Background: The tight junction protein CLDN1 is a potential therapeutic target in many cancers where its dysregulation is associated with invasiveness and migration. The effect of CLDN1 expression on outcomes in BTC is unknown. We examined the molecular and clinical correlates of CLDN1 expression in a real-world cohort of patients (pts) with advanced BTCs, including across subtypes.

Methods: We used Tempus Lens, a platform used to query multimodal data from millions of de-identified patient records in the Tempus Database to identify 1,897 pts with BTCs whose tumors were tested with Tempus xT and xR panels. RNA-seq data were normalized by computing transcripts per million (TPM) and transforming them by log2(TPM+1). PD-L1 status was assessed with the 22C3 clone and immune cell infiltration was estimated using quanTIseq. Pts were classified into CLDN1-high (CLDN1-H) vs CLDN1-low (CLDN1-L) by median CLDN1 expression. P-values were calculated by either χ2 or Fisher’s exact tests. Real-world overall survival (rwOS) was defined as the time from the initiation of first-line (1L) treatment with chemotherapy(chemo) +/- immunotherapy (IO) (N = 919) to the earliest of death, last known follow up, or max 5 years. Median rwOS (mOS) was estimated with Kaplan-Meier curves and log-rank p-value. Statistical significance was set at p≤0.05; all tests were two-sided.

Results: The 1,897 patient cohort was comprised of 51% IHCC, 14% EHCC, 22% GB, 13% CCA NOS. Median age 65 yrs; 53% female. IHCC tumors had more frequent CLDN1-H expression (70%), while EHCC (81%) and GB (78%) tumors had more frequent CLDN1-L expression (p < 0.001). CLDN1-H tumors had more frequent FGFR2 fusions (16% vs 2.6%) and IDH1 (18% vs 3.1%), PBRM1 (12% vs 5%) and BAP1 mutations (19% vs 5.3%); CLDN1-L tumors had more frequent KRAS (25% vs 8.2%) and TP53 mutations (60% vs 25%), ERBB2 amplification (6.8% vs 1.5%) (all p < 0.05). CLDN1-H expression was associated with a significantly higher percentage of M1 macrophages and NK cells and lower M2 macrophages, CD8 T-cells, Tregs and neutrophils (all p < 0.001). There was no difference in PDL1 (22c3) status by CLDN1 expression. Pts with CLDN1-H tumors had a trend towards improved mOS vs those with CLDN1-L tumors (12.2 vs 10.2 m, p = 0.094). Among pts receiving 1L chemo+IO, CLDN1-H expression was associated with significantly improved mOS (14.8 vs 10.0 m, p = 0.0464); no difference in outcomes by CLDN1 expression in pts receiving chemo alone.

Conclusions: High CLDN1 expression appears to be associated with immune cell infiltration in this cohort of advanced BTC, and with improved survival in pts treated with 1L chemo+IO. Furthermore, relevant molecular alterations in BTC seem to differ with high vs low CLDN1 expression. Larger studies are warranted to evaluate the predictive and prognostic role of CLDN1 in BTC to identify novel therapeutic strategies.

VIEW THE PUBLICATION