DR SOCINSKI: This gentleman is a 64-year-old guy, presents with actually left leg pain. He has a 50-pack-year history of smoking but he had quit 11 years earlier.
There was a lytic lesion in the left iliac bone and, at the same time, chest x-ray, then shortly after that a chest CT showed a right upper-lobe 4-cm mass with some associated mediastinal adenopathy.
He actually got biopsied in an outside institution, as well as a bronch. Both of them returned squamous carcinoma of the lung consistent with a lung primary.
So I treated him with carboplatin/paclitaxel for 4 cycles. He was having a little bit of chest pain there. That got better. By RECIST criteria he would have almost had a PR. He had a nice reduction in tumor after 4 cycles, actually felt great, really didn’t have a whole lot of toxicity associated with that. We met after his fourth cycle. He was very happy to take a break from therapy. We talked about the pros and cons of his maintenance options, either — let me ask you this. After 4 cycles he’s better. Would you have considered maintenance in him, or what would you have thought of in a squamous carcinoma patient?
DR BELANI: That’s a very good question, actually. And at this present point in time, there is no approved treatment for squamous cell carcinoma except erlotinib, which is approved in the maintenance setting based on the SATURN trial. So that is the option, which is an approved option. But whether the benefit is there, the benefit is modest at best in patients who are EGFR wild type. So I think we wanted to ask this question with weekly paclitaxel after carboplatin and paclitaxel, which he got. And we had done the trial with weekly paclitaxel and carboplatin followed by maintenance paclitaxel. And if we look back to the 3-arm study where we gave maintenance paclitaxel versus nothing, there was a benefit, but the trials were not designed to look at maintenance paclitaxel. So I would have considered paclitaxel to maintain his disease, because he was having a RECIST response.
DR LOVE: So you would have actually considered it in him?
DR BELANI: I would have considered it based on the data that we have generated from prior studies, though it is not approved, because it seems like he did not want to continue the combination regimen, and with the maintenance data out there in nonsquamous non-small cell lung cancer.
DR LOVE: How was he tolerating it? I mean, could he have continued with both? So would you consider keeping it going?
DR BELANI: I usually give 4 cycles of chemotherapy, and I usually stop and I take them to a single agent thereafter.
DR LOVE: But again, you take them to a single agent, that is, what I’d call maintenance in squamous cell?
DR BELANI: In squamous cell, if they are having a response. If there’s a continued response, the data says to add 2 more cycles of chemotherapy. So if Bruce Johnson were sitting here, he would say, “Add carboplatin and paclitaxel, 2 more cycles.”
DR SOCINSKI: And there’s a great wealth of data supporting that strategy.
DR LOVE: So just to be clear again, when he even said, “Would you consider maintenance,” I’m like, “What? What are we talking about,” kind of — I was flashing on maintenance rituximab in diffuse large B-cell lymphoma, which kind of got thrown out. And I thought it was kind of not on the table with squamous cell. John, is it? Do you agree with Chandra?
DR HEYMACH: Yes. I mean, I think all those options. I guess one other to put out there — and I don’t end up doing this, but there is data supporting it, is from the French Intergroup study, the IFCT. And they actually — there was a maintenance erlotinib arm and there was a maintenance gemcitabine arm. And it was positive for PFS with an OS trend that was pretty strong. It was a little underpowered for OS.
DR LOVE: With what? With gem?
DR HEYMACH: With maintenance gemcitabine. It was squamous specific. It was, I think, all comers. Now, with that said, I don’t end up doing that, typically. I mean, I think I’d just follow the options that Chandra said, but there’s data to consider.
DR LOVE: I mean, he brought up the option of continuing paclitaxel. Do you ever do that?
DR HEYMACH: I don’t continue single-agent paclitaxel. I would bet if you did the properly powered study, like JMEN in squams, it would probably be positive, the same way the Fidias study was — the Fidias study, of course, was the docetaxel maintenance. And that was strongly trending toward positive. It was underpowered, but it was a pretty strong effect.
DR BELANI: I would say it is not an approved way of doing it. And we cannot really base it on the data that is available. But based on the data generated from 2 large clinical trials, there is a role if the patient continues to respond. It is better to continue the treatment, if it is not —
DR LOVE: Yes, I’m not — a lot of times, we’re not trying to find what the data is. We’re just trying to find what you do based on everything that you can integrate. I’m just saying that I haven’t heard a whole lot about maintenance in squamous cell. Just real quick, to see whether or not you’re an outlier or I missed a trend here, maintenance in squamous cell?
DR SPIGEL: Yes. I agree with these guys that I stop after 4cycles. I do not switch to erlotinib.
