Background
Phase III double-blind RCT, N=948, patients with resectable gastric or gastroesophageal junction (GEJ) adenocarcinoma (stage ≥T2 or N+). Stratified by region, PD-L1 CPS, and tumor location. FLOT chemotherapy (docetaxel 50 mg/m², oxaliplatin 85 mg/m², leucovorin 200 mg/m², 5-FU 2600 mg/m² q2wk) given perioperatively (4 pre-op + 4 post-op cycles). Enrolled 2019–2023 across Europe, Asia, and North America.
Interventions and follow up
Arm A: Durvalumab 1500 mg q3wk + FLOT (perioperative, 4+4 cycles), then durvalumab 1500 mg q4wk maintenance for up to 12 cycles post-surgery
Arm B: Placebo + FLOT (perioperative, 4+4 cycles), then placebo maintenance
Primary endpoint: Event-free survival (EFS)
Follow-up: 33.2 mo
Arm B: Placebo + FLOT (perioperative, 4+4 cycles), then placebo maintenance
Primary endpoint: Event-free survival (EFS)
Follow-up: 33.2 mo
Results
EFS: HR 0.71, 95% CI 0.57–0.87, P=.0008
OS: HR 0.78, 95% CI 0.61–0.99, P=.021 (ESMO 2025 final analysis)
pCR rate: 19% vs 7% (Arm A vs B)
R0 resection: 78% vs 75%
OS: HR 0.78, 95% CI 0.61–0.99, P=.021 (ESMO 2025 final analysis)
pCR rate: 19% vs 7% (Arm A vs B)
R0 resection: 78% vs 75%
Adverse events
Grade ≥3 (any): 76% (Arm A) vs 73% (Arm B); treatment discontinuation due to AE 20% vs 12%; surgical complication rates comparable between arms.
Immune-mediated: any grade 34% vs 10%, grade ≥3 8% vs 2%; most common were hypothyroidism, rash, hepatitis.
Immune-mediated: any grade 34% vs 10%, grade ≥3 8% vs 2%; most common were hypothyroidism, rash, hepatitis.
Conclusions
Perioperative durvalumab + FLOT followed by durvalumab maintenance significantly improved EFS and OS compared to FLOT alone in resectable gastric/GEJ adenocarcinoma, establishing a new perioperative immunotherapy standard. The pCR rate nearly tripled with durvalumab addition (19% vs 7%), suggesting meaningful pathologic downstaging.
Key Limitations
EFS is a composite endpoint (recurrence, progression, death) that can be driven by early events; the OS benefit, while significant at ESMO 2025, requires longer follow-up for full characterization. PD-L1 CPS did not clearly define a predictive subgroup — benefit was observed broadly — raising questions about biomarker-guided selection. Post-operative durvalumab maintenance adds 12 cycles beyond FLOT; the relative contribution of perioperative vs maintenance dosing is unclear. Applicability to regions where FLOT is less standard (e.g., parts of Asia where doublets predominate) is uncertain.
Clinical Context
FDA approved durvalumab (Imfinzi) + FLOT for resectable gastric/GEJ adenocarcinoma in 2025, the first immunotherapy approval in the perioperative resectable setting. This adds to a rapidly evolving landscape where nivolumab is established in metastatic first-line (CheckMate 649) but did not show benefit in the adjuvant setting (ATTRACTION-5). MATTERHORN positions durvalumab + FLOT as a standard-of-care option for perioperative-eligible gastric/GEJ patients globally; ESMO endorsement is anticipated.