Design: global, randomised, double-blind, placebo-controlled, multicentre phase 3 trial, 147 centres in 20 countries
Population: 948 adults with untreated resectable stage II-IVa gastric or gastro-oesophageal junction adenocarcinoma, median age 62, 72% male
Primary outcome: overall survival in the intention-to-treat population (key secondary endpoint of the trial)
Effect: hazard ratio 0.78 (95% CI 0.63-0.96), P = 0.021; treatment-related deaths 6/475 vs 2/469

MATTERHORN randomised 948 adults with resectable stage II-IVa gastric or gastro-oesophageal junction adenocarcinoma, across 147 centres in 20 countries, to perioperative durvalumab plus FLOT chemotherapy or placebo plus FLOT. Event-free survival and pathological complete response had already been reported; this is the overall survival result, a key secondary endpoint. Overall survival improved with durvalumab: hazard ratio 0.78 (95% CI 0.63-0.96, P = 0.021).

For a clinician outside oncology, the point is what changes in the referral pathway. Perioperative chemotherapy for resectable gastric cancer has been FLOT for several years; adding a checkpoint inhibitor through the perioperative period and for ten further cycles is a longer, more complex treatment than a surgical referral used to imply. Treatment-related deaths were 1% against under 1%, so the safety cost is small but not zero.

The India-relevant caveat is access. Durvalumab is expensive and the regimen runs for around a year beyond surgery; whether it is available to a given patient is a different question from whether it works. Say that plainly at the point of referral rather than after the multidisciplinary meeting.

Refer resectable gastric and junctional adenocarcinoma before any chemotherapy is started — perioperative treatment begins preoperatively
Warn the patient that treatment now extends well past surgery, not just before it
Immune-related toxicity can present to any clinician: colitis, thyroiditis, pneumonitis and hepatitis in someone on a checkpoint inhibitor are drug effects until proven otherwise
Confirm funding and drug availability early — the regimen is long and interruption matters

Why it matters

Perioperative checkpoint blockade in this cancer had improved event-free survival without a survival result; now it has one.

The statistics, in plain English

A hazard ratio of 0.78 with an interval reaching 0.96 clears 1.0 only narrowly, and P = 0.021 was tested against a threshold of 0.0499 set aside for this endpoint — so the result is positive but not overwhelming. Overall survival was a key secondary endpoint here; the trial's primary endpoint, event-free survival, was reported previously.