Liposomal irinotecan, carboplatin or oxaliplatin (LyRICX) with or without nivolumab in the first-line treatment of metastatic or irresectable esophagogastric adenocarcinoma: A randomized phase 2 study.

Abstract

LBA287 Background: Oxaliplatin-based regimens are standard of care for metastatic/irresectable esophagogastric cancer, but can cause substantial neurotoxicity, limiting quality of life and eligibility for subsequent therapies. This study aims to identify the most favorable first-line chemotherapy backbone, balancing efficacy and neurotoxicity. Methods: This multi-center, open label, randomized phase II clinical trial enrolled adults with previously untreated pathologically confirmed metastatic/irresectable HER2 negative esophagogastric adenocarcinoma from 31 medical centers in the Netherlands. Until August 2022, patients were randomized (2:2:1) to one of three arms: 1) nanoliposomal irinotecan, leucovorin and fluorouracil (F-Nal-Iri); 2) capecitabine and carboplatin (CapCar); 3) capecitabine and oxaliplatin (CapOx). Following approval of nivolumab, patients with Combined Positive Score (CPS) <5 or contraindications for nivolumab were randomized to F-Nal-Iri, CapCar, or CapOx (2:2:1), while patients with CPS ≥5 were randomized to CapCar or CapOx (2:1) with nivolumab. Primary outcomes were grade 2-4 neurotoxicity and progression-free survival (PFS), with a predefined pick-the-winner strategy to select the most favorable regimen. Results: From September 2019-January 2025, 320 patients (median age 65y; 81% male) were randomized (F-Nal-Iri: 83; CapCar: 150 (74 with nivolumab); CapOx: 80 (36 with nivolumab)); median PFS follow-up 24.1 months. Grade 2-4 neurotoxicity occurred in 0 patients in the F-Nal-Iri arm, 4 patients (2.5%) in the CapCar +/- nivolumab arm, and 37 patients (46.2%) in the CapOx +/- nivolumab arm. Fishers exact tests showed no difference in neurotoxicity between CapCar and F-Nal-Iri (p=0.301), but significant differences between CapCar vs CapOx and CapOx vs F-Nal-Iri (both p<0.001). No clinically meaningful differences in other toxicities were observed. Median PFS was 4.5 (90%CI 4.14-6.34) months for F-Nal-Iri, 5.8 (90%CI 4.53-6.27) months for CapCar +/- nivolumab and 6.4 (90%CI 5.88-7.36) months for CapOx +/- nivolumab. In the subgroup without nivolumab, one-sided log rank tests and Cox models indicated no significant differences in PFS between F-Nal-Iri (4.5 months) and CapCar (5.7 months) (p=0.169, HR 1.17 (90%CI 0.890-1.534) or CapOx (5.9 months) (p=0.296, HR 0.89 (90%CI 0.636 -1.258). Between CapCar +/- nivolumab and CapOx +/- nivolumab, there was also no significant difference in PFS (p=0.118, HR 1.20 (90%CI 0.932-1.540)). Conclusions: Relative to CapOx, CapCar and F-Nal-Iri yielded markedly lower rates of grade 2-4 neurotoxicity with similar PFS and no excess of other toxicities. Given its ease of use—no central line required—and its relatively low cost (all drugs off-patent), CapCar can be considered the most favorable first-line chemotherapy backbone. Clinical trial information: 2023-509287-26-00 .