greater than 1% on label. And both nivo and atezo are approved regardless of the PD-L1 expression status. Now, the efficacy of all of these drugs in the second-line space seem very similar. If you look at hazard ratios from these Phase III trials for survival, it’s consistently around 0.72, 0.73. So we don’t think their efficacy is much different. The main differences boil down to how often you give these drugs. Nivo is given every 2 weeks. Pembro and atezo are given every 3 weeks. So for that reason, from a patient convenience standpoint, atezo and pembrolizumab have become now more go-to drugs in the second-line space in our practice. So we chose atezo because at the time we started him on therapy, we did not have the PD-L1 expression status. We didn’t want to wait for it. We said, “With atezo, we can just start him right away without having that information.” Necitumumab or docetaxel/ramucirumab as second-line treatment for advanced squamous cell NSCLC DR LOVE: Another potential consideration in him up front with having squamous cell would be necitumumab. Is that something that you thought about in this younger patient, 63 years old? Is that something you thought about? Do you ever utilize it? DR RAMALINGAM: So necitumumab is an approved agent in combination with platinum doublet in squamous histology. The PFS and overall survival results were modestly improved. So it is a consideration in this setting. We did not choose it for this patient because of some of the comorbid conditions that he had. And we were worried about toxicity of putting a 3-drug combination and, therefore, we gave him 2 drugs. But it’s certainly an option that's FDA approved and is available for patients who have good performance status. DR LOVE: So I guess statistically, the most likely situation for him is going to be to have disease progression. Hopefully he will respond on the atezo, but more likely he won’t. Assuming his performance status is still maintained, what do you think you’re going to be thinking about next? DR RAMALINGAM: So in third-line space, if the performance status is appropriate then docetaxel with or without ramucirumab is the treatment option we go to outside of a clinical trial. If the patient’s performance status is very good, I would give him the doce/ram combination. If they are not candidates for the combination, then I would just go with docetaxel monotherapy. DR LOVE: Any comments on ramucirumab in terms of tolerability issues? I’ve heard it’s kind of like bev. Is that what you find? DR RAMALINGAM: Yes, the profile appears to be similar. They have the class effect toxicity, such as hypertension/proteinuria and a small risk of bleeding. We also see some thrombotic phenomenon or arterial/venous thrombi in some of these patients. So you just want to select patients based on comorbid conditions and risk for developing clots or bleeding. And then if they are appropriate otherwise, ramucirumab is well tolerated. Using metaphors to explain molecular testing to patients with cancer DR LOVE: So I wanted to run through a few of the papers that you’ve been involved with and also some other papers I’m just kind of curious about, beginning — you had a couple of papers I found interesting, “Using Metaphors to Explain Molecular Testing to Cancer Patients.” Can you talk about this work and what you saw there? DR RAMALINGAM: Sure. So as we talk about genome testing and results of genomic testing and what it means to patients, we did a study in our institution where it was very clear that it was not easy for patients to understand. And there is a subset of patients who have a hard time understanding what a mutation is, what a driver mutation is. And, therefore, we felt that it’ll be helpful to develop some terminologies. And in the study, we had some investigators. We went to them and asked them, “How do you explain molecular testing to your patients?” And then we studied patients who were given those analogies and see which of those metaphors resonated more with them. And that’s what this paper was. It’s a small group of patients. But what we felt was, when you use certain examples — for example, I often use the situation of a driver on a bus. If you are in a bus and the bus is going down fast, the best way to stop it is by — the mutation is the driver in this case. The driver is responsible for the bus going fast. And the targeted therapies end up working at the level of the driver. So using those kinds of examples, we felt, resulted in a better understanding by the patients about what the whole genomic testing and treatment options, based on that, are about. DR LOVE: And from your own experience with your own patients, what do you find works best? DR RAMALINGAM: I use either the driver analogy, and I also use the switch on/switch off. I tell them, “Think about cancer cells as a room full of wires. And there is a switch to turn on and off. And the mutation in your case is the switch keeping the lights on. And by using a drug to switch off, you can actually turn the lights off and, therefore, the cancer cell could die.” Biology of resistance to TKIs in patients with EGFR mutations DR LOVE: I wanted