Gastroesophageal adenocarcinoma (GEA) and squamous esophageal cancer (ESCC) are responsible for >1 million deaths annually globally. Until now, patients with metastatic GEA and ESCC could anticipate survival of less than one year. Anti-PD-1 monotherapy has demonstrated modest efficacy in previously treated GEA and ESCC. In 2020, four pivotal trials have established anti-PD1 therapy as a new standard of care for selected GEA and ESCC patients as first-line advanced and adjuvant therapy. In this review, we discuss the recent results of the CheckMate 649, ATTRACTION-4, KEYNOTE-590 and CheckMate 577 trials. We consider these results in the context of current standards of care and historical trials of immune checkpoint blockade in GEA and ESCC. We explore biomarker selection for anti-PD-1 therapy and appraise the future of combination therapies. In CheckMate 649, treatment with oxaliplatin-fluoropyrimidine chemotherapy plus nivolumab in patients with combined positive score (CPS) ≥5 GEA tumors provided a clinically meaningful and statistically significant improvement in overall survival. The ATTRACTION-4 trial did not see a similar overall survival benefit, despite a clear improvement in progression-free survival. We review potential explanations for this result. KEYNOTE-590 showed profoundly improved survival when pembrolizumab was added to cisplatin-fluoropyrimidine chemotherapy in ESCC patients with CPS ≥10 tumors; this benefit was less convincing in unselected ESCC. Finally, CheckMate 577 provides proof of concept for the improvement in disease-free survival with adjuvant nivolumab in high-risk resected GEA and ESCC following trimodality therapy. Immune checkpoint blockade has come of age in GEA and ESCC, and will now be integrated into first-line and earlier lines of therapy, providing benefit for a larger proportion of patients. Biomarker standardization will be critical to select the patients most likely to benefit from treatment. For patients with immune evasive tumors, novel combinations under development show promise; however, global trials are needed.
E C Smyth, V Gambardella, A Cervantes, T Fleitas