COMMENTARY

In Advanced NSCLC, Immunotherapy Evolving to Become First-Line Treatment

H. Jack West, MD

Disclosures

June 25, 2018

By all measures, the 2018 annual meeting of the American Society of Clinical Oncology (ASCO) was big for lung cancer, as we would have expected by the array of press releases leading into it. The most anticipated presentation was KEYNOTE-042 (KN-042), a randomized phase 3 trial of first-line pembrolizumab versus doublet chemotherapy administered to 1274 patients with squamous or nonsquamous non–small cell lung cancer (NSCLC) without an EGFR mutation or ALK rearrangement, and with tumor PD-L1 expression of 1% or more. This trial, already reported as positive for a significant improvement in overall survival (OS),[1] was slated to be featured in the ASCO 2018 plenary session and presented by Gilberto Lopes, MD.[2] This advance information about the positive KN-042 trial cast a long shadow that limited our ability to interpret the clinical significance of KEYNOTE-189 (KN-189), the randomized phase 3 trial of first-line chemotherapy (cisplatin or carboplatin with pemetrexed), with or without pembrolizumab in EGFR and ALK wild-type patients with advanced nonsquamous NSCLC across the full spectrum of tumor PD-L1 expression, that was highly positive for an OS benefit in an overlapping population.[3]

KN-042 delivered on the promised OS benefit for the broad trial population for pembrolizumab over doublet chemotherapy (median 16.7 vs 12.1 months, respectively; hazard ratio [HR], 0.81). However, the breakdown of OS by tumor PD-L1 expression demonstrated that the benefit was most pronounced in the subgroup of patients with high PD-L1 expression (> 50%) and, as illustrated in the Table, was progressively diluted as the patient populations with > 20% and > 1% expression were included.

Table. Efficacy in KEYNOTE-042 by Subgroup, Defined by PD-L1 Expression

Patient population Median PFS
(pembro vs chemo, respectively)
HR
PD-L1 ≥ 50% 20.0 vs 12.2 months 0.69
PD-L1 ≥ 20% 17.7 vs 13.0 months 0.77
PD-L1 ≥ 1% 16.7 vs 12.1 months 0.81
PD-L1 1%-49% 13.4 vs 12.1 months 0.92

The most instructive comparison presented by Dr Lopes was the analysis of pembrolizumab versus doublet chemotherapy in patients with tumor PD-L1 expression of 1%-49%, which revealed that there was no difference in OS between the two arms in this subset. Overall, the findings clarify that the benefit of pembrolizumab over chemotherapy was limited to the subgroup of patients with high tumor PD-L1 expression for whom pembrolizumab monotherapy is already a current standard of care based on the KEYNOTE-024 trial, which previously demonstrated the superior efficacy of pembrolizumab over chemotherapy in patients with tumor PD-L1 expression > 50%.[4] These findings indicate that the positive results for the overall, broader trial population were because patients with low tumor PD-L1 were "carried" by the especially strong results in the patients with high PD-L1. Other notable shortcomings included the lack of an improvement in progression-free survival (PFS) with pembrolizumab and the absence of crossover in the trial to allow patients assigned to first-line chemotherapy to receive pembrolizumab upon progression. As mentioned by Leena Gandhi, MD, PhD, in her plenary session discussion,[5] because the trial was conducted globally, only approximately 20% of patients ultimately received the benefit of an immune checkpoint inhibitor as subsequent treatment off protocol, meaning that the vast majority of patients did not benefit from second-line treatment with an established OS benefit.

KEYNOTE-042 seemed like it had overpromised and underdelivered as a positive NSCLC trial.

We should expect the broadening of the label for pembrolizumab to include patients with low PD-L1, based on the positive results of KN-042, but does this mean that this is a strong option for them? While the OS with pembrolizumab monotherapy in these patients is comparable to that of chemotherapy and includes a more favorable toxicity profile for immunotherapy over chemotherapy, this is also relatively disappointing in the context of the handicap favoring the pembrolizumab arm due to the lack of crossover for the chemotherapy arm.

At the same time, results using the combination of pembrolizumab concurrent with doublet chemotherapy are now emerging as a stronger option. Data from the just-reported KEYNOTE-407 (KN-407) trial provide a compelling first-line treatment approach for patients with advanced squamous NSCLC across the full spectrum of tumor PD-L1 expression and complement the excellent results seen for patients with advanced nonsquamous NSCLC with platinum/pemetrexed/pembrolizumab in the KN-189 study.

If KN-042 seemed like it had overpromised and underdelivered as a positive NSCLC trial featured in the plenary session, KN-407 was more of a sleeper, with results that proved to be far more interesting and clinically relevant than had been anticipated leading up to ASCO 2018. KN-407, a trial randomizing 559 patients with advanced squamous NSCLC and any level of tumor PD-L1 expression to first-line carboplatin and either paclitaxel or nab-paclitaxel with either pembrolizumab or placebo, was reported in early May as positive for significant improvement in the secondary endpoint of objective response rate,[6] but this endpoint is not as well correlated with survival and generally not as valued as PFS or especially OS. The trial picked up momentum when the ASCO abstract was released in mid-May and reported a difference in ORR of 23% favoring chemo/pembrolizumab over chemo/placebo (58.4% vs 35%, respectively), but its importance rose dramatically when a press release in late May revealed a significant improvement in both PFS and OS.[7]

The trial results presented at ASCO[8] proved so impressive that some felt it should have received top billing in the lung cancer program, if only the results had been available before 2 weeks prior to the conference. The significantly greater OS with chemo/pembrolizumab was highly clinically significant (median OS 15.9 vs 11.3 months; HR, 0.64; P = .0008) and was comparable across multiple different clinical subgroups: age, sex, performance status, and whether patients had received paclitaxel or nab-paclitaxel. Most notably, the OS benefit with first-line chemo/pembrolizumab was present across all of the PD-L1 subgroups (HR, 0.61, 0.57, and 0.64 for populations of < 1%, 1%-49%, and > 50% tumor PD-L1 expression, respectively). There was also a highly significant improvement in PFS with chemo/pembrolizumab (median 6.4 vs 4.8 months; HR, 0.56; P < .001), also present across all levels of tumor PD-L1 expression, along with the striking difference in response rate favoring the pembrolizumab arm (58.4% vs 35%; P = .0004). The toxicity profile was only marginally higher with the addition of pembrolizumab.

Taken together, the results from KN-042 and KN-407 can be combined with previously reported results from KN-189 to offer a comprehensive array of options that move immunotherapy into the first-line setting for patients with advanced NSCLC with either squamous or nonsquamous EGFR and ALK wild-type NSCLC. There are now two highly positive trials for first-line concurrent chemo/immunotherapy with histology-specific doublets combined with pembrolizumab that show a highly significant OS benefit across the spectrum of tumor PD-L1 expression. Pembrolizumab monotherapy remains a feasible option to consider for patients with PD-L1 > 1%, but I would argue that the evidence today suggests that chemo/immunotherapy is a stronger option for all but the subset of patients with high tumor PD-L1 expression of > 50%. With new press releases delivering immunotherapy trial results in advanced NSCLC every few weeks, we can only wonder whether this new standard will become a lasting standard, or whether it will be challenged by new options and biomarker-defined subgroups as the new results roll in.

We've come a long way from the days, not many years ago, when the lung cancer field would bat "O for ASCO" without any positive trials. Much more to come!

Comments

3090D553-9492-4563-8681-AD288FA52ACE
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.
Post as:

processing....