to also ask you about your work on mechanisms of resistance to EGFR TKIs with people with EGFR tumor mutations. And maybe you can talk a little bit about what you’ve seen in the laboratory and also something — actually, it really wasn’t on my radar. I always think about EGFR T790. And recently I started to hear about MET alterations, even HER2 alterations that occur. Can you just provide an overview of what we’ve learned about the biology of EGFR resistance? DR RAMALINGAM: Sure. So when we treat patients with erlotinib or gefitinib, you see a few escape pathways. T790 is by far the most common resistance mechanism. MET amplification or MET pathway activation is seen in about 5% to 10% of the patients. And then there are other rare escape pathways, such as HER2 pathway. And you can even see an occasional BRAF mutation. So these are all signaling through various growth factor pathways. Now, what agent you give in the first line will dictate what potential mechanisms of resistance are going to develop. So in preclinical models, when we study EGFR-mutated cells with a drug like osimertinib, what happens is, these cells do not develop T790, because you’re blocking that pathway. And when you look at what remains, the MET pathway becomes an even more important escape mechanism in these patients. It may be in the order of 20% to 30% where you see MET pathway activation. And that’s where we’re now developing combination approaches, combining a third-generation EGFR inhibitor with, say, a MEK inhibitor. And at least in the preclinical paper that we just published, there seems to be a significant tumor inhibition with that combination approach. So we’re moving towards getting that to the clinic. ECOG-ACRIN 2511: A Phase I/II study of cisplatin and etoposide with or without the PARP inhibitor veliparib as front-line therapy for extensive-stage SCLC DR LOVE: I want to ask you also about some of the work you’ve done in small cell cancer. And I was interested to see that you’re actually looking at PARP inhibitors. Actually, PARP inhibitors obviously are now approved in ovarian cancer. Really exciting data came out in breast cancer at ASCO. But I see here that you’ve been involved with work with the drug veliparib, which is not approved, but combining it with chemo. What do we know about that in small cell? DR RAMALINGAM: Sure. So veliparib as a PARP inhibitor basically inhibits DNA repair. So when you give a platinum compound to a patient, you are trying to cause DNA damage. And that results in apoptosis and cell death. And that’s how tumors respond to platinum. But the tumors have in-built repair mechanisms. So when you give a drug like veliparib, it blocks DNA repair and makes platinum work better. And that’s the rationale behind these combination approaches. What you’re referring to is a trial that we reported at ASCO from ECOG, which was a randomized Phase II trial of platinum/etoposide with veliparib versus platinum/etoposide with placebo as first-line therapy for patients with small cell lung cancer. This was a randomized Phase II to look for a signal. And what we found was, the PFS was improved in patients who got veliparib plus chemotherapy, to a modest extent. Now, the hazard ratio is approximately 0.75. And the question is, can we identify biomarkers to select patients who benefit from this? Now, there have been some exciting advances in the preclinical understanding of PARP inhibition. And there are some exciting biomarkers such as SLFN11 that are being proposed as putative biomarkers for PARP inhibition. So we’re now looking at these samples to see what their SLFN11 expression is and see if this responds. The HRD score is the other score that has been linked with PARP inhibition. So we’ll be looking at HRD score as a potential predictor as well and help finalize plans on further development of this strategy. Rova-T in DLL3-positive SCLC DR LOVE: What else is going on with small cell? I know there’s an antibody-drug conjugate, rova-T. It’s hard to pronounce. What do we know about that therapy? DR RAMALINGAM: Rova-T is an antibody-drug conjugate. The antibody itself targets a protein called delta-like ligand 3, which is seen in about 70% of small cells. And this drug conjugate has a toxin attached to an antibody, so essentially you deliver the drug into the cancer cell by using this targeted approach. And the interesting part has been, they’ve seen response rates of about 40% in patients with small cell lung cancer after they have received standard platinum doublets and even topotecan. This is in patients who are selected for the biomarker DLL3 expression. And this drug is given intravenously. They give every 6 weeks. You just give 2 doses of the drug. And you see responses within the first several weeks. So this is an exciting development in the second-line small cell space. The drug is in advanced-stage clinical trials. There are certain toxicities that we’ve seen with this drug, such as recurrent effusions, some skin toxicities that need to be managed proactively. But it is an interesting agent and hopefully will enter the treatment landscape for small cell lung cancer. DR LOVE: Have you used