Several weeks ago, at a very crowded plenary session for the European Society of Medical Oncology (ESMO) in Copenhagen, Denmark, results with first line immunotherapy compared to standard first line chemotherapy for patients with advanced non-small cell lung cancer (NSCLC) were presented that simultaneously ushered in a new era for testing for PD-L1, the leading predictive marker for sensitivity to immunotherapy, and indicated both the new promise and limitations of PD-1 immune checkpoint inhibitors such as Keytruda (pembrolizumab) and Opdivo (nivolumab). In fact, the remarkably different results from two similarly designed trials leave us in a new world, but also one in which further change is coming.

First, let’s discuss the positive results for first line. The KEYNOTE-024 trial, which enrolled 305 patients with high level expression of the PD-L1 antibody marker known as 22C3 (>50% of tumor cells staining positive, seen in about 30% of patients with advanced NSCLC overall), randomized patients to either first line treatment with standard chemotherapy with any of several chemotherapy doublets or Keytruda at a new fixed dose of 200 mg IV every 3 weeks. Notably, patients with an EGFR mutation or ALK rearrangement were excluded from the trial, based on evidence that these patients are far more likely to benefit from the targeted therapies against their driver mutations than either chemotherapy or immunotherapy. Patients assigned to Keytruda were to continue until significant side effects or progression. Patients who were assigned to chemotherapy could receive maintenance Alimta (pemetrexed) if they had not demonstrated progression after 4-6 cycles of initial doublet chemotherapy; patients on the chemotherapy arm were eligible to cross over to receive Keytruda as second line therapy. 

The trial demonstrated a highly significant improvement in progression-free survival (PFS); the median PFS (the time when half of patients have progressed and half have not) being 10.3 months for the first line Keytruda recipients vs. 6.0 months for the first line chemo patients. The differences became more pronounced with longer follow up, so that by 1 year from the start of treatment, 48% of patients assigned to Keytruda still hadn’t progressed, while 15% of the patients starting on chemo hadn’t progressed. In terms of response rate, the probability that measurable cancer will shrink significantly, it was significantly better with Keytruda – 45% vs. 28%.  As is typical with immunotherapy trials, chemotherapy caused more side effects, though a minority of patients will have challenging and even rarely serious side effects with immunotherapy.

Both groups of patients did relatively well in terms of overall survival (OS), but a higher proportion of those starting with Keytruda remained alive a year into the trial (70% vs. 54%). Based on these differences in efficacy favoring Keytruda, the Data Safety Monitoring Committee following the trial recommended stopping the trial because it would have been considered unethical to continue to randomize patients to chemotherapy in light of the emerging findings. Notably, however, while this survival benefit was seen despite the built-in crossover of chemo patients to Keytruda, only about half of the progressing patients had received immunotherapy, a low proportion that is unexplained, disappointing, and partly challenges the idea that it is critical to get immunotherapy first, because too many patients assigned to first line chemo failed to ever get immunotherapy, despite the fact that this is a treatment that has been repeatedly proven to improve survival as a second line therapy.

With that presentation and the simultaneously published article in the New England Journal of Medicine, the standard of care for advanced NSCLC changed, as indicated by the remarkably quick update in the NCCN guidelines (the leading treatment recommendations put forth by a group of cancer experts as defining our best treatment) and a new approval for Keytruda as first line therapy, specifically for patients with high level expression of PD-L1. This means that it is now necessary to have the tumors of newly diagnosed patients with advanced squamous or nonsquamous NSCLC tested for PD-L1, and for the 30% of patients with high level expression of PD-L1, to favor single agent Keytruda.

But despite this clear victory for immunotherapy in advanced NSCLC, this doesn’t mean that most or all patients should get immunotherapy as initial treatment. Though some provocative data came out looking at chemo combined with immunotherapy as first line therapy, that relatively small trial didn’t show a survival benefit compared with first line chemo followed by immunotherapy. We need to also remember that the trial excluded patients with an EGFR mutation or ALK rearrangement, as the oral targeted therapies for these patients are remarkably effective, and we’ve seen disappointing (though still very limited) results with immunotherapy overall in these populations.  Importantly, we must remember that good results with first line Keytruda are seen thus far only in the high PD-L1 expression group, the 30% of patients with the best probability of benefit from immunotherapy, and that we can’t presume that immunotherapy would be better for the 70% of patients with a lower probability of benefiting greatly from immunotherapy.

And that brings us to the humbling results of the Checkmate-026 trial of Opdivo vs. chemotherapy, which I’ll cover in a post later this month.