DR SPIGEL: Patients who had received crizotinib for ALK-rearranged lung cancer as part of the pivotal Phase I study that got crizotinib approved in patients that were part of a Phase II experience. They looked back at those patients who had all responded to crizotinib. So that was one of the criteria. You couldn’t have gotten crizotinib and then progressed. And then asked the simple question: At progression, those patients that went on to stay on crizotinib, who were clinically stable, their physicians felt like they were benefiting in some meaningful way, those patients that remained on crizotinib, how did they do compared with patients that didn’t, that went on to other therapies?
So a retrospective analysis, we know there’s a lot of bias involved in who these patients might be in terms of selection. It was about 120 patients that were examined, who went on to receive crizotinib, 74 that went on to receive other therapy. And basically what was discovered was that patients who had the opportunity to stay or continue beyond progression actually had better outcomes than the group of patients who went on to other therapies that did not include crizotinib. So the overall survival was improved. The time to progression was improved.
And interestingly enough, if you looked at the overall survival from the time that crizotinib was first used — so this is before they progress — the overall survival was 29.6 months versus about 11 months. So that’s pretty impressive.
Now the faults of this analysis, I think, are obvious, that you actually took the best patients, the patients who were clinically stable, doing the best, who had a chance to remain on crizotinib. We would expect them to do the best anyway. What if you randomized those patients, those very patients, to crizotinib or pem or something like that? How would they do? And we don’t know that. This is the limitation of a retrospective analysis. But certainly supports the idea of continuing crizotinib in those patients that are doing well, at least until we have more data.
DR LOVE: So Greg, you were the second author on this paper. Can you put in context what you think the value of continuation of crizotinib is beyond progression, for example, compared to what we were talking about before with the EGFR TKIs?
DR RIELY: Yes. I think for both situations, EGFR and ALK, we have to understand that resistance as it develops is likely a minor population of cancer cells. And it may occur in individual sites as an escape event that doesn’t occur in multiple places. So I think both those things together make us think that we likely would get some benefit from continued treatment.
Now, one of the big differences between ALK and EGFR, though, is the data that we have about safety of adding something to the underlying tyrosine kinase inhibitor. So for erlotinib, gefitinib, we have lots and lots of data saying you can add just about any chemo to erlotinib or gefitinib, in contrast with crizotinib, with a very razor-sharp focus as they developed that drug. And there’s very little data about combinations. There’s very little Phase I data to say it’s safe to give erlotinib and pemetrexed or — sorry, crizotinib and pemetrexed. There’s no data that says it’s safe to give in any other combinations.
So when these patients were continued on crizotinib in the trial, for the most part they were continued on just crizotinib. That’s all they were given.
DR LOVE: In spite of progression.
DR RIELY: In spite of progression. Sometimes local therapies were administered, and sometimes it was just a sense that they were growing slowly. So I think that’s the type of patient that we’re thinking of here. And I think Dave’s criticisms are perfectly appropriate, that there are a lot of biases built into that selection, that we probably can’t put down on paper. But there certainly are biases there.
DR LOVE: So in terms of what, quote, the standard is or what you’re doing right now with your patients with ALK positivity who progress on crizotinib, Ram, is continuation of crizotinib, either by itself or with something else, like chemo, something that you commonly bring up? I kind of was getting the feeling in terms of the comments you all made before that maybe the standard of care is to find a trial on these second-generation agents.
DR GOVINDAN: That’s easier said than done sometimes. So that’s the first choice. And if you cannot do that — a patient cannot travel somewhere, so what do you do? So the challenges, I think, as you heard from Greg before, do we have the safety for combination studies? For example, with erlotinib, we have very good safety data with pretty much all chemotherapy drugs, perhaps except with the exception of vinorelbine, whereas with crizotinib, that’s the problem. We don’t have very many combination studies that would tell us it’s safe to combine them. So for the most part, I think at the moment, if you’re going to combine them, you’ve got to watch them carefully. I suppose you can combine pemetrexed and crizotinib safely without problems. But you have to be very cautious about that. That’s why I feel that things are a little bit different here compared to EGFR TK inhibitors. So in practice I just don’t want to leave it vague, I would continue chemotherapy with crizotinib if the chemotherapy is pemetrexed. That’s about as far as I would go. In other situations, I would just do chemotherapy alone and, if they progress, I will retry crizotinib for now.
DR LOVE: So just to be clear, a patient who’s clearly progressing, maybe even a little bit symptomatic on crizotinib, Mark, again, let’s assume there isn’t a protocol, one of these second-generation agents, just bottom line, what are you likely to do? If you give chemo, what kind of chemo? Would you combine it with crizotinib? What do you do?
DR SOCINSKI: Well, I mean, if they’re clearly progressing and they need some form of effective therapy, I think the second-generation agents are of great interest. But we don’t always have access to those. If I’m going to treat them, I think I’d revert back to platinum-based doublets. Here, the preferred agent is pemetrexed. I don’t know why these patients shouldn’t get bevacizumab, if it’s indicated in this population.