it yourself in a trial? DR RAMALINGAM: Yes. We have used rova-T in clinical trials for a few patients in our trials. DR LOVE: Have you seen any patients with clinically meaningful responses? DR RAMALINGAM: We have seen at least 1 patient in our small cohort that has seen a good response to rova-T. This is a patient with liver metastases and extensive chest disease who showed a good response to treatment and has now completed therapy and is on follow-up. DR LOVE: So when you think about the other antibody-drug conjugates out there — T-DM1 in breast cancer, brentuximab vedotin in lymphomas — they seem generally well tolerated. I mean, there’s a chemotherapy moiety in there, but you don’t really see chemotherapy toxicity. Is that the case with this agent? DR RAMALINGAM: So the toxicity for the ADCs depends on where else your target protein is expressed. So if your protein is expressed in certain normal cells, then you’re going to see toxicity related to that. So whether there is something to do with DLL3 in the serosal surfaces is not clear to me. But certainly we see ascites and pleural effusions in these patients. DR LOVE: In those few patients that you’ve treated, did any of them have significant clinical problems with effusions? DR RAMALINGAM: We have had at least 1 patient who ran into significant effusion problems and another patient who had some skin issues related to photosensitivity. Activity of combined anti-PD-1 and anti-CTLA-4 inhibitors in SCLC DR LOVE: The other thing that seems to be out on the table with small cell is checkpoint inhibitors, but particularly I’m hearing a lot about combinations of anti-CTLA-4 and anti-PD-1, nivo/ipi, for example. What about checkpoint inhibitors in small cell? DR RAMALINGAM: Still under investigation, but there are some studies that have read out. And what we’ve seen in PD-1 inhibition alone seems to have a response rate of about 10% to 11%. And when you combine CTLA-4 inhibition with a PD-1 inhibition, the response rates tend to be about 25%. So the combination approach is now being studied in randomized trials, but it’s also already been put in the NCCN panel for salvage therapy of small cell lung cancer and can be used. And I take it that they decided to do that based on the fact that there isn’t much to do for these patients at this time. So I think that the data with the combination of CTLA-4 and PD-1 are promising. I have 1 patient that I’ve treated who’s been on this approach for almost 2 years with no evidence of disease. So I do think that this works in at least a subset of patients. Toxicity comparison of anti-PD-1 and anti-PD-L1 antibodies DR LOVE: A couple of final questions from some of the work you’ve done on checkpoint inhibitors. One, you had a poster looking at the toxicity profile of PD-1 versus PD-L1 agents. I’ve heard that debated ever since these things came into play. What did you all see? And in general, what has been seen? Initially, I was hearing things like maybe there’s going to be less pulmonary toxicity with anti-PD-L1. What do we know about this? DR RAMALINGAM: So the initial premise was that PD-1 inhibitors will have less autoimmune toxicities compared to PD-1 inhibitors. And that’s what prompted our investigation. So we did an analysis of close to 6,000 patients that have been treated across PD-1/PD-L1 inhibitors in lung cancer and look at what were the adverse events of patients treated with PD-1? What were the adverse events of PD-L1? What were the efficacy results? Are there any key differences of note? And this was presented in the World Conference on Lung Cancer by Dr Pillai from our group. And she showed that there weren’t many differences in toxicity between these 2 classes of drugs. They were very similar in terms of adverse events. And this paper should be coming out in Cancer journal very soon. So this leads us to believe that these agents are very similar in terms of efficacy and toxicity as of now. Cardiac allograft rejection as a complication of PD-1 checkpoint blockade DR LOVE: So I have to ask you about this case report you have of cardiac allograft rejection in a patient getting a PD-1 agent. DR RAMALINGAM: That was a very unfortunate situation where a patient who had received a heart transplant at Emory 20 years earlier and was doing very well about a year and a half ago had developed a skin cancer. And he was treated at the West Coast, which is where he was living. And they had given him platinum doublet, which did not result in a significant response. And subsequently they decided to give him a checkpoint inhibitor. And he got 1 dose of a checkpoint inhibitor and within a few days developed significant heart failure. And he was brought to our ICU, and when we saw him on consultation he was in full heart failure. And it was clear that that was prompted by the immune checkpoint inhibitor. So it is important that patients who have received organ transplants are not treated with checkpoint inhibition, because you are going to start facilitating the rejection process. These have to be very carefully managed and, at least for the time being, until we know better we should not be giving them