Background and Aim

Intrahepatic cholangiocarcinoma (iCCA) is a highly aggressive and heterogeneous liver cancer with limited therapeutic options. Its pronounced intratumoral diversity at both molecular and cellular levels poses a significant barrier to effective treatment. To better understand the
complexity and spatial dynamics of the tumor microenvironment (TME), we employed spatial transcriptomics (ST) to resolve the architecture of clonal populations and their interactions with surrounding stromal components.

Methods

We combined Confetti-based lineage tracing in a genetically engineered mouse model of iCCA with 10x Genomics Visium CytAssist spatial transcriptomics across multiple tumors (n=4),
generating high-resolution spatial maps of tumor-stroma interactions. ST data were integrated with single-cell RNA sequencing (scRNA-seq) to identify cell types and reconstruct cell–cell
communication networks across clonal territories. Bioinformatic analyses included spatial
deconvolution (CARD), ligand–receptor inference (LIANA), copy number variation mapping
(inferCNV), dimensionality reduction (PaCMAP), and a custom random forest classifier for clonal identity. Findings were further validated in four human iCCA samples.

Results

Spatial transcriptomics revealed structured gradients of gene expression and niche-specific
transcriptional programs that aligned with histological features and clonal architecture.
Integration with scRNA-seq demonstrated that clonal identity not only drives intrinsic tumor cell programs but also actively shapes the surrounding microenvironment. Notably, cancer-
associated fibroblasts (CAFs) formed spatially restricted subpopulations that engaged in clone-specific signaling interactions, significantly influencing tumor cell proliferation and spatial
organization. This integrated spatial approach uncovered highly organized clonal ecosystems
coexisting within the same tumor mass.

Conclusion

Spatial transcriptomics was essential in decoding the spatial logic of iCCA. By mapping clonal-stromal interactions at high resolution, our study revealed how tumor clones shape—and are
shaped by—their microenvironment. These insights lay the groundwork for spatially targeted
and clonally informed therapeutic strategies in iCCA.