Nivolumab-Ipilimumab for Localized Microsatellite Instability–High Gastric/Gastroesophageal Junction Adenocarcinoma

Importance: Neoadjuvant nivolumab plus ipilimumab in localized deficient mismatch repair/microsatellite instability-high gastric or gastroesophageal junction (GEJ) adenocarcinoma was evaluated in the NEONIPIGA study. The primary end point of pathological complete response (pCR) was previously reported. Objective: To report long-term follow-up results and secondary end points, including event-free survival (EFS) and overall survival (OS), of the NEONIPIGA study. Design, Setting, and Participants: The NEONIPIGA study is a French, multicenter, single-arm, phase 2 nonrandomized clinical trial that included patients with histologically proven cT2-T4NXM0 deficient mismatch repair/microsatellite instability-high GEJ adenocarcinoma who were eligible for surgery. Patients were enrolled between October 23, 2019, and June 4, 2021, and the database was locked on November 13, 2025. Intervention: Neoadjuvant nivolumab, 240 mg, every 2 weeks for 6 cycles plus ipilimumab, 1 mg/kg, every 6 weeks for 2 cycles, followed by surgery 5 weeks after the last nivolumab dose, and adjuvant nivolumab, 480 mg, every 4 weeks for 9 cycles. Main Outcomes and Measures: EFS and OS. Results: A total of 32 patients (median [range] age, 65 [40-80] years; 8 [28%] female and 21 [72%] were male) were included: 16 (50%) with gastric cancer and 16 (50%) with GEJ cancer. Three patients did not undergo surgery (1 with metastatic disease and 2 who refused). Twenty-seven patients (84%) completed neoadjuvant therapy and 14 (44%) received all treatment as planned. Neoadjuvant and adjuvant therapy-related grade 3/4 adverse events occurred in 6 (19%) and 4 patients (17%), respectively. Median follow-up was 48.3 months (95% CI, 44.8-52.0 months). Among the 32 patients in the intention-to-treat population, 4-year OS was 84.1% (95% CI, 65.8%-93.0%). In the 31 patients eligible for surgery, 4-year EFS was 83.5% (95% CI, 64.8%-92.8%). One patient relapsed with cerebral metastases. Six deaths occurred: 2 related to gastric cancer or treatment (1 patient had postoperative complications and 1 cerebral metastases) and 4 unrelated. All patients who did not undergo surgery achieved a clinical complete response after immune checkpoint inhibitor therapy. One patient subsequently died of causes unrelated to their cancer. Overall, 17 of 29 patients (59%) achieved pCR on surgical specimen. Safety was previously reported, and no additional late toxic effects were observed. Conclusions and Relevance: In this phase 2 nonrandomized clinical trial, the immunotherapy regimen evaluated was associated with sustained long-term survival, but no clear evidence of improved survival over outcomes reported in other studies evaluating surgery alone in this patient population was observed. The high pCR rates and durable responses observed may provide useful information for designing future clinical studies, including randomized trials investigating organ-preserving strategies. Trial Registration: ClinicalTrials.gov Identifier: NCT04006262.