checkpoint inhibition. DR LOVE: Did you say the patient had a skin cancer? DR RAMALINGAM: Yes, cutaneous squamous cell cancer. DR LOVE: Hmm. Interesting. And what happened with the patient? DR RAMALINGAM: We managed to improve the patient’s heart failure. And he was able to be discharged home, but by the time he got home, he was on hospice. And he died a few weeks later. DR LOVE: From heart failure? DR RAMALINGAM: From heart failure and the consequences of his cancer as well. The heart failure improved, but his cancer was progressive. DR LOVE: Wow! Actually when I saw that I was struck, because I had heard anecdotally of patients with kidney transplants who got checkpoint inhibitors who, quote, did well. Safety of immune checkpoint inhibitors for patients who have undergone organ transplants DR LOVE: What do we know in general about people who’ve had organ transplants who get checkpoint inhibitors? DR RAMALINGAM: I think there’s at least 1 study, the Hopkins group was looking at transplant patients and specifically in kidney setting, they have reported that it was safe to give them. My approach is, in very highly controlled situations it may be okay on a clinical trial to study how these patients do. But for our audience in the community who are taking care of these patients, it would be better not to treat these patients with checkpoint inhibition. Rationale for combining the anti-CD38 antibody daratumumab with immune checkpoint inhibitors for NSCLC DR LOVE: It’s funny, because somebody came up to me at ASCO and handed me an abstract and said, “Have you ever heard of daratumumab in lung cancer?” We were about to do a myeloma session. I asked the myeloma people. They didn’t know, but they were hypothesizing that it was based on the immune effects. And here I see that that’s your study. Can you talk about what we know about daratumumab? I see you’re looking at it with a checkpoint inhibitor. DR RAMALINGAM: Yes. So daratumumab is, as you know, approved for myeloma. It’s a CD38 inhibitor. And there are preclinical rationale that suggest that when you combine CD38 inhibition with a PD-1 or PD-L1 inhibition, there is considerable enhancement of antitumor immune response. So there is a randomized Phase II study that’s presently ongoing that combines atezolizumab with daratumumab. It’s a small signal-finding Phase II study. We had a lead-in cohort of patients where we just looked at the safety of the combination, and then we moved on to the randomized Phase II part. So it’s still early. The safety so far has been good. And we are anxiously awaiting the results of the trial. DR LOVE: I’m curious what you’ve been seeing with infusion reaction, because the myeloma people really have struggled with that. DR RAMALINGAM: That is correct. The infusion reactions with daratumumab require a prolonged infusion stay or prolonged stay in the infusion center. So those things are not much different when you use it in lung cancer patients, but they can be managed. It’s usually in the first or the second time. DR LOVE: Have you thought about trying the subQ? That’s out there now, too. DR RAMALINGAM: At this point, we’re still using the IV version. And when the signal in terms of PFS and efficacy results of the Phase II are read out, we’ll make a decision about whether that’s a better approach to do. First-line pembrolizumab compared to chemotherapy for patients with PD-L1-positive NSCLC DR LOVE: Let’s move on now and talk about the use of checkpoint inhibitors, specifically in first-line metastatic disease where we’ve seen some new approvals and data sets, beginning with the use of pembrolizumab. DR RAMALINGAM: So pembrolizumab is now approved for first-line therapy of non-small cell lung cancer. And this is in biomarker-selected patients as monotherapy. And this is a Phase III trial but in all patients with advanced-stage non-small cell lung cancer. And they selected patients whose PD-L1 expression in the tumor was greater than 50%. And this is the immunohistochemistry test that uses a specific antibody. And this level of positivity is seen in approximately 25% to 30% of patients with non-small cell lung cancer. And they randomized patients who were positive for PD-L1 at this 50% or greater threshold to get the standard platinum doublet chemotherapy or pembrolizumab as monotherapy. And the primary endpoint of this trial was to show an improvement in progression-free survival. And what you see is a significant improvement in progression-free survival, from 6 months to 10.3 months, favoring the use of pembro as monotherapy and a significant improvement in overall survival as well. The hazard ratio for PFS was approximately 0.5. And the hazard ratio for overall survival was approximately 0.6, resulting in the first positive trial in the front-line setting that showed improvement in survival with a novel agent over standard platinum doublet chemotherapy. So this is now an approved option in the front-line setting for patients with this high PD-L1 expression. Duration of response to first-line pembrolizumab DR LOVE: I’m kind of curious, when you start out first-line pembrolizumab in this identical situation, we see the