Association between obesity, molecular subtype, and immunotherapy outcomes in intrahepatic cholangiocarcinoma: The obesity paradox.

Abstract

587 Background: Obesity and elevated body mass index (BMI) are associated with increased cancer risk yet paradoxically correlate with improved outcomes after immune checkpoint inhibition in lung cancer and melanoma. Their impact on treatment outcomes in intrahepatic cholangiocarcinoma (iCCA) remains unclear. We investigated the associations between BMI, genomic alterations, tumor immune microenvironment, and outcomes with chemoimmunotherapy (chemo-IO) in patients (pts) with iCCA. Methods: This retrospective study included pts with iCCA treated at a single center. Demographics, clinicopathologic characteristics, genomic profiles, and outcomes were extracted from electronic health records. Pts were stratified as obese (BMI ≥30) or non-obese (<30). Tumor microenvironment profiling included RNA-sequencing and CODEX multiplex tissue imaging. Multivariate logistic regression was used to assess associations between obesity and genetic alterations. Kaplan-Meier and log-rank tests evaluated overall survival (OS). Results: We identified 661 pts with iCCA of all stages (I–4%, II–11%, III–17%, IV–59%, unknown–9%) treated between Aug 2015 and Nov 2024. Median age was 62 (range 24–92); 51% were female. Comorbidities included diabetes (15%), hypertension (48%), and hyperlipidemia (32%). Obesity was present in 28%. Surgical resection was performed in 453 pts, and 289 received RT. Median OS (mOS) was similar between obese and non-obese pts (29.6 vs 28.5 mo, p=0.73). Among chemo-IO-treated pts (n=240), obese pts had numerically higher (not significant) mOS (40.2 vs 29.4 mo, p=0.55). Landmark analysis of long-term survivors (≥18 mo; n=96) demonstrated a superior mOS in obese pts after chemo-IO (39.7 vs 18.8 mo, p=0.02). Genomic profiling (n=653) revealed that obese pts had a lower prevalence of FGFR2 fusions (OR=0.57, p=0.039), KRAS (OR=0.41, p=0.023), and IDH1 mutations (OR=0.64, p=0.051), but higher prevalence of IDH2 (OR=2.4, p=0.054) and NRAS mutations (OR=1.93, p=0.09). RNA-sequencing analysis (n=86; obese=37, non-obese=49) showed increased immune infiltration of CD8 + T-cells (p=0.047) and resting mast cells (p=0.044) in obese pts. CODEX spatial profiling (n=38; obese=17, non-obese=21) demonstrated increased abundance of CD141 + dendritic cells (p=0.023) with concurrent reduction in CD163 + M2 macrophages (p=0.023). Conclusions: Obesity in iCCA is associated with distinct genomic and immune activation features which may contribute to improved outcomes in long-term survivors after chemo-IO. Its potential role as a prognostic biomarker with IO for iCCA should be further evaluated in larger, independent cohorts.