Final survival results of ipilimumab or FOLFOX in combination with nivolumab and trastuzumab in previously untreated HER2 positive esophago gastric adenocarcinoma: The randomized AIO INTEGA trial.

Abstract

4026 Background: In metastatic esophagogastric adenocarcinoma (EGA), the addition of PD-1 inhibitors (i) to chemotherapy has improved the outcome in selected patient populations. The randomized INTEGA trial investigated trastuzumab and PD-1i with FOLFOX or CTLA-4i in 1st line treatment of advanced HER2+ EGA. Methods: Patients with previously untreated, metastatic HER2+ (local IHC3+ or 2+/ISH+) EGA, and adequate organ function were randomized to trastuzumab and nivolumab (1mg/kg Q3W x4 /240mg Q2W for up to 12 months) in combination with ipilimumab (3mg/kg x4 Q3W; Ipi arm) or mFOLFOX6 (FOLFOX arm) accompanied by a large translational program. The primary endpoint was the 12month overall survival (OS) rate. The trial was registered at ClinicalTrials.gov, NCT03409848. Results: Between March 2018 and May 2020, 97 patients were enrolled and 88 randomized across 21 German sites with the following baseline characteristics: female/male 18/70, median age 60.5 (range 41-80), ECOG 0/1 54/34, GEJ/stomach 66/22. Central post hoc biomarker analysis on 82 evaluable patients showed PD-L1 CPS≥1/≥5 in 59/46 patients and confirmed HER2 (central IHC3+ or 2+/ISH+) positivity in 77 patients. The observed OS rate at 12 months was 70% (95% confidence interval (CI) 54-81%) in the FOLFOX arm, and 57% (95% CI 41-71%) in the Ipi arm. The progression free survival was 10.7 and 3.2 months, and the overall response rate was 53.5% and 34% in the FOLFOX and the Ipi arm, respectively. Although these data are in favor of the FOLFOX arm, the overall survival curves crossed with increased follow-up (median follow-up of 18.8 months) and thus the final median OS was 22.1 and 23.3 months, numerically favoring the Ipi arm. According to the centrally assessed CPS the OS was: 22.1 and 32.2 months in the CPS<5 group, and 22.7 and 12.6 months in the CPS≥5 group for the FOLFOX and the Ipi arm, respectively. Notably, the median tumor burden calculated as the sum of target lesions was numerically lowest in the Ipi arm with higher than median OS. Conclusions: In contrast to limited short term efficacy of the Ipi arm as reflected by the lower PFS, the Ipi arm showed favorable OS, that was inversely correlated to the CPS status. Further analysis of baseline clinical and molecular data to better identify the subgroup of patients benefitting the most of trastuzumab/nivolumab/ipilimumab will be presented at the meeting. Clinical trial information: NCT03409848 . [Table: see text]