For patients with advanced gastric/GEJ cancer expressing PDL1, is there data to support frontline pembrolizumab?

I agree with Dr. @Dr. First Last's comments regarding the Keynote 059, 061, and 062 studies.

However, there are new data presented at the 2020 ESMO meeting.

The Keynote 590 is a phase III study enrolled 749 patients with advanced esophageal and GEJ cancer (including esophageal squamous cell carcinoma and adenocarcinoma, also GEJ Siewart type I adenocarcinoma) for first-line treatment. The patients were treated with chemotherapy (fluorouracil and cisplatin up to 6 cycles) with/without pembrolizumab (200mg q3week, up to 35 cycles) irrespective of the PD-L1 expression level. The median overall survival was 12.4 months for the pembrolizumab/chemo arm compared to 9.8 months for the chemo only arm. One year survival rate was 51% vs. 28%, two-yea...

Like tweets from some of our politicians, my previous answer has not aged well. I absolutely agree with Dr. @Dr. First Last that both Checkmate-649 and KEYNOTE-590 have transformed the treatment of esophagogastric cancer (EGC; both adenocarcinoma and SCC). Broadly speaking, there are now phase III data to add an anti-PD-1 antibody (nivolumab or pembrolizumab) to any EGC.

I think it's pretty unambiguous that any esophageal SCC benefits from the addition of pembrolizumab to a fluoropyrimidine/platinum doublet (the KEYNOTE-590 study used the dreaded 5-FU/cisplatin q3W regimen but I have every intention of combining it with FOLFOX). As Dr. Jin suggested, it remains unclear exactly what group of patients with esophagogastric adenocarcinoma benef...

There are no phase III data. The only data so far come from one arm of the KEYNOTE-059 study, where 25 Pts were treated with 5-FU/cisplatin and pembrolizumab (Bang, J Clin Oncol 2017;35:4012 [abstr]). The response rate was 60% and survival data were encouraging. However, in the absence of phase III data, I could not recommend this as a first-line regimen. The KEYNOTE-062 study of first-line therapy in patients with PD-L1 positive tumors with pembrolizumab vs. 5-FU/cisplatin vs. pembro/-FU/cisplatin has completed accrual and results should be reported in the next year. Checkmate 649, which evaluates FOLFOX vs. nivolumab/FOLFOX vs. ipilimumab/FOLFOX in all-comers, will shortly complete accrual. Therefore, there will be definitive phase III da...

I will be in the minority, but I am not sure the available data is convincing to use upfront immunotherapy in most patients with esophageal cancer.

1. KN590 really just mimics the results of KN181: there is some benefit for patients with esophageal SCC with CPS>10.

The benefit in the overall population in KN590 seems to be driven by SCC tumors (3/4 of patients on the trial), the presented Forest plots do not show a benefit for adenocarcinomas and CPS<10 tumors. I believe there is a reason BMS did not show survival curves for adenoca with CPS>10 during ESMO 2020.

2. CM649 was designed as IPI/NIVO vs FOLFOX in pts with TPS>1%. The trial evolves over time and we see the result for FOLFOX vs FOLFOX+nivo in CPS>5% with accrual increased beyond ...