Prevalence of claudin 18.2 (CLDN18.2) and the association of biomarkers with clinical activity in patients with gastric or gastroesophageal adenocarcinoma (G/GEJa) treated with zolbetuximab.

466 Background: CLDN18.2, a targetable biomarker, is a tight junction protein normally confined to gastric mucosa of healthy tissue and often retained in G/GEJa. Zolbetuximab, a chimeric IgG1 monoclonal antibody, binds to CLDN18.2 and mediates cancer cell death via antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC). Phase 2 FAST study results showed prolonged survival with zolbetuximab+EOX (epirubicin, oxaliplatin, capecitabine) vs EOX in CLDN18.2-positive G/GEJa. In ILUSTRO Cohort 2, confirmed partial responses (PRs) and stable disease (SD) were previously reported in 63.2% and 10.5%, respectively, of patients (pts) treated with zolbetuximab+mFOLFOX6 (modified 5-FU, folinic acid, oxaliplatin). Here is an analysis of CLDN18.2 expression and exploratory biomarkers from 2 studies of zolbetuximab alone or with mFOLFOX6 in G/GEJa. Methods: Data are from phase 1 (NCT03528629) and phase 2 (NCT03505320, ILUSTRO Cohorts 1 and 2) studies. Pts with CLDN18.2-positive locally advanced/metastatic G/GEJa received zolbetuximab alone (phase 1 [ineligible for SOC], n = 18; phase 2 Cohort 1 [≥3rd-line], n = 30) or with mFOLFOX6 (phase 2 Cohort 2 [1st-line], n = 21). Archival (any time before treatment), baseline (during screening or ≤3 months before first study treatment), and on-treatment (Cycle 3) tumor samples were collected when possible. CLDN18.2 expression and immune cell populations were assessed by IHC. Blood samples collected longitudinally while on treatment (phase 2 study) were analyzed for ADCC activity (ex vivo cell-based assay), circulating tumor antigens, and circulating tumor DNA (ctDNA). Results: Of 416 screened pts in both studies, 26.2% (n = 109), 9.6% (n = 40), and 36.1% (n = 150) had tumors that were CLDN18.2-positive in 1%–49%, 50%–74%, and ≥75% of tumor cells ( = strong to moderate staining intensity), respectively; 28.1% (n = 117) were CLDN18.2-negative (0/1+). Analysis of pre- and on-treatment tissue samples (n = 8 matched pairs) with zolbetuximab alone showed an increased trend in immune cell infiltration, with CD8+ T cells and CD163+ cells showing the most consistent trend. A trend of increased on-treatment vs baseline ADCC activity was observed in PBMCs collected from pts treated with zolbetuximab alone and with mFOLFOX6. Rapid and deep decreases in circulating tumor antigens (in pts with elevated antigen levels) and ctDNA were observed with best overall responses of both PR and SD. Conclusions: CLDN18.2 is a high-prevalence biomarker in G/GEJa. Treatment with zolbetuximab is associated with tumor and peripheral biomarker changes related to its proposed mechanism of action. Clinical responses observed in pts treated with zolbetuximab and mFOLFOX6 are associated with correlative biomarkers of activity including rapid deep molecular responses (ctDNA). Clinical trial information: NCT03528629 .