CRLM Prognosis Reimagined: From Static Scores to Dynamic Precision Oncology ESMO Open | April 2026

Introduction

Colorectal cancer liver metastases represent a major determinant of survival in colorectal cancer, affecting more than half of patients. While hepatic resection remains the cornerstone of curative therapy, recurrence rates remain high, reflecting the biological heterogeneity of disease. Traditionally, clinical risk scores such as the Fong score have guided decision-making, but these models are largely based on static anatomical and clinical parameters, limiting their relevance in the era of precision oncology.

Problem Statement

Conventional clinical risk scores fail to capture tumor biology and dynamic disease behavior, limiting accurate prognostication and personalized treatment decisions in CRLM.

Summary

This review highlights a paradigm shift in CRLM risk stratification—from static clinical models to biologically informed and dynamic frameworks. Classical scores, although useful, are insufficient to predict long-term outcomes or treatment response. Emerging tools now incorporate molecular profiling (e.g., RAS, BRAF, MSI status), histological growth patterns, and liquid biopsy techniques such as circulating tumor DNA (ctDNA).

Molecular markers refine prognosis and guide therapy, with KRAS and BRAF mutations indicating worse outcomes, while MSI tumors may respond better to immunotherapy. Histological growth patterns further stratify risk, with desmoplastic patterns associated with better prognosis and replacement patterns linked to aggressive disease.

Most importantly, dynamic biomarkers such as ctDNA are transforming management. Preoperative ctDNA predicts recurrence risk, while postoperative ctDNA identifies minimal residual disease and can guide adjuvant therapy decisions with high accuracy.

Overall, the integration of clinical, molecular, and dynamic data enables adaptive, real-time risk stratification. This evolving approach supports better patient selection, optimized timing of surgery and systemic therapy, and represents a major step toward true precision oncology in CRLM.