DR LOVE: Let’s go back for the next step in this patient’s course.
DR SOCINSKI: Okay. So he actually had a good 3 months on nothing. He came back over the summer and had a bit more right hip pain. Actually, without telling me, he went back to see his radiation oncologist, and guess what happened?
DR BELANI: He got radiation.
DR SOCINSKI: He got radiation to the right hip. And it didn’t help him. And then when I saw him after that, he had developed kind of recurrence of his right upper chest pain and, in fact, on CT that day he had progression. It looked like this was starting to invade the chest wall, and that’s what was the basis of his pain. So that’s kind of where he was. Other than that symptom — and I can’t remember — I don’t even think he was taking narcotics for this pain. He was taking Tylenol or nonsteroidals. He’s otherwise fit.
DR BELANI: Performance status good?
DR SOCINSKI: Yes.
DR BELANI: No symptoms.
DR SOCINSKI: Yes. PS 0. Other than chest pain. So his performance status was pretty good. He had no new sites of disease. He had bone only and this enlarging right upper lobe.
DR LOVE: So before I ask Chandra what he would do outside a trial setting, I’ve got to say that I found it really interesting that, of course, you discussed clinical trial options and the 2 clinical trials you brought up, which I guess — did he end up on a trial?
DR SOCINSKI: Yes.
DR LOVE: Okay. One was nanoparticle docetaxel, which I didn’t know existed, and erlotinib and rilotumumab.
DR LOVE: What is nanoparticle docetaxel?
DR SOCINSKI: So, the nanoparticle docetaxel thing is a compound I refer to as BIND-014. Dan Van Hoff did a Phase I study, presented it at AACR. And much like nab paclitaxel is a different formulation and, presumably, by putting these compounds, drugs that are water insoluble, into nanoparticles, they can get to the tumor where they need to be preferentially. And this is what this compound looks like.
DR LOVE: And no corticosteroids with this?
DR SOCINSKI: Correct.
DR LOVE: Interesting.
So what I want to ask Chandra is: How would you think through this situation outside a protocol setting? What are the options you might be thinking about? And would you consider a VeriStrat assay?
DR BELANI: So outside of a clinical trial I guess the 2 options are to give him docetaxel as a single agent, which is approved in the second-line setting, or to give him erlotinib based on BR.21. And if you are going to give him erlotinib, which I think that the activity, again, is modest at best, because if BR.21 were to be done today, it would probably be a negative trial in this subgroup of patients who are EGFR wild type and squamous cell carcinoma.
So would I do a VeriStrat? I have not done VeriStrat. I’m not convinced of the idea. If you have to give erlotinib, you might as well give it. VeriStrat is only for that group of patients which is VeriStrat poor, where you should not give erlotinib and you would give chemotherapy.
DR LOVE: So just, again, I know that an option that you all think about are clinical trials, and I’m going to ask Mark to talk about the paper that was presented on VeriStrat at ASCO, which I thought was very interesting. But just before he gets into what happens to what happened to this patient and the data that was presented, I’ll just go to John and ask you: What would you likely do in this situation?
DR HEYMACH: Regarding the VeriStrat, I haven’t used it routinely and, as Chandra raised, there is benefit for squamous cell cancer with erlotinib. It’s small, so usually I’d say if somebody can get chemotherapy, I’d give them the next-line chemotherapy. We usually favor or I’m typically starting with gemcitabine, because we have a lot of taxane-based or docetaxel-based second-line studies, so we’d usually go gem/platinum first, docetaxel second, and then if we don’t have a clinical study, erlotinib third, would be our typical route.
DR LOVE: So Ram, are you using VeriStrat? And how would you approach this man?
DR GOVINDAN: The answer to your first question is no. VeriStrat, I don’t use it. And in patients who don’t have EGFR mutation I tend to use chemotherapy in the second-line setting and, rarely, erlotinib in the third-line setting.
DR LOVE: If you believed — and, again, we’ll let Mark present the data. But if you believe that VeriStrat can identify a group of patients who do as well with erlotinib as with, let’s say, docetaxel, if you believe that, do you think it would be helpful? Do you think that, for example, erlotinib would be better tolerated?
DR GOVINDAN: No. The problem with all these tests is that we need to know exactly what we are measuring. With proteomics, it’s evolving right now. And I’m sure we’ll know a little bit more in the future. The question really is, are we going to base a test that we are convinced that it is the right way to identify patients? As Chandra said, if you have to give erlotinib, you may as well give it a free trial for a few months. So I also want to point out that erlotinib is not a benign drug. It has toxicities. It has side effects that are not that different from some chemotherapy drugs. Arguably, docetaxel has higher rates of myelosuppression. But I don’t feel that the test that we are not convinced of — or if you don’t — if you’re not convinced of the utility of a test, I would use that primarily to decide who should get what.
