High-resolution spatial multi-omics enables prognostic scoring in intrahepatic cholangiocarcinoma

Abstract

Intrahepatic cholangiocarcinoma (iCCA) represents a highly aggressive hepatic malignancy with escalating global incidence and dismal clinical outcomes. Despite therapeutic advancements through targeted therapies and immune checkpoint inhibitors, the field critically lacks robust prognostic tools for precise risk stratification and treatment optimization. Here, we present a comprehensive spatial multi-omics investigation integrating cutting-edge technologies: image mass cytometry for single-cell proteomic profiling, expansion gel-enabled spatial proteomics for regional resolution, complemented by single-cell RNA sequencing, bulk proteomics, and multiplex immunofluorescence. This integrated approach enabled high-dimensional mapping of the tumor microenvironment at spatial and molecular resolution. Leveraging these insights, we developed a novel spatial multi-modal scoring system through systematic analysis of a 155-patient training cohort at micro-scale resolution (1 mm²). To ensure clinical translatability, we performed rigorous cross-scale validation in an independent 214-patient cohort, addressing critical challenges in whole-slide pathologically intra- and intertumoral heterogeneity. Through multi-omics characterization encompassing single-cell, spatial, and proteomic dimensions, we elucidated the biological foundations underlying the prognostic power of our spatial multi-modal scoring system. This innovative framework offers insights into iCCA biology and holds potential as a practical tool for risk stratification and guiding therapeutic decisions in iCCA management.

Keywords: intrahepatic cholangiocarcinoma; multi-modal model; prognostic stratification; spatial proteomics; tumor microenvironment.