Association of gut microbiota with therapy efficacy and prognosis in biliary tract cancer.

Abstract

4111 Background: Biliary tract cancers (BTC) are rising in incidence and have poor prognosis. Microbial exposure via gut–liver axis may contribute to their development and progression. We profiled the gut microbiome of patients with BTC and assessed its correlation with chemoimmunotherapy outcome. Methods: For this prospective study, we collected baseline stool samples from newly diagnosed patients with BTC treated at a single center. Microbiome profiles were generated using 16S rRNA gene sequencing. Microbial alpha (intrasample variability) and beta (interindividual variability) diversity indices were calculated and correlated with treatment outcome. Taxonomical differential abundance was quantified using MaAsLin2. Statistical analyses included descriptive statistics for demographics, Kaplan-Meier for survival & Spearman's correlation for microbiome-clinical outcome. Results: From May 2021 to October 2024, 72 patients were enrolled. Median age was 66 years (range: 29–86) with 43% females. Primary sites included intrahepatic cholangiocarcinoma (CCA) (61%), distal extrahepatic CCA (18%), hilar CCA (11%), gallbladder Ca (3%), and mixed (7%). Stages were I (14%), II (33%), III (46%), IV (7%). 52% underwent surgery (14% upfront/others post neoadjuvant), 74% progressed to stage 4 disease and 67% received immune check point inhibitors. Therapy responses included complete (CR 3%), partial (PR 19%), stable (SD 32%), and progressive (PD 28%). At a median follow-up of 20 mos, mPFS was 14.6 (8.3-NR) mos for entire cohort [8 (5–11) mos for stage 4 patients] and mOS was not reached. Microbiome analysis of stool samples provided by immunotherapy-treated patients (n = 26) suggested that disease control (DCR: CR/PR/SD) correlated with higher alpha diversity (p = 0.063), while progression showed a trend toward lower diversity across all indices. Beta diversity (calculated by Bray-Curtis distances) showed no significant differences by response or progression. Differential abundance analysis using MaAsLin2 identified significant microbial associations (LogFC > 1.5, adjusted p < 0.25) with treatment outcomes. DCR associated bacterial genera included Ruminococcus , Subdoligranulum , Romboutsia & Collinsella , while PD correlated with higher abundances of Streptococcus , Eggerthella , Paraprevotella & Enterococcus . Genera such as Ruminococcus , Romboutsia , Coprococcus & Christensenellaceae R-7 group were linked to non-progression, whereas Parasutterella , Streptococcus , Prevotella-9 & Monoglobus were elevated in progressive disease. Conclusions: Our preliminary results suggest that gut microbiome should be studied as a multidimensional biomarker for predicting therapy response/prognosis in BTC. Higher microbial alpha diversity may link with better immunotherapy responses, and specific microbial taxa may correlate with treatment outcomes.