DR LOVE: What happened with this patient?
DR SOCINSKI: I did get a VeriStrat test on him. He turned out to be good. My view of the VeriStrat is that it clearly has both a prognostic and predictive aspect of it. I was really looking for a reason not to use erlotinib in him, but that would have been the case in the poor. I’ve always been struck by the BR.21 curves. I think most of us would agree that even though erlotinib has utility in the wild-type population, there are a fair amount of patients who really get no benefit in the wild-type population. And we’ve never been able to identify by clinical and molecular factors a group that clearly doesn’t have benefit, say for the whole KRAS. And KRAS is controversial, because that’s mostly a response rate, not necessarily a survival issue in this setting. And so I think VeriStrat has clearly identified — it’s interesting the way this has evolved. I think often we forget that this test was actually developed independently, by some fairly credible lung cancer investigators, Dr Carbone at Vanderbilt and — and the Colorado group. And this serum proteomic test identifies a binary result, for better or for worse. We call them good and poor. I’m not quite sure how that language started out. But the breakdown is about 70-30. Seventy percent of the patients end up being good. Thirty percent end up being poor.
One of the problems with the development of this test early on is that all the data was relatively small subsets looked at retrospectively. And then they were able to look at the BR.21 database retrospectively. And I think the bottom line from all of that retrospective data, including BR-21, is, we were all convinced that this test really had a prognostic aspect to it. But the question was, how predictive was it one way or the other? And that kind of led to the PROSE trial that was presented at ASCO this year, a trial done in Italy in the second-line setting.
Patients were tested for VeriStrat results. Seventy percent of them were good. Thirty percent poor. That was blinded to the patients and the investigators, and then they were randomized to that second-line choice that all of us have, and it’s very common in the community: Do you go the cytotoxic chemotherapy route, or do you go the EGFR TKI route? In this case, PROSE randomization was to a choice of docetaxel or pemetrexed versus erlotinib as the TKI.
And what they showed actually in that trial was really 2 things. If you look at the VeriStrat good population, their median survival in the second line was about 11 months. And there was no difference whether you got cytotoxic chemotherapy or whether you got erlotinib. If you were in the 30 percent that were VeriStrat poor, your median survival on erlotinib was 3 months and your median survival on chemotherapy was 6.5 months. So that data suggests — in the interaction test for VeriStrat status versus treatment received was quite — I don’t remember the exact number. I think 0.031 or something like that. So clearly statistically significant, an interaction between treatment choice in the second-line setting and what the VeriStrat status was.
DR LOVE: I was fascinated that there was a prospective Phase III study of a predictor of treatment response. And you might argue whether it showed anything useful, but I just think the fact that there was a Phase III study like that was quite interesting. I’m curious, Greg. Ram was concerned about the empiric nature of this. Are there any clues or hints about what they’re actually measuring and why it might correlate, or is this just purely an empiric thing?
DR RIELY: I think those are good words to describe it, clues and hints. But there’s nothing clear, to my knowledge, that really explains what we’re seeing, what the cause of the VeriStrat poor or VeriStrat good signature is.
DR LOVE: Do you believe that there actually is antitumor benefit of EGFR TKIs like erlotinib in patients with ER – EGFR-wild disease? I mean, you can cite various data, but in your heart, do you think it actually causes antitumor effect?
DR RIELY: I don’t. I don’t. I guess, what’s the lower limit of detection here? I mean, is it possible that occasionally, yes, absolutely. But I think that in my calculus when I’m seeing a patient, if I have an option of a chemo of any sort, I feel like that has a better chance of delivering antitumor response to a person by giving chemo, whether that is a vinca alkaloid, a taxane, gemcitabine. I think that’s better than erlotinib in that setting.
DR GOVINDAN: So we have to define what we mean by EGFR wild type. If you look at the specific mutations that attach — that actually is only the L858R and exon 19 deletion, a lot of those uncommon mutations that you may not pick up when you don’t do the next-gen sequencing.
I’m also unclear what proportion of patients in the VeriStrat good group had EGFR mutation. Do we have the data?
DR SOCINSKI: Yes. The data’s actually here.
DR GOVINDAN: In other words, is this all driven by the EGFR mutations?
DR SPIGEL: It was 6 versus 8 patients.
DR SOCINSKI: There were 6 versus 8 patients in the 2 arms.
DR GOVINDAN: But that was only when the tissue was available. It was not all comers, so that data is not exact.
