Expression Profiling of Claudin 18.2 in Resected Specimens and Endoscopic Biopsies from Patients with Gastric and Gastroesophageal Junction Adenocarcinoma: A Review

Abstract

Claudin 18.2 is now a treatment-defining biomarker in gastric and gastroesophageal junction adenocarcinoma, and eligibility turns on a single number: whether at least 75 per cent of tumour cells stain moderately or strongly. That threshold is where the difficulty lies, because expression within a tumour is uneven and a small endoscopic sample may not represent the whole. Objectives: To review the published evidence on claudin 18.2 expression in gastric and gastroesophageal junction adenocarcinoma: reported prevalence,
agreement between biopsies, resected specimens and metastatic sites, the influence of the positivity threshold on that agreement, and the treatment outcomes that follow a positive result. Material and Methods: Narrative review of randomised trials, trial screening and correlative analyses, and matched-sample immunohistochemistry studies indexed in PubMed and the major oncology and gastroenterology journals. Studies were eligible if they reported claudin 18.2 prevalence in a defined population, concordance between two sample types in the same patients, or outcomes stratified by claudin 18.2 status. No new patient data were generated; every prevalence, concordance and survival figure quoted is one published by the original investigators. Results: Across the screening populations of two phase 3 trials, positivity at the 75 per cent threshold was 38.4 per cent, and was similar between gastric
and junctional primaries, between biopsy and resection, and between primary and metastatic samples. Independent series have reported 47.0 and 24.0 per cent. Concordance depends strongly on the threshold applied: between archival and baseline samples it was 61.1 per cent at the 75 per cent cutoff but 88.9 per cent for any staining. Intratumoral heterogeneity was found in 38.5 per cent of surgical specimens and 25.2 per cent of paired primary-metastatic comparisons were discordant, with peritoneal lesions showing the highest positive concordance at 31.4 per cent and liver lesions the lowest at 12.8 per cent. Adding zolbetuximab to first-line chemotherapy extended median overall survival from 15.5 to 18.2 months in one trial and from 12.2 to 14.4 months in the other. Conclusion: Prevalence is stable enough across sample types to make biopsy-based testing reasonable, but concordance in individual patients is not high, and it falls as the threshold rises. Because eligibility rests on a dichotomised cutoff applied to a continuously distributed and spatially uneven marker, a negative result from a small or single-site sample should not be treated as final where the therapeutic question remains open. Abbreviations: CLDN18.2 – Claudin 18 Isoform 2, G/GEJ – Gastric or Gastroesophageal Junction, IHC – Immunohistochemistry, HER2 – Human Epidermal Growth Factor Receptor 2, PD-L1 – Programmed Death-Ligand 1, CPS – Combined Positive Score, MSI – Microsatellite Instability, dMMR – Deficient Mismatch Repair, PFS – Progression-Free Survival, OS – Overall Survival, HR – Hazard Ratio, FFPE – Formalin-Fixed Paraffin-Embedded.