Safety run-in results from C-SOLVE (JACCRO GC-12): A randomized, non-comparative phase II trial of CapeOX or SOX plus zolbetuximab in patients with HER2-negative, CLDN18.2-positive unresectable advanced gastric or GEJ cancer.

Abstract

352 Background: C-SOLVE (JACCRO GC-12; jRCTs031240510) is an ongoing randomized, non-comparative phase II trial investigating first-line treatment for patients with HER2-negative, CLDN18.2-positive (≥75% of tumor cells with moderate-to-strong membranous staining), unresectable or recurrent gastric or gastroesophageal junction (GEJ) cancer. Zolbetuximab (Zol), a CLDN18.2-targeted monoclonal antibody, has shown efficacy with capecitabine plus oxaliplatin (CapeOX) in global studies, but has not been evaluated with S-1 plus oxaliplatin (SOX), the most used chemotherapy backbone for gastric cancer in Japan. C-SOLVE was designed to assess the safety and efficacy of SOX+Zol, using CapeOX+Zol as a reference. A safety run-in was conducted prior to full enrollment (target N=140). Methods: Eligible patients had HER2-negative, CLDN18.2-positive unresectable or recurrent gastric/GEJ cancer, ECOG performance status 0–1, measurable disease (RECIST v1.1), no prior systemic therapy for advanced disease, and adequate organ function. Zol was administered at 800 mg/m² on day 1 of cycle 1, followed by 600 mg/m² on day 1 of subsequent 3-week cycles. In the safety run-in phase, the first 12 patients in each treatment arm were evaluated for the incidence of nausea, vomiting, and other key adverse events (AEs), including any serious AE, grade ≥3 AE, or AE leading to treatment discontinuation or death within 3 months of initiation. An independent data monitoring committee reviewed the safety data. Results: Among the first 12 patients treated with SOX+Zol, grade ≥3 nausea and vomiting were observed in 3 (25.0%) and 1 (8.3%) patients, respectively. In the CapeOX+Zol arm, these events occurred in 1 (8.3%) and 0 patients, respectively. The use of corticosteroids, 5-HT3 receptor antagonists, and NK-1 receptor antagonists were generally comparable between the two groups. Other notable grade ≥3 AEs in the SOX+Zol arm included anorexia (n=4), peripheral sensory neuropathy (n=3), hypoalbuminemia, hypokalemia, diarrhea, and fatigue (each n=2). In the CapeOX+Zol arm, grade ≥3 anorexia (n=3) and diarrhea (n=2) were reported, along with one case of grade 4 hyperamylasemia. One patient in the SOX+Zol group discontinued treatment due to grade 2 cognitive impairment and grade 3 COVID-19 infection. No treatment-related deaths occurred in either arm. Conclusions: The safety run-in analysis of C-SOLVE demonstrated manageable and comparable toxicity profiles between SOX+Zol and CapeOX+Zol. No unexpected or unacceptable safety signals were identified. Based on these findings, the independent data monitoring committee recommended continuation of the study as planned. As of August 25, 2025, 87 patients have been enrolled. Accrual is ongoing at 110 JACCRO-affiliated sites across Japan. Clinical trial information: 031240510 .