response rate is about 45%. If a patient says to you, “What’s the chance that I’m still going to be without disease progression in a couple of years,” what would you say? DR RAMALINGAM: So that is the most intriguing part about immunotherapy, which is, the patients who respond have a greater median duration of response. They form the tail of the curve, as we like to call it, which is, can we increase the longevity of patient survival? And with immunotherapy, the best responders seem to hold onto their response a lot longer. So when you project some of the early studies that we have, the curves from those early studies, the 3-year survival rate with immune checkpoint inhibition appears to be on the order of 30% to 35%. So that’s a much higher number than what we would expect to see with chemotherapy even among some of the best responders. DR LOVE: So just a related question, which is, for how long do you continue treatment in a patient who’s having a great response? DR RAMALINGAM: The standard protocol on the clinical trials has been to give it for a defined duration of either 2 years or 1 year for certain agents. So for pembro trials at our institution, they have required that the patients receive them for 2 years. So I think of this 2-year as the de facto duration of therapy at this point. Clearly, this is an important question as to what is the optimal duration of therapy that’s required. And there are studies being planned to address this. So for now, if a patient is tolerating it well and does not have any toxicities related to the autoimmune adverse events, then I continue it at least for a period of 2 years. First-line carboplatin/pemetrexed/pembrolizumab for advanced nonsquamous NSCLC DR LOVE: So, of course, the recent approval of the combination of pembrolizumab with carboplatin/pemetrexed brought a very interesting dynamic into the situation. Maybe you can talk a little bit about the data there. DR RAMALINGAM: Absolutely. The approval, as you know, was based on a randomized Phase II trial. This is an accelerated approval for pembro in combination with chemo for nonsquamous patients in the front-line setting. This trial is now published. It enrolled about 120 patients. These were newly diagnosed patients with Stage IV nonsquamous histology. And they were randomized to treatment with the doublet regimen, carboplatin and pemetrexed or carboplatin/pemetrexed with pembrolizumab. And for patients who get the 3 drugs, they were given the option of maintenance therapy with either pem or pem plus pembro or both. And the majority of the patients, in fact, got both agents as maintenance therapy. The primary endpoint actually was response rate. There’s a significant improvement in response rate for combining chemotherapy with pembrolizumab. And the median PFS was also improved. The median overall survival has not been reached for both groups. And the curves at this point don’t look too different in terms of overall survival but a clear separation for median progression-free survival. So this is the basis on which the drug was approved in combination with chemo. Keep in mind that they did not use PD-L1 expression as a selection criteria, so this trial included patients with all kinds of PD-L1 expression, starting from none to high expression. They also noted that the outcomes based on PD-L1 expression were not consistent across groups. For instance, when you look at the high expressers, greater than 50%, you see that the chemo plus pembro group did much better in terms of response rate compared to chemo alone. And for the overall group as well, we saw it was better. But when you look at the 1% to 49% expressers, the results were somewhat conflicting. The chemo plus pembro group didn’t do quite as well even as compared to chemo. But the problem is, you’re looking at a small trial, in 60 patients in each arm, and when you break them down into smaller and smaller subgroups, what you’re left with are statistically nonsignificant, wide confidence intervals type of results. So at this point, this trial is a first step, in my mind, towards integrating chemo with pembrolizumab. There are Phase III trials that have been either completed or are ongoing, combining chemo with pembro, which will have much larger patient numbers and will hopefully help us determine which subset of patients should get, say, a checkpoint inhibitor alone, which group of patients should get chemo plus a checkpoint inhibitor and whether there is a group where the combination doesn’t work and we should just give chemo. So for my practice at this point, with a greater than 50% group, I’m still giving them pembro as monotherapy. For the less than 50% group, it’s a conversation with the patient about chemo alone versus chemo plus pembro in the nonsquamous histology. Choosing between pembrolizumab monotherapy and pembrolizumab with chemotherapy for patients with advanced NSCLC DR LOVE: So we’ve surveyed investigators about this issue. And it seems like in the greater than 50%, most, but not all, the majority are doing what you do in terms of single-agent pembrolizumab, although maybe 25% to 30% are using the combination. What about