DR SOCINSKI: Yes. And they’ve done the same analysis, excluding the known mutants.
DR GOVINDAN: Yes. So that’s 6 patients.
DR SOCINSKI: And there was really no difference here.
DR LOVE: But just to follow up with this man, he actually did end up receiving nanoparticle docetaxel. Right?
DR SOCINSKI: Yes.
DR LOVE: And what happened? How’s he doing?
DR SOCINSKI: He literally just received his first dose like 2 weeks ago. He had no problems. Since he was the first patient I put on the trial, our oncology pharmacy had more problems getting their act together.
DR SPIGEL: It’s an issue. One point I want to make about the VeriStrat, so I — there was one patient I shared with William Pao, a very motivated patient who asked if he could get the testing done, because he was EGFR wild type. And so he’s the only patient I’ve ever ordered the test in. And whatever you think about its merits as a predictive or prognostic test, the one thing I will say that struck me as amazing was how fast I got my results.
DR SOCINSKI: Two days.
DR SPIGEL: Yes. I actually got it in 24 hours. Maybe because they knew Spigel was ordering it for the first time. But it was crazy that you potentially, if that test is valuable, you could have results within 24 or 48 hours to make a treatment decision. That’s much different than waiting 5, 14, 22 days to select therapy. That mechanism of getting your result back based on a blood-based test is pretty fascinating.
DR BELANI: Is it flip of a coin?
DR LOVE: I’m kind of curious, though, whether you agree with Ram’s statement, because again, this came up in the discussion at ASCO – that there’s really not that much difference if — tell me if I’m quoting you correctly — in terms of tolerability and quality of life, erlotinib versus, let’s say, docetaxel?
DR SPIGEL: Yes. I disagree with you. I hate docetaxel. Our randomized trials — so we’re in the “BIND-014” as well, and we were just talking about the nivolumab versus docetaxel studies. It’s hard to have equipoise on a trial like that, not because — you know that nivolumab is going to be a great therapy, but because you hate docetaxel. For the same reason I like pemetrexed over paclitaxel. But docetaxel is a harder drug. And it’s all anecdotal. And I’m the first to say erlotinib is not the drug of choice in wild-type patients. But I have several patients with squamous lung cancer who’ve been through docetaxel, who’ve had terrible nail issues, who are not candidates for chemotherapy, who are on erlotinib, who month after month after month, they come in, and nothing’s going on. And I can’t explain it. They’re having no side effects and the best you would call it is stable disease. And that’s all anecdotal, but I’ll take that. I’ll take that, if that gives patients a few months.
DR GOVINDAN: Let me clarify what I said. Erlotinib is not a drug without side effects. I said arguably docetaxel has more side effects — myelosuppression, so I would never say that docetaxel is a comparable drug or easier drug to take compared to erlotinib. But erlotinib is a drug that has side effects, too. We cannot ignore that.
DR LOVE: I mean, maybe they’re both placebos and you’re trying to decide how to have less toxicity. But just globally, if I’m the patient and I say, “If I get this blood test done that I can get back in 24 hours and I’m going to do just as well on erlotinib,” understanding, as you say, certainly it’s not a benign drug, “as docetaxel,” I mean, it seems like it would be a practical thing, if you believe — does anyone have any questions in terms of the reproducibility of the test or —
DR HEYMACH: So a couple of points about it. Firstly, I’d say this is a really well-done study and it’s really — being involved in biomarkers, I can say it’s an impressive thing to do a study like this, to make the point. And it’s actually very well done. And second is that in terms of reproducibility, the findings, at least on erlotinib arms across multiple studies, have been very consistent with this. So I actually — and interacting with the groups a bit, I have no concerns about the reproducibility.
The issue I have a bit more concerns about is we have substantial data now that in the wild-type EGFR population, chemotherapy works better than erlotinib. And if you think about the “TAILOR” study, for example, that was just presented — and it doesn’t tend to be that close of a call in the overall group.
And so I guess I don’t see a large population where I think I’m choosing between the two. I tend to think of it — I’m typically going to be giving them sequentially, I think. I don’t see anybody that has wild-type EGFR that I think, “I’m only going to get one shot and that’ll be it.” So typically I’d be saying, “We’ll give chemotherapy first,” and almost everybody could get erlotinib afterward. But I’m not sure about the benefit of saving erlotinib there.
DR SOCINSKI: But on the other hand, we know that there is a huge drop-off from first to second to third line, and that when you say you think you’re going to get in multiple lines of therapy, I think the evidence would suggest that there’s a continuing drop-off. And even though you think you’re going to get third-line treatment in, it doesn’t always happen.