A phase 2 pivotal study of savolitinib in patients with <i>MET</i> -amplified gastric cancer or gastroesophageal junction adenocarcinomas.

Abstract

4011 Background: MET amplification is a poor prognosis for survival in patients (pts) with gastric cancer (GC) or gastroesophageal junction adenocarcinoma (GEJa). Currently no approved targeted therapy is available for MET-amplified GC/GEJa. Savolitinib is a potent and highly selective oral MET-TKI. Here we reported the primary analysis results from a pivotal phase 2 trial investigating savolitinib in MET-amplified GC/GEJa, conducted with registration intent (NCT04923932). Methods: Eligible pts had locally advanced or metastatic GC/GEJa with MET amplification (gene copy number ≥10 by FISH) and had failed ≥2 lines of prior standard therapy. Pts received savolitinib 200 mg (BW &lt; 50 kg) or 300 mg (BW ≥50 kg) BID. The primary endpoint was ORR assessed by an Independent Review Committee (IRC) per RECIST 1.1, with a pre-specified efficacy threshold defined as the lower limit of the 95% CI of ORR exceeding 15%. Secondary endpoints mainly included DCR, DoR, TTR, PFS, OS and safety. An exploratory OS analysis was conducted in pts with MET-amplified GC/GEJa, who had previously received first-line, second-line, or later-line standard therapies, but were ineligible for study enrollment and did not receive the study drug. Plasma samples were also collected at baseline and end of treatment (EOT) for comprehensive genomic profiling using a 671-gene sequencing panel (Benrui GCP, OrigiMed). Results: As of Oct 8, 2025, 65 eligible pts were enrolled and received savolitinib. Baseline characteristics were: median age of 57.4 years, ECOG PS of 0 (13.8%) or 1 (86.2%), primary tumor sites of GC (84.6%) or GEJa (15.4%), and baseline ctDNA MET-positive rate of 64.6% (42 pts). Per the IRC assessment, ORR was 32.3% (95%CI: 21.2%, 45.1%), which exceeded the pre-specified efficacy threshold; in baseline ctDNA MET-positive pts (42 pts), IRC-assessed ORR was 45.2% (95%CI: 29.8%, 61.3%). IRC-assessed secondary efficacy endpoints were: DCR of 63.1%; mTTR of 1.4 months; mDoR of 9.7 months; and mPFS 4.0 of months. mOS was 6.9 months in the 65 treated pts, versus 4.8 months in 139 pts who received alternative treatments rather than the study drug in the exploratory OS analysis. Savolitinib-related TEAEs of Grade ≥3 occurred in 23 pts (35.4%). Exploratory ctDNA biomarker analysis showed that pts (N = 14) harboring baseline alterations in FGFR2, EGFR, PIK3CA, BRAF, or KRAS had lower ORR (14.3%, 2/14) and shorter mPFS (1.4 months). Additionally, in pts (N = 26) who provided PD/EOT samples, 12 pts developed alterations in these genes that were identified as putative acquired resistance mechanisms, and 4 pts had acquired MET mutations. Conclusions: Savolitinib showed clinical efficacy and a tolerable safety profile in pts with MET-amplified GC/GEJa who had received ≥2 lines of prior therapy, supporting it to be a future treatment option for this genetically defined population. Clinical trial information: NCT04923932 .