An exploratory study on the relationship between renal cell carcinoma and CAFs infiltration by integrating Pathomics and collagen features

Comprehensive Summary

This study examined the correlation between cancer-associated fibroblasts (CAFs) and disease progression in clear cell renal cell carcinoma (ccRCC). It used a non-invasive pathomics approach based on routine H&E-stained slides based on data from 354 TCGA ccRCC patients. The authors developed a machine-learning model that extracted 465 histopathological features to generate a Pathomics Score reflecting CAF-associated stromal remodeling. The model demonstrated good predictive performance for CAF infiltration (AUC = 0.813 in the training set and 0.782 in the validation set). Higher Pathomics Scores were linked to poorer survival of patients. This conveys that stromal features play an important role in disease outcomes. The score also reflected levels of tissue fibrosis, supporting its biological relevance. Overall, the study uses routine pathology slides, which makes it accessible and potentially easier to adopt in practice, but its reliance on retrospective TCGA data means further validation in real-world settings is required.

Outcomes and Implications

This paper highlights the clinical relevance of stromal biology in ccRCC, where five-year survival for advanced disease remains as low as 10-20%. The findings show that features related to cancer-associated fibroblasts on routine H&E slides are linked to patient survival, suggesting that the tumor stroma provides useful prognostic information. The ability to estimate CAF infiltration without additional staining or molecular testing helps overcome a common limitation in routine pathology; cost and scalability often restrict advanced analyses. Although the Pathomics Score is not yet validated for clinical use, its association with patient survival, tissue fibrosis, and common ccRCC mutations (VHL and PBRM1) suggests that it could be useful for risk assessment and future research on the tumor microenvironment. Validation in larger, multi-center studies will be needed before it can be applied in clinical practice.