Association of angiogenic, fibrotic, and immunosuppressive tumor microenvironment with immunotherapy outcomes in metabolic disease–associated biliary tract cancers.

Abstract

e16004 Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) is an increasing risk factor for biliary tract cancers (BTC), including cholangiocarcinoma (CCA). Although MASLD is characterized by chronic hepatic inflammation, its impact on the efficacy of immune checkpoint inhibitors (ICI) remains poorly defined. We evaluated clinical outcomes of ICI-treated MASLD-associated BTC and explored underlying genomic and transcriptomic features. Methods: We retrospectively analyzed BTC patients treated with durvalumab, cisplatin, and gemcitabine between June 2015 and December 2024 using an institutional database. MASLD was defined as metabolic dysfunction and hepatic steatosis visible on imaging or biopsy, based on American Association for the Study of Liver Diseases (AASLD) criteria. Hepatic steatosis was quantified on pre-treatment non-contrast CT scans using liver proton density fat fraction (PDFF), with PDFF ≥5% classified as steatosis. Overall survival (OS) following ICI therapy was compared between MASLD and non-MASLD groups using multivariable analyses. A molecularly characterized subset underwent whole-exome sequencing and bulk RNA sequencing (RNA-seq). Tumor microenvironment (TME) features were investigated using established transcriptomic classifiers (BostonGene) and cell deconvolution methods. Results: Among 151 BTC patients, 74 (49%) received ICI, and 53 (35.1%) met AASLD criteria for MASLD. Median OS for the overall cohort was 18.4 months. Non-MASLD patients had a longer median OS compared with MASLD patients (23.0 vs 16.7 months; p = 0.056). Transcriptomic profiling (n = 93) revealed that MASLD tumors exhibit a distinct TME relative to non-MASLD controls. Key features: (i) enhanced angiogenesis with upregulation of the angiogenic gene signature (p = 0.01), increased endothelial cell fraction (p = 0.04) (ii) an immunosuppressive infiltrate marked by regulatory T cells (p = 0.06), with increased B cells (p = 0.03) (iii) a fibrotic TME with upregulated matrix remodeling (p = 0.15) and cancer-associated fibroblast signatures (p = 0.16). Genomic analyses revealed no significant differences in tumor mutational burden or copy number variation between MASLD and non-MASLD, suggesting TME remodeling, rather than genomic alterations, as the main driver of the clinical differences observed. Conclusions: MASLD-associated BTC is characterized by a distinct TME marked by angiogenesis, fibrotic remodeling, and immunosuppressive cell infiltration, features that may contribute to diminished responses to ICI. These observations provide a rationale for comprehensive molecular profiling incorporating both DNA and RNA analysis, as well as combination approaches that pair ICIs with anti-angiogenic or anti-fibrotic agents to enhance therapeutic outcomes in this expanding patient population.