Hossein Borghaei, DO: CheckMate-227 is a large, randomized, complicated study in terms of its design that basically tests nivolumab in combination with either chemotherapy or ipilimumab and compared it against chemotherapy in patients who have either no PD-L1 expression or have at least 1% PD-L1 expression. For the patient population with over 1% PD-L1 expression, the randomization is to 3 separate arms, 1:1:1, of ipilimumab plus nivolumab versus histology-based chemotherapy versus nivolumab alone. For the patient population with less than 1%, which is basically negative PD-L1 expression, the randomization again is to ipilimumab/nivolumab, histology-based chemotherapy, or combination nivolumab plus histology-based chemotherapy. It should be emphasized that both squamous cell and nonsquamous cell histologies were allowed to participate, and the rest of the criteria were very similar to other phase III studies we have seen: no EGFR or ALK translocation, no active brain metastases, and a number of other inclusion/exclusion criteria that are customary. The study is still ongoing because overall survival is an endpoint on the study that has not been reached, but at this year’s AACR meeting, the primary data, top-line data, were presented in the patient population that has high tumor mutational burden.

The way the study was designed initially, it did not have TMB as a coprimary endpoint, TMB referring to tumor mutational burden. But as the results of several retrospective studies became available—and 1 of these studies was this large phase II study called CheckMate-568—the TMB cutoff of 10 mutations per megabase was shown to be a predictor of clinical efficacy for a combination of the PD-L1 inhibitor nivolumab. CheckMate-227 was reamended to include the coprimary endpoint of TMB.

In CheckMate-227, the primary analysis that was done was on patients with high TMB, defined as 10 mutations per megabase. Compared to standard chemotherapy, the combination of a PD-L1 inhibitor, nivolumab, in that patient population showed a much better PFS and response. The survival data weren’t mature enough, but that is something that is going to be looked at. In terms of the safety of the combination, we have to keep in mind that this particular regimen of ipilimumab/nivolumab was actually studied rather heavily as part of CheckMate-012, which actually was a dose-optimization type of study where different dose levels and schedules of delivery for both drugs were investigated in multiple cohorts of patients. That came up with the current schedule and dose that’s part of CheckMate-227.

From the toxicity point of view, clearly when you add ipilimumab to nivolumab, you get more toxicity. If I recall correctly, the percentage of patients with grades 3 and 4 toxicity, the severe toxicity, was about 25%, which is definitely higher than what you get with nivolumab alone. On the other hand, the clinical efficacy you get in terms of response and PFS is far better than what you get with nivolumab.

When you compare that—this study allows us to do those kinds of comparisons with chemotherapy—I think any time you add a third agent to a platinum doublet, you anticipate getting additional toxicities. But I think most of the toxicities were in line with what you would get from a chemotherapy combination. I don’t think they are out of step with what we’ve seen from other I-O plus chemotherapy combinations that have been reported so far. We’re not ignoring grades 1 and 2 toxicities, but there were definitely higher rates of grade 3 or 4 toxicities with either ipilimumab/nivolumab compared to nivolumab alone. Again, I would say that ipilimumab/nivolumab has less of those side effects, grades 3/4 severe toxicities, as compared to nivolumab plus chemotherapy.

CheckMate-227 allowed both patients with squamous and nonsquamous histology to participate. The percentage of patients who had squamous cell histology was actually a little bit lower in the study. Sometimes, when you do these subgroup analyses, smaller numbers make it difficult to make conclusive statements about the efficacy, toxicity, or whatever it is that you might be seeing. I think that we need larger trials to differentiate between squamous and nonsquamous patients. I think we’re going to be getting some of these common chemotherapy combinations for squamous cell patient populations. But when you look at the overall data, it becomes a little bit difficult to compare CheckMate-227 versus KEYNOTE-189 exactly, just because KEYNOTE-189 had 1 chemotherapy combination backbone. Everybody got carboplatin/pemetrexed and everybody had nonsquamous histology. But in CheckMate-227, there are different chemotherapy backbones and patients have squamous and nonsquamous disease. I think that we have to keep that in mind as we interpret the results of these trials.

Transcript Edited for Clarity