Patients with advanced gastroesophageal adenocarcinoma (GEA) have a poor prognosis. Despite recent improvements in treatments, overall survival (OS) for most of these patients remains < 18 months. Immune checkpoint inhibitors in combination with chemotherapy are now approved for treatment of advanced GEA in the first-line setting (pembrolizumab for esophageal cancers and nivolumab for GEA). These approvals are based on the results of the randomized phase 3 studies KEYNOTE-590 and CheckMate 649. 1,2 However, there remains a need to better define patients who are more likely to benefit from these agents.
KEYNOTE-062 was a phase 3 study of pembrolizumab in patients with advanced gastric and gastroesophageal junctionadenocarcinoma. 3 Patients were randomly assignedto receive chemotherapy, chemotherapy plus pembrolizumab, or pembrolizumab alone. This study did not lead to pembrolizumab approval for gastric cancer or GEA, although, consistently with other studies, it suggested that patients whose tumors had higher programmed death-ligand 1 expression were more likely to benefit from immunotherapy treatments.
Lee and colleagues published results of a prespecified exploratory analysis from this study that evaluated the association between tumor mutational burden (TMB) and patient outcomes. Those who had tumors with higher TMB had better outcomes from immunotherapy treatments, but not chemotherapy alone. In patients with TMB ≥10 mut/Mb, pembrolizumab vs chemotherapy significantly improved the objective response rate (ORR; 55.6% vs 41.2%), progression-free survival (PFS; 11.1 vs. 7.0 months; hazard ratio [HR] 0.52) and OS (31.6 vs 13.4 months; HR 0.34). Similarly, pembrolizumab plus chemotherapy vs chemotherapy alone improved ORR (73.3% vs 41.2%), PFS (31.1 vs 7.0 months; HR 0.62), and OS (31.1 vs 13.4 months; HR 0.54). However, after patients with microsatellite instability–high (MSI-H) tumors, who represented 44% of patients (22 of 50), with TMB ≥10 mut/Mb were removed from the analysis, the positive association between TMB and immunotherapy effect on OS and PFS was no longer significant. The small number of patients with high TMB limits the interpretation of these exploratory data. However, the results suggest that TMB is unlikely to emerge as a significant clinically relevant biomarker in patients with microsatellite-stable upper gastrointestinal cancers.
Ultimately, novel therapeutic strategies are needed to improve outcomes. The phase 3 BRIGHTER study reported by Shah and colleagues evaluated the activity of napabucasin, a generator of reactive oxygen species, in advanced GEA. This was a phase 3 study that enrolled 714 patients whose tumors progressed on one prior line of therapy. Patients were randomly assigned in a 1:1 fashion to receive eitherpaclitaxel plus napabucasin or paclitaxel plus placebo. The primary endpoint was OS. Patient characteristics were well balanced between the treatment arms. At the preplanned interim analysis, there was no difference in OS between the two treatment arms (6.97 months with napabucasin vs 7.29 months with placebo; P =.5699).As such, the study was terminated before enrollment completion because it was deemed futile to continue. There were no new safety signals. It is important to note that since the study was launched, ramucirumab and paclitaxel became the new standard treatment in the second-line setting per the phase 3 RAINBOW study. 4 As such, this combination regimen should be offered to patients who are candidates for systemic treatment in the second-line setting and who have no contraindications to these agents.