Phase II study of ABSK061 (selective FGFR2/3 inhibitor) combined with ABSK043 (oral PD-L1 inhibitor) ± CAPOX in advanced FGFR2-positive gastric cancer and gastroesophageal junction cancer (GC/GEJC): Preliminary results from ABSK061-201.

Abstract

e16036 Background: Aberrant FGFR2 signaling is a validated driver associated with poor prognosis in advanced GC/GEJC. FGFR2b overexpression is observed in 16%-30% of patients in this tumor type, while approximately 5%–10% of GC cases have FGFR2 gene amplification. ABSK061 is a highly selective oral FGFR2/3 inhibitor designed for treating GC/GEJC. Methods: This ongoing phase II study (NCT06632262) is to evaluate the safety and preliminary efficacy of ABSK061 plus ABSK043 with or without chemotherapy. In dose escalation cohort and expansion cohort 4, patients with solid tumors including pretreated (2L+) GC/GEJC with FGFR2b overexpression and/or FGFR2 amplification, were enrolled and received ABSK061 plus ABSK043. Expansion cohort 1 specifically enrolled treatment-naïve (1L) advanced HER2- GC/GEJC patients with FGFR2b overexpression and/or FGFR2 amplification, who received ABSK061 plus ABSK043 and CAPOX. Results: As of January 2026, a total of 37 patients (24 2L+, 13 1L) were enrolled with ABSK061 75mg BID and ABSK043 800mg BID ± CAPOX. No dose-limiting toxicities (DLTs) were observed. Treatment emergent adverse events (TEAEs) occurred in 97.3% of patients, of which 51.4% were Grade ≥3. The most common TEAEs included increased aspartate aminotransferase (54.1%), increased alanine aminotransferase (45.9%), and anaemia (43.2%). In expansion cohort 1, TEAEs occurred in 92.3% of patients, of which 53.8% were Grade ≥3. The most common ( > 10%) Grade ≥3 TEAEs included platelet count decreased (23.1%), hypokalaemia (15.4%) and neutrophil count decreased (15.4%), which were most likely attributable to chemotherapy. Across cohorts, all FGFR-related AEs including ocular events (2.7% grade 3), stomatitis (2.7% grade 3) and nail disorder ( all grade 1-2) were reversible and well managed through dose modifications without leading to permanent discontinuation. Among 10 treatment-naïve patients with GC/GEJC who received at least 1 cycle of treatment, all patients showed target-lesion shrinkage and remain on treatment, 9 achieved partial response (PR) per RECIST v1.1, yielding an objective response rate (ORR) of 90%. Among 16 evaluable pretreated (2L+) patients with GC/GEJC, 5 achieved PR (ORR: 31.3%), with the longest duration of treatment being 7.6 months. Conclusions: ABSK061 combined with ABSK043, with or without CAPOX, has demonstrated a manageable safety profile and encouraging anti-tumor activity in both 1L and 2L+ FGFR2-positive GC/GEJC. These promising results warrant further clinical development of the doublet and quadruplet regimens for GC/GEJC. Clinical trial information: NCT06632262 .