Divide and Conquer to Treat Lung Cancer
Mené sur 305 patients atteints d'un cancer du poumon non à petites cellules de stade avancé n'ayant jamais été traité et dont les tumeurs exprimaient un taux élevé de PD-L1 (pourcentage de cellules tumorales ≥ 50 %), cet essai de phase III évalue l'efficacité, du point de vue de la survie sans progression, et la toxicité d'une chimiothérapie à base de sels de platine ou du pembrolizumab, un anticorps monoclonal anti PD-1
The tactic of divide and conquer, which has been attributed to Philip of Macedonia and Julius Caesar in warfare, is transforming the treatment of lung cancer. Biomarkers are being used to identify subsets of patients with lung cancer who can receive initial treatment with a checkpoint inhibitor or a targeted agent. The results of the KEYNOTE-024 trial, now reported in the Journal by Reck and colleagues,1 show therapeutic benefits for the checkpoint inhibitor pembrolizumab for the subset of patients with non–small-cell lung cancer (NSCLC) who have a high level of programmed death ligand 1 (PD-L1) expression, which is defined by a PD-L1 tumor proportion score of 50% or greater (i.e., membranous PD-L1 expression on at least 50% of tumor cells, regardless of the staining intensity); this subset represented 30% of the screened KEYNOTE population. The patients treated with pembrolizumab achieved longer progression-free survival than did patients treated with chemotherapy (hazard ratio for disease progression or death, 0.50), as well as longer overall survival (hazard ratio for death, 0.60). In addition, grade 3, 4, or 5 treatment-related adverse events were half as frequent in the patients treated with pembrolizumab as in those treated with chemotherapy, a finding that further favors treatment with the checkpoint inhibitor over chemotherapy.
Earlier studies showed that previously treated patients with either nonsquamous or squamous lung cancers who were treated with the checkpoint inhibitors pembrolizumab and nivolumab had a survival advantage over those who were treated with chemotherapy, with hazard ratios for death of 0.59 to 0.73; these studies led to Food and Drug Administration (FDA) approval of these therapies.2-4 In the study favoring pembrolizumab,4 patients needed to have PD-L1 expression on at least 1% of tumor cells, whereas in the studies favoring nivolumab,2,3 patients were not selected on the basis of PD-L1 expression. The trial conducted by Reck and colleagues may establish a new standard of care for previously untreated patients with NSCLC and a PD-L1 tumor proportion score of 50% or greater.
FDA-approved targeted agents that are being used as initial treatment for patients with NSCLC include gefitinib, erlotinib, and afatinib for the 15% of patients with mutations of epidermal growth factor receptor (EGFR), as well as crizotinib for both the 5% with rearrangements of anaplastic lymphoma kinase (ALK) and the 1% with rearrangements involving the gene encoding the ROS1 proto-oncogene receptor tyrosine kinase (ROS1).5-9 The patients who participated in the KEYNOTE-024 trial had clinical characteristics that were different from those of the patients who participated in the trials of these other targeted agents. The majority of the patients in the KEYNOTE-024 trial were men, more than 90% were current or former smokers, and approximately 20% had squamous lung cancer. In the other trials,5-9 approximately 90% of the three oncogenic drivers (EGFR mutations or ALK or ROS1 rearrangements) were found in patients with adenocarcinomas, and the majority of those patients were women and had never smoked. In addition to differences in the patient characteristics, the foremost difference between the KEYNOTE-024 trial and the trials of the other targeted agents in previously untreated patients with NSCLC is the survival advantage for the patients treated with pembrolizumab over those treated with chemotherapy (hazard ratio for death, 0.60). In contrast to KEYNOTE-024, the other studies showed that previously untreated patients with oncogenic drivers who were treated with the targeted agents had longer progression-free survival than did those who were treated with chemotherapy (hazard ratio for disease progression or death, 0.30 to 0.47) but did not have statistically significant prolonged survival.
There are important issues regarding the use of PD-L1 as a predictive biomarker to identify patients who are most likely to benefit from treatment with pembrolizumab and other checkpoint inhibitors. The PD-L1 tumor proportion score cutoff point of 50% was prospectively defined in a phase 1 study that yielded a tumor response rate of 45% among patients with a score of 50% or greater, as compared with a tumor response rate of 12% among patients with a score lower than 50%; patients with a score of 50% or greater also had longer progression-free and overall survival than did patients with a lower score. In the three KEYNOTE reports,1,4,10 23 to 30% of patients with NSCLC had a PD-L1 tumor proportion score of 50% or greater, so this subset represents a substantial proportion of patients with NSCLC — a greater proportion than the approximately 20% who could receive initial treatment with targeted agents. However, the following points still need to be defined: the proportion of patients with NSCLC who have a PD-L1 tumor proportion score of 50% or greater; the presence of PD-L1 expression in patients with targetable oncogenic drivers, to determine whether those patients should receive initial treatment with a targeted agent or a checkpoint inhibitor; the relationship between the assay that uses 22C3 antibody and other existing PD-L1 assays in assessing PD-L1 expression; whether different cutoff points or other biomarkers can identify additional subsets of patients who may benefit from checkpoint inhibitors over chemotherapy; and the ability to predict the long-term outcomes in patients who have a response to checkpoint inhibitors.
Immunotherapy with checkpoint inhibitors will displace chemotherapy in yet another subset of patients with lung cancer. Oncologists may work with diagnostics laboratories to get appropriate, timely testing for PD-L1 expression in order to make determinations about the initial treatment of their patients with lung cancer. We await the results of long-term follow-up of large cohorts of patients who were treated with another checkpoint inhibitor, nivolumab, to see whether the initial survival advantage in the population translates to survival of 20% or greater at 3 years and beyond.11
New England Journal of Medicine , éditorial, 2015