How would you manage locally advanced, resectable gastric cancer that is MSI-H?

I would not recommend FLOT + durvalumab for MSI-H gastric/GEJ patients. I say this because there is mounting evidence that chemotherapy may not provide a significant benefit for these MSI-H patients at all (1). All efforts should be made to spare this patient population from getting cytotoxic chemotherapy. I would even make the argument that we may be doing harm to this patient population by giving them chemotherapy.

We also have evidence showing the dramatic responses these patients receive from immunotherapy alone. The GERCOR NEONIPIGA phase II trial showed a 60% pathologic complete response; most patients did not have any cancer left at the time of surgery after exposure to immunotherapy (2). Anecdote to my practice, I have never had a lo...

This is a very good question and lives in a largely "data-free zone" at this time. A recent meta-analysis shows that a number of small studies have been reported with an overall beneficial effect, including pCR (Schwengber et al., PMID 41687159). Based on our experience with rectal cancer (Cercek et al., PMID 40293177) and colon cancer (NICHE trials), use of dual ICI seems like a very justifiable approach, with far less toxicity than FLOT (even without Durva). I believe this can be readily discussed with appropriate patients and is a viable option pending randomized trials of ICI vs Chemo-antiPD1.

This is a timely question, as I currently have several patients in this (highly favorable) situation.

What surprised me is the outcomes of these patients in MATTERHORN, as described in the published manuscript (Janjigian et al., PMID 42716074). As has been pointed out, there were only 25 patients with MSI tumors who received durvalumab/FLOT. Their pathologic complete response (CR) rate was only 28%, and only 68% of patients had event-free survival at 18 months. While these are obviously very small numbers, there does not appear to be any indication that the addition of chemotherapy to immunotherapy is beneficial compared to immunotherapy alone.

At this time, the options for preoperative immunotherapy include ipilimumab/nivolumab (Andre et al....

To echo Dr. @Dr. First Last excellent commentary, I think this is a largely evidence-limited zone where decisions have to be patient-centered. Factors that I take into account in these discussions include patient performance status, co-morbid conditions, and the type of surgery that may need to be offered (a subtotal versus total gastrectomy has very different functional outcomes).

The original data from MAGIC (Smyth et al., PMID 28241187) strongly suggested that preoperative chemotherapy was not beneficial, and even potentially harmful for MSI-H patients with gastroesophageal cancer. I think that this suggests that FLOT Durvalumab is unlikely to be beneficial, as the immunosuppressive effect of chemotherapy in this context may override the ...

Chemotherapy will likely add more toxicity than benefit. If PD1 monotherapy is not effective, can switch to dual; if dual is not working, would try NK-based immunotherapy.