Artificial Intelligence-Driven Quantification of Tumor-Stroma Ratio and Fibroblasts Enables Precise Classification of Stroma Quality and Quantity in Predicting Colorectal Cancer Recurrence.

Abstract

The tumor microenvironment (TME) significantly influences the progression and prognosis of colorectal cancer (CRC). Key components, including the tumor-stroma ratio (TSR) and cancer-associated fibroblasts (CAFs), have been recognized as important prognostic markers in CRC. However, the conventional assessment of TSR and CAF density is often subjective and labor-intensive, limiting its clinical application. In this study, we employed an artificial intelligence (AI)-powered whole slide image (WSI) analyzer, Lunit SCOPE IO, to objectively quantify TSR and CAF density in stage II and III CRC specimens from 207 treatment-naïve patients. Our analysis revealed that both TSR (log-rank p<0.0001) and CAF (log-rank p=0.017) density were independently associated with disease-free survival (DFS), providing superior prognostic accuracy compared to conventional risk factors. Notably, incorporating TSR and CAF density with traditional high-risk criteria allowed for the reclassification of additional patients as high-risk, significantly improving DFS prediction and reducing false-negative rates. These findings highlight the potential of integrating AI-based histopathological analysis into routine clinical practice to enhance diagnostic precision, improve risk stratification, and ultimately optimize patient management in CRC.