Computational pathology identifies immune-mediated collagen disruption to predict clinical outcomes in gynecologic malignancies

This paper introduces CollaTIL- a pathology feature family that quantifies immune-collagen relationships within TME. This study extracts collaTIL features from gynecologic cancers (ovarian, cervical, and endometrial), and finds associations between immune infiltration and chaotic collagen architecture.

CollaTIL features were associated with OS, demonstrating it's value as a prognostic biomarker. CollaTIL low risk patients had high TIL density and greater disorder in collagen fiber orientations.

CollaTIL is a biomarker for drugs targeting immune-collagen interactions. CollaTIL may be used to identify subgroups of patients based on TME characteristics (TIL density, collagen disorganization, etc) allowing for improved patient selection. 

Read the full article HERE