# A Single-Cell Atlas of Pan-Cancer Liver Metastasis Reveals Dynamic Cellular Programs Driving Metastatic Progression and Immune Modulation

## TL;DR

This study creates a detailed map of liver metastases across various cancers, revealing how different cell types and interactions drive cancer spread and immune evasion.

## Contribution

The paper introduces a pan-cancer single-cell atlas of liver metastases and identifies four dynamic cellular programs linked to metastatic progression and immune modulation.

## Key findings

- Four cellular programs were identified that reflect transitions from immune-active to immunosuppressive tumor environments.
- Natural killer cells and macrophages are associated with early immune surveillance, while regulatory T cells dominate immunosuppressive states.
- The study highlights tumor-intrinsic adaptations and stromal remodeling as key drivers of metastatic niche formation.

## Abstract

Liver metastasis remains a major challenge in cancer treatment, yet its cellular and molecular landscape remains poorly defined at the pan-cancer level. Here, we construct a single-cell transcriptomic atlas of liver metastases across multiple cancer types by analyzing 100 single-cell RNA sequencing samples, profiling over 460,000 cells, and identifying 121 distinct cellular subtypes. We define 4 representative cellular programs (CPs) associated with liver metastasis, revealing how cellular composition and intercellular interactions within the tumor microenvironment drive metastatic progression and immune modulation. These CPs recapitulate a dynamic transition from immunoactive states, marked by natural-killer-cell-mediated immune surveillance and macrophage-driven angiogenesis, to immunosuppressive environments dominated by regulatory T cell infiltration and immune exclusion. The shift is marked by progressive alterations in immune infiltration, stromal remodeling, and tumor-intrinsic adaptations, elucidating key mechanisms of immune evasion and metastatic niche formation. Our study provides a high-resolution framework for understanding the heterogeneity and evolution of liver metastasis and highlights the potential of CP-based stratification to inform therapeutic strategies targeting the metastatic tumor microenvironment.