Claudin 18.2 expression—concordance between primary and metastatic lesions in gastric and gastroesophageal adenocarcinomas: a systematic review and meta-analysis

Background Claudin 18 isoform 2 (CLDN18.2) is a clinically validated therapeutic target in gastric and gastroesophageal adenocarcinomas. Biomarker assessment in pivotal trials has largely relied on primary tumor (PT) tissue, despite growing evidence of spatial heterogeneity between PTs and metastatic lesions (MLs). This systematic review and meta-analysis evaluated concordance of CLDN18.2 expression between paired PT and ML samples. Materials and methods PubMed, Web of Science, and Scopus were searched for studies reporting paired CLDN18.2 immunohistochemistry (IHC) results in PTs and MLs. Random-effects meta-analyses of proportions estimated overall discordance, PT-positive to ML-negative (PT + →ML - ) conversion, and PT-negative to ML-positive (PT - →ML + ) conversion. Agreement metrics, heterogeneity, subgroup analyses, and sensitivity analyses were conducted. Results Eleven studies comprising 896 paired PT-ML samples were analyzed. Overall, 20.2% of cases were discordant, with a pooled discordance proportion of 21% [95% confidence interval (CI) 18-24; I 2 = 35%]. Concordance beyond chance was moderate (Cohen's κ = 0.57). Significant asymmetry was observed, with PT + →ML - conversion occurring in 35.0% (95% CI 24.0-48.0) of PT-positive cases, compared with PT - →ML + conversion in 14.2% (95% CI 9.8-20.3) of PT-negative cases (McNemar P = 0.0047). Discordance was numerically higher in distant metastases and with higher IHC cut-offs. Findings were consistent across sensitivity analyses. Conclusions Findings indicate that CLDN18.2 expression demonstrates moderate spatial heterogeneity between PTs and MLs. Approximately one in five patients show discordance, most commonly reflecting loss of CLDN18.2 in MLs. These results suggest that single-site PT testing may not fully capture biomarker distribution across disease sites.