Practice Changing Trials in Advanced Non-Small Cell Lung Cancer - Episode 6
Benjamin P. Levy, MD: Let’s move on to some of the other data that have emerged, next talking about 2 trials that have looked at combining immunotherapy with chemotherapy in an unselect group of patients in the advanced-stage setting. Hoss, you want to walk us down 021G and maybe the 189 story that’s beginning to unravel?
Hossein Borghaei, DO, MS: Sure. So, we participated in the KEYNOTE-021 trial. This is a larger study with multiple cohorts, and the one that we’re going to concentrate on is the now famous cohort G subpopulation. This ended up being a randomized phase II design after going through all of the other cohorts for patients with nonsquamous non—small cell lung cancer. It was unselected based on PD-L1, so everybody could get in. And the randomization was simple. We would use the standard carboplatin/pemetrexed followed by maintenance pemetrexed versus the same backbone with pembrolizumab grafted on to the platinum doublet given every 3 weeks. I have to admit, and I have said this before, personally I was a skeptic that adding cytotoxics to checkpoint inhibitors would be a reasonable strategy. I didn’t really anticipate positive results as we’re seeing, but nonetheless, obviously I’m proven wrong. So, roughly 60 patients per arm were recruited. I’ve got to tell you again, because we put a number of patients on this study, it was a well-tolerated regimen. I really didn’t have any major issues managing patients on this.
So, at the time of the first reporting that Dr. Langer did at ESMO and the subsequent publication in Lancet, PFS and response rate were superior in the group that got the triple therapy as opposed to the platinum doublet. Crossover was allowed such that if patients on the chemotherapy arm had any evidence of progression, they were allowed to get the additional pembrolizumab and all of that. So, the PFS and response rate made this an interesting combination, and the PFS improvement was quite significant. We don’t really have data, in terms of PFS improvement based on level of PD-L1, because the subsets were so small that it would have been really difficult to have any logical conclusions drawn from these smaller subsets. But what I think is even more surprising is that with further follow-up, overall survival started trending toward the triple combination. Now, that analysis that I presented at World Lung just last year doesn’t have the statistical power behind it because alpha was spent for the primary analysis, but nonetheless, you see a separation of the curves, perhaps suggesting that chemotherapy can add to the additional benefit we see with PD-L1 or PD-1 inhibition. The interesting part, again, is that it does not seem to be necessary right now, depending upon the level of PD-L1 expression.
And then we have a press release that the exact design, but in a phase III format, KEYNOTE-189, is positive for both PFS and OS. Obviously, I haven’t seen the data so I can’t really tell you what they really show or whether there are subgroups that are benefitting more. Is the overall survival driven, for instance, because of the high PD-L1 subgroup or not? We don’t have any of that data. But I think this study, along with what we’re going to discuss next, IMpower, makes you wonder if the addition of chemotherapy could at least have an additive effect, if not synergistic, to the checkpoint inhibitors that we’re using right now. So, interesting results. I would not have predicted the overall survival advantage, but here it is. I think it’s becoming part of what we’re using because based on a phase II study, the FDA has approved a triple combination, as we know, so it gives us an additional option for patients who come in without molecularly driven tumors.
Benjamin P. Levy, MD: Excellent summary. I have a lot of questions.
Transcript Edited for Clarity