the patient who’s sick and needs a response, they have a lot of tumor? You do see here on this slide, at least, that the response rate seems higher. Would you be more likely to consider the combination in that situation? DR RAMALINGAM: Yes, absolutely. So there are some patients, as you said, where the clinical situation is such where you may just have 1 shot at giving them an effective therapy. If you don’t turn them around, there may be no second-line therapy. So for those patients, clearly, giving chemo plus pembro, which is associated with a higher response rate, makes perfect sense. And these are symptomatic patients with larger disease burden, rapidly declining performance status. Those are early indicators, to me, where I would use the combination. Now, when I talk about “declining performance status,” we still want them to be at a point where they can receive these therapies, not somebody who’s got a PS 3 or 4. These trials were done in patients with a good performance status. DR LOVE: So, of course, when we talk about chemo plus a checkpoint inhibitor, we’re talking about nonsquamous. What do we know about chemo plus checkpoint inhibitor in squamous disease? Is there any reason to think it wouldn’t kind of play out in a similar way? DR RAMALINGAM: Not that we have — let me take a step back. If you look at all the trials done with checkpoint inhibitors in lung cancer as monotherapy, the results in squamous and nonsquamous have both been uniformly impressive. So I would not expect there to be a big difference between chemo plus checkpoint inhibition in squamous versus nonsquamous. And those trials are underway or have been completed, so we should get those results in the very near future to see whether that combination can also be used in squamous subsets. Molecular testing algorithm for patients with nonsquamous NSCLC DR LOVE: Can you talk about your overall algorithm for the management of patients with metastatic non-small cell? DR RAMALINGAM: When I have a nonsquamous patient, molecular testing includes not just EGFR, ALK and ROS1, I also test them for BRAF, because there is an approved therapy. So for nonsquamous patients, my initial testing panel includes at least those 4 genetic abnormalities for which there are FDA-approved treatment options. DR LOVE: Could I just throw in MET exon 14? Because we’re seeing more investigators talk about using targeted therapy up front there. What about that? DR RAMALINGAM: That is another group where we have a promising ability to treat those patients with MET inhibitors. I didn’t put that there because it’s not an FDA-approved option at this point. But certainly in our institution, we do a wider panel. That includes MET, HER2 and other additional abnormalities. But MET exon 14 is clearly a mutation which is seen in about 3% to 4% of the patients with lung cancer. It’s higher in sarcomatoid histology, in the mixed sarcomatoid tumors of the lung. And for those patients, giving a therapeutic trial of a MET inhibitor, like crizotinib or cabozantinib, once they’ve gone through the platinum doublet would be a reasonable approach. DR LOVE: So technically the approval — I believe checkpoint inhibitors, or pembrolizumab first line — states specifically that it’s ALK- and EGFR-negative. Biologic background of the activity of immune checkpoint inhibitors in patients with NSCLC and driver mutations DR LOVE: But I am curious, though, about the issue of where checkpoint inhibitors fit in in patients with a driver mutation, particularly the common ones, EGFR, ALK, ROS1 and these others that you mentioned. Can you comment a little bit about the biology of this situation? DR RAMALINGAM: So what we’ve learned so far is that the higher the mutation burden in a tumor, the greater the likelihood of the immune system recognizing some of these mutated neoantigens. So mutation burden has been linked with sensitivity to immune checkpoint inhibition. Now, mutation burden tends to be higher in patients with a smoking history because of the repeated carcinogen exposures. And non-small cell lung cancer is among the tumors with the highest mutation burden. Now, when we look at EGFR and ALK tumors, we know that these tumors are seen more commonly in patients with minimal smoking history. And these patients, while they have a dominant driver mutation, their overall mutation burden seems to be much lesser compared to smoking-related cancer patients’ mutation burden. We’ve also seen some reports that the PD-L1 expression in these tumors are lower. And they may not have the right immune microenvironment to respond to immune checkpoint inhibition. So at this point, what we can say is, for patients with an EGFR, ALK or ROS1 or BRAF, targeted therapy seems to result in better outcomes compared to trying immune checkpoint inhibition, even though those comparative trials have not been done at this point. What we know is from trials that compared chemo to checkpoint inhibition in the salvage therapy setting for non-small cell lung cancer. The patients with EGFR mutation actually fared better if they had received chemotherapy. And the hazard ratio across these trials was approximately 