What is your approach for perioperative chemotherapy in MSI-H/dMMR localized gastric cancer?

This question gets at aspects of MSI-H biology, biomarker testing and post-hoc data from the MAGIC and CLASSIC trials suggesting a lack of benefit, and potential negative impact of treatment in MSI-H patients. The following is restricted to gastric cancer, and the rates of MSI-H in true esophageal and siewert I-II GEJ cancers are exceedingly low.

First, the post-hoc data raises the importance of MSI-H testing in stage II/III gastric cancer patients at diagnosis as has largely been adopted in colon cancer where MSI-H impact on adjuvant benefit has been more extensively studied. In a recently published post-hoc analysis of the Asian CLASSIC trial by Choi and colleagues the authors observed a 6.8% rate of MSI-H (8% in stage II, 5% in stage III)...

As ever, @Dr. First Last provides an exhaustive analysis of this topic and has the receipts/references to back it up.

I would approach this answer with a slightly different emphasis. Firstly, in addition to the references he cites, there is the individual patient data meta-analysis by Pietrantonio and colleagues, PMID 31513484 that collectively analyses the MAGIC, CLASSIC, ARTIST, and ITACA-S trials. This meta-analysis comprises phase III studies from Europe (MAGIC, ITACA-S) and East Asia (CLASSIC, ARTIST) and in the peri-operative (MAGIC) and adjuvant setting (the others). Out of 1,556 studies for whom MSI status was available, 7.8% had MSI tumors, in line with the studies Sam cites above. More importantly, this study shows very favorable o...