KEYNOTE-189 and CheckMate-227: Practice-Changing Results, but How Should We Prioritize Them?

H. Jack West, MD


April 24, 2018


CheckMate-227 was another high-profile study published in NEJM[5] concurrent with the presentation at AACR. It had also been reported as positive in a press release[6] that described a significant improvement in PFS for recipients of the combination of nivolumab (nivo) with ipilimumab (ipi) compared with a standard-therapy arm of platinum-doublet chemotherapy (any of several potential histology-appropriate regimens) as first-line therapy for advanced EGFR and ALK wild-type squamous or nonsquamous NSCLC in patients with a high tumor mutational burden (TMB).

The trial randomly assigned patients with either tumor PD-L1 < 1% (n = 550) or, in a separate cohort of the study, PD-L1 1% or higher (n = 1189) in a 1:1:1 fashion to a chemotherapy doublet, nivo alone, or nivo/ipi. The study design was modified, however, to pool results across PD-L1 cohorts and compare outcomes in the subset of 299 patients who had tissue available for TMB testing who had received either chemotherapy (n = 160) or nivo/ipi (n = 139) and whose tumors were found to have a high TMB. This was defined as 10 mutations per megabase based on prior research,[7] is not correlated with PD-L1 expression, and was present in 44% of tumors tested.

The KEYNOTE-189 trial should have a greater impact.

Across multiple measures of PFS, results were significantly better with nivo/ipi in the high-TMB population: median PFS was 7.2 versus 5.5 months; 1-year PFS was 42.6% versus 13.2% (HR 0.58, P < .001). Recipients of nivo/ipi also demonstrated a higher RR (45.3% vs 26.9%). These results were independent of PD-L1 expression or NSCLC histology. OS results were relatively immature, with a nonsignificant trend toward improved OS with nivo/ipi among patients with high TMB tumors (HR 0.79). Notably, patients who received nivo monotherapy did not show superior efficacy. The toxicity profile, though different for the two arms, was relatively comparable in frequency and severity with nivo/ipi versus chemotherapy.

The Clinical Impact

Although additional trials were presented beyond these data, these two studies provide plenty to consider in terms of how they should change our clinical practices for patients with EGFR and ALK wild-type advanced NSCLC.

My view is that the KEYNOTE-189 trial should have a greater impact, leading us to be far more inclined to favor the already FDA-approved carboplatin/pemetrexed/pembrolizumab combination for eligible patients, at least those with tumor PD-L1 expression less than 50%.

For those with high PD-L1 expression, pembrolizumab monotherapy remains a strong consideration that I will be more inclined to favor, given the relatively comparable efficacy, lower risk for side effects, and the fact that this preserves the option of pursuing platinum-based doublet chemotherapy as a subsequent option upon progression.

You don't take a knife to a gunfight, and you can't take PFS to an OS fight.

The results from CheckMate-227 leave me intrigued to see more work with nivo/ipi, and I believe that there may be a subset of patients for whom this could be an especially favorable regimen. TMB assessment is still plagued by practical challenges, ranging from questions about the reliability of the metrics to identify optimal patients, to the need for significant additional tissue, to a turnaround time that routinely extends to at least 2-3 weeks. These issues limit the utility of TMB for selecting first-line treatment, particularly when the benefit is limited now to PFS rather than OS. I believe that on a different day, when not presented immediately after the results of KEYNOTE-189, the CheckMate-227 trial would be considered more notable. But you don't take a knife to a gunfight, and you can't take PFS to an OS fight.

A Dynamic Moment in Time

We should conclude with a reminder that these results represent just this particular moment in time. We have already learned from a recent press release that the KEYNOTE-042 trial of first-line pembrolizumab versus doublet chemotherapy for patients with tumor PD-L1 expression of 1% or greater is positive for an OS benefit with pembrolizumab in a broader population of advanced NSCLC patients (squamous or nonsquamous, EGFR and ALK wild type) than that for which pembrolizumab monotherapy is currently indicated.[8]

We can anticipate that this will leave us with an enviable dilemma: evidence and likely future FDA approvals supporting the use of either pembrolizumab alone or in combination with carboplatin/pemetrexed in the approximately two thirds of advanced NSCLC patients with PD-L1 of at least 1%. Another press release reported that the IMpower131 study of first-line atezolizumab added to carboplatin with nab-paclitaxel or conventional solvent-based paclitaxel in advanced squamous NSCLC is positive for a significant PFS benefit with carboplatin/nab-paclitaxel/atezolizumab versus the same chemotherapy backbone alone.[9] We await the upcoming presentation of both of these important trials as well as several others that speak to the dynamism of the field.

In the meantime, as the options for our patients increase, our understanding of how best to integrate immunotherapy into lung cancer treatment will probably continue to evolve rapidly in the coming months and years as new trial results are released. I expect that we will need to revisit our treatment standards following our upcoming review of highlights from ASCO.

Stay tuned.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.