Biomarkers of response in patients with gastric/gastroesophageal junction adenocarcinoma (GEA) treated with telisotuzumab adizutecan.

c-Met protein (MET protein) overexpression (OE) and amplification (amp) occur in several cancer types, including GEA, and are associated with poor prognosis. The c-Met–directed antibody-drug conjugate telisotuzumab adizutecan (ABBV-400) is being assessed in advanced solid tumors in an ongoing, first-in-human study (NCT05029882). Results showed promising antitumor activity and a tolerable safety profile in GEA (Strickler et al. ESMO 2024, 1439P). Here, we present our analysis on associations between c-Met OE, amp, and ctDNA-derived molecular response with clinical outcomes. c-Met expression was evaluated retrospectively in FFPE tissue (archival/fresh) collected at study screening using IHC (SP44 Ab clone) or whole transcriptomic RNA sequencing (Caris). Plasma ctDNA collected at baseline (BL) and cycle 3 day 1 (C3) was analyzed with the GuardantINFINITY assay. Molecular response (MR) was assessed using circulating tumor fraction (cTF), estimated on the basis of ctDNA variant allele frequencies in a 74-gene panel (Guardant360 CDx) and methylation signals (GuardantINFINITY methylome platform). MR was defined as a 50% decrease in cTF from BL. amp status was determined retrospectively by the Guardant Health algorithm. Radiographic response was assessed per RECIST v1.1. Clinical data from 42 patients (pts) with advanced GEA were included in this biomarker analysis. There was a negative correlation between c-Met expression and tumor size change from BL (R=-0.59). Of 42 pts, 12 (29%) had tumors with amp per ctDNA analysis/local site reports, which was higher than the bioinformatic analysis of real-world data (Caris-ConcertAI linked RWD360 and PT360; ~3%). Concordance between c-Met expression per IHC and RNA was observed (R=0.75, p<10 ). There were 12 responders; 7 (58%) had amp, and 5 (42%) did not. Responders had lower ctDNA percentage change from BL vs nonresponders (1.5-fold and 1.6-fold difference based on 74-gene and methylation panel, respectively). Pts with MR had a greater decrease in tumor size from BL vs those without MR. A positive correlation between change in ctDNA levels (BL to C3) and change in tumor size was observed using the 74-gene panel and methylation panel (R=0.73 and 0.69, respectively). Additional biomarker analyses focused on the association between and other key biomarkers relevant in GEA will be discussed. In pts with advanced GEA treated with telisotuzumab adizutecan, c-Met OE/ amp and ctDNA-based MR were associated with enhanced antitumor activity. Compared with real-world data, amp was enriched in this study population. Phase 2 studies will continue to explore c-Met OE and amp prevalence and their potential predictive value with other biomarkers in a clinically representative population. .

Duke Scholars

Author John Strickler Medicine, Medical Oncology