Correlation of immune remodeling linked to obesity with immunotherapy outcomes in cholangiocarcinoma: An externally validated international cohort study.

4134 Background: Obesity is an established risk factor for cancer, yet paradoxically may be associated with improved responses to immune checkpoint inhibitors (ICIs). We investigated whether obesity defines a clinically and biologically distinct cholangiocarcinoma (CCA) subtype when treated with ICIs. Methods: Retrospective analysis of Body Mass Index (BMI)-stratified patients (pts) with CCA treated at a single US center (US cohort) and enrolled in a prospective multicenter Italian registry (Italian cohort). High BMI was defined per CDC criteria (overweight 25.0-29.9; obesity ≥30 kg/m²), with cohort-specific cutoffs for analysis (US ≥30; Italy ≥25). Tumor microenvironment (TME) was profiled with bulk RNA-seq and CODEX multiplex imaging and body composition radiomics with TotalSegmentator tool in the US cohort. Chi-square, Pearson correlation or Wilcoxon tests were used to test associations with BMI; overall survival (OS) was assessed with Kaplan-Meier and Cox models. Results: 989 pts were included (661 US; 328 Italy). In the US and Italian cohorts, median age was 62 and 65 years, 51% and 47% were female, and 28% and 43% were obese and overweight, respectively. Metabolic comorbidities included diabetes (US: 15%; Italy: 19%), hypertension (US: 48%; Italy: 45%), and hyperlipidemia (US: 32%; Italy: 15%). Higher subcutaneous-to-visceral fat ratio was associated with improved OS with chemotherapy-ICI (HR 0.89; 95% CI 0.79-0.99; p = 0.036). In both cohorts, patients with high BMI treated with chemotherapy-ICI had significantly longer OS compared with lower-BMI patients (US: 31.6 vs 19.2 months, p = 0.005; Italy: 21 vs 12 months, p = 0.025). In multivariable models (US cohort) both obesity (HR 0.35; 95% CI 0.23-0.52, p < 0.0001) and higher visceral-to-subcutaneous fat ratio (HR 1.79; 95% CI 1.04-3.07, p = 0.037) were independently associated with OS. Combined genomic profiling (n = 906) showed a lower prevalence of KRAS alterations in high-BMI pts (OR 0.56, p = 0.006). RNA-seq (n = 86; obese = 37, non-obese = 49) demonstrated enrichment of fibroblast-associated signatures ( αSMA, PDPN, collagen IV, vimentin ), upregulation of immune checkpoints ( PD-1, PD-L1, TIGIT, LAG-3, VISTA ), and higher GATA3 and IFN-γ expression in obese pts (both p < 0.05). Multiplex CODEX analysis (n = 38; obese = 17, non-obese = 21) revealed an increased co-expression between GATA3 , a regulator of T-cell dysfunction, and granzyme B , a cytotoxic T-cell marker (Pearson ρ = 0.24, p < 0.05), indicating increased abundance of GATA3 ⁺granzyme B⁺ cells in obese pts. Conclusions: In CCA, obesity characterized by predominant subcutaneous adiposity is linked to improved outcomes with ICIs and to a biologically distinct TME enriched in dysfunctional cytotoxic T-cells that may be reinvigorated by checkpoint inhibition. BMI may represent a pragmatic biomarker warranting prospective validation.