Top Line: Immune checkpoint inhibitors have ousted chemo completely in some disease sites, perhaps most notably in certain non-small cell lung cancers.
The Study: KEYNOTE-062 looked to establish the same with pembrolizumab for advanced gastroesophageal cancers by randomizing 763 patients 1:1:1 to first line  pembro,  pembro + chemo (cisplatin/5FU/capecitabine), or  chemo alone. A key to understanding the results is understanding the “PD-L1 combined positive score (CPS),” which is the number of cells (tumor and immune) staining positive for PD-1 / PD-L1 divided by the total number of viable tumor cells multiplied by 100. There were multiple primary endpoints, including establishing non-inferiority of the pembro monotherapy arm and superiority of the pembro + chemo arm relative to the chemo alone arm in regards to overall survival (OS) with a CPS ≥1 or with a CPS ≥10. In the end, pembro monotherapy indeed achieved non-inferior OS in patients with a CPS ≥1 (11 months with either pembro or chemo) as well as with a CPS ≥10 (17 months with pembro and 11 months with chemo). Pembro + chemo did not prove superiority for OS with CPS ≥1 (13 versus 11 months, respectively) nor with CPS ≥10 (12 versus 11 months). Most notably, rates of grade 3-5 toxicity were 69-73% with chemo versus 17% without.TBL: Pembro monotherapy for advanced gastroesophageal cancers appears to achieve survival outcomes non-inferior to those with traditional chemo, especially among those with greater PD-L1 staining, with substantially less toxicity. | Shitara, JAMA Oncol 2020