1.05 in favor of chemo. And this was based on a meta-analysis that was reported recently. So in my practice, I use the targeted therapies for patients with EGFR, ALK and ROS first. And if they have the option of a second-line targeted therapy, I give them a second-line targeted therapy like osimertinib in EGFR patients, or alectinib or brigatinib and ceritinib in ALK patients. And then I go to a platinum doublet. Only then I would give them a checkpoint inhibitor outside of a clinical trial. DR LOVE: Could you just go back through this? Is there anything — specifically you talked about mutational load. Anything else you want to point out in this slide to maybe better explain the biology? DR RAMALINGAM: So what you see in this slide is a diagram that was put together by Katie Politi and Scott Gettinger from Yale in an editorial they wrote in Clinical Cancer Research, where they talked about potential differences between an EGFR-mutated untreated tumor versus a tumor that’s been treated with an EGFR TKI. And even there you start seeing changes in the tumor immune microenvironment as the tumor develops acquired resistance. And this can also be said for the ALK-positive situation. So this tells us that there is a very dynamic process under play in patients whose tumors harbor these mutations. And it’s not the same as what it was when you saw them before they got any treatment, to the point where you see them after they have developed acquired resistance. And this has various implications on how antigens are presented, antigens are recognized, the immune microenvironment, including T regulatory cells. And, therefore, there is a lot more left to be learned at this point. It’s a bit too early in the game to see how best to combine immune checkpoint inhibition with targeted agents for patients with treatable driver mutations. DR LOVE: But at a more macro level, these biologic changes that are occurring as patients receive TKIs for target lesion or for mutations, does this make them more susceptible to immunotherapy? DR RAMALINGAM: One would make the case that as they go through acquired resistance and acquire more mutations, you might see a higher likelihood of response to immune checkpoint inhibition. But contrary to that, some of the studies that have been done so far have been quite slow to develop further because of some toxicity concerns that we’ve seen when you combine a checkpoint inhibitor with a PD-1 or a PD-L1 antibody. So the initial experience has been somewhat mixed. And I would say as these trials proceed further, we’ll see what’s happening in the clinic to our patients. Meta-analysis comparing the efficacy of immune checkpoint inhibitors to that of chemotherapy in patients with EGFR wild-type versus mutated NSCLC DR LOVE: Okay. And can you go through this graph, exactly? What it looks at. DR RAMALINGAM: Sure. This was a meta-analysis that was done to look at how the checkpoint inhibition strategy compared to chemotherapy alone based on whether a patient had EGFR mutation or wild-type EGFR. Now, we all know that there have been multiple randomized trials with nivo, pembro and atezo that compared these agents to docetaxel in the second-line setting. And these Phase III trials all consistently showed that use of a checkpoint inhibition was associated with favorable survival compared to docetaxel alone. And this study basically pulls all those data together to see how many EGFR-mutated patients were included in these Phase III trials and how did their outcomes compare to patients with EGFR wild-type tumors. So on the top half of the slide, what you see are outcomes for patients with EGFR wild-type tumor. And there you can see that the hazard ratios consistently favor the use of a checkpoint inhibitor. And the cumulative hazard ratio is somewhere in the ballpark of 0.7. When you look at the lower half, which specifically looks at EGFR-mutated tumor patients enrolled in these trials, you see that the hazard ratios are, on the right, favoring docetaxel. And the composite hazard ratio was 1.05, suggesting that for these patients with EGFR mutation, chemotherapy did better than immune checkpoint inhibition. Now, we don’t have information on ALK on these patients, because those were a relatively small number of patients. But this goes on to say that it’s not a given that if it works for everybody, it should work for EGFR-mutated patients. DR LOVE: We should note that subsequent to recording these discussions, results from the Phase III FLAURA study were presented at the ESMO meeting demonstrating that osimertinib significantly improved progression-free survival compared to standard-of-care EGFR TKIs for patients with EGFR-mutated metastatic lung cancer. Results from the Phase III PACIFIC study were also presented at the ESMO meeting and published in The New England Journal of Medicine demonstrating that durvalumab significantly improved progression-free survival compared to placebo for patients with Stage III non-small cell lung cancer after chemoradiation. In addition, alectinib was approved for first-line therapy in patients with ALK-rearranged metastatic non-small cell lung cancer. |