An accompanying commentary notes several salient issues:
First, number of patients with PD-L1 TPS of ≥50% was close to 50%, which is much more than 30% usually seen in general population. Thus, because magnitude of treatment benefit rose with increasing PD-L1 TPS, results obtained in all three cohorts were skewed towards the high PD-L1 expression.
Second, after results of the KEYNOTE-024 study in patients with TPS >50% became public, whether pembrolizumab monotherapy improves progression-free survival and overall survival in patients with PD-L1 TPS <50% became clinical question of interest. Unfortunately, in an exploratory analysis there was no discernible improvement with pembrolizumab in overall survival in this population. Thus, this population might initially derive more benefit from chemotherapy and an as-yet unidentified group of patients with PD-L1 TPS of 1–49% could derive long-term benefits from pembrolizumab.
It advises that the conclusion of KEYNOTE-042 that pembrolizumab is an appropriate treatment option for patients based only on a PD-L1 TPS <50% should be regarded with caution. In the absence of a tumour mutation burden result, which is unlikely to be available for most patients in the near future, patients with low PD-L1 TPS might be managed best with a combination of chemotherapy and PD-1 antagonists.