WCLC 2021 | KEYNOTE-598: pembro + ipi in metastatic NSCLC

I presented the results of the KEYNOTE-598 study at World Lung in the last few days. KEYNOTE-598 is a placebo-controlled randomized double-blind trial, which compares the combination of pembrolizumab and ipilimumab to pembrolizumab plus placebo as first-line treatment for patients with non-small cell lung cancer and a tumor proportion score for PD-L1 of 50% or more, and also no activating mutations in EGFR or ALK.

And so this is an important study to try and understand for the first time in an appropriately powered way, the contribution that ipilimumab makes. The study accrued just under 570 patients or 568 patients in total, they were evenly split between the two treatment arms, which were well-balanced for the characteristics. This included both squamous and non-squamous cancers, about a quarter was squamous, three quarters non-squamous.

The major finding in the study was that there was no difference in overall survival between the two study arms. The medians were almost identical, 21.4 and 21.9 months. The hazard ratio of 1.08 really indicates that there’s no difference and there was a non-significant P-value. Similarly, for progression-free survival, there was no difference between the two arms. Median progression-free survival was approximately eight months in each of the study arms. And again, the hazard ratio was 1.06. So indicating not any real difference between these two.

Same sort of results for response rate and duration of response. So overall, in terms of the effectiveness of the treatment, there’s really no difference between these two arms. By contrast, when you start looking at side effects and start looking at some of the adverse events, there are differences. And it doesn’t matter whether you look at treatment-related adverse events or immune-mediated adverse events. There are more adverse events with the pembrolizumab-ipilimumab combination than with pembrolizumab alone. And that applies to all grade toxicities, to grade 3 to 5 toxicities, to the serious adverse events, to the ones that led to death and the ones that led to discontinuation of treatment. So, when you take all this into account, there are more adverse events. It’s important to point out though, that those adverse events for the pembro-ipi combination are very much in keeping with what’s been seen when nivolumab is combined with ipilimumab, for example. And so, really the study arms performed as we expected based on prior data.

When you wrap this all up and you say, “Well, what does it mean?” I think what it means is, that there’s no value for this patient population, so that’s the tumor proportion score greater than 50% population, there’s no value in the addition of ipilimumab to pembrolizumab. It just adds side effects without adding any effectiveness. And it means that pembrolizumab as monotherapy remains a standard of care for that patient group.

So one of the interesting questions is why didn’t the addition of another drug make a difference? And there are a few possible explanations. Firstly, what is the finding not due to? We know it’s not due to the fact that pembrolizumab did better than expected or pembrolizumab and ipilimumab did worse than expected. These drugs behave the way we would expect them to behave based on other studies. So, it’s not that something did extraordinarily well. I think one of the possibilities, and it was touched on in the discussion of the paper at the conference, is that when you take that tumor proportion score greater than 50% population, you’ve got a population that already has an activated immune response. So there’s already immune cells infiltrating in this kind of tumor.

And it may be that even by adding a CTLA-4 inhibitor like ipilimumab, you can’t make that any better. So it might be a population of patients where it really doesn’t add to that. Another bit of evidence that supports that, is if you go back to the CheckMate 227 study, although it wasn’t really the primary endpoint. In fact, it was the group of patients with the very lowest TPS, less than 1%, who actually had the highest or the most positive hazard ratios. So it may be that, that’s the group of patients where there’s something to gain. But certainly based on our results, there’s no justification for adding ipilimumab to a PD-1 inhibitor, in particular pembrolizumab for this group of patients.