Abstract B042: STAR (Stroma-Tumor AI Risk) Assessment: Association of AI-derived tumor-stroma proportion with patient survival provides added prognostic value beyond KELIM in epithelial ovarian cancer

Abstract Background: There remains a critical need for biomarkers predicting treatment response in epithelial ovarian cancer (EOC). The KELIM score, derived from the rate of CA-125 elimination during the first 100 days of treatment, is a clinically available biomarker of response to platinum-based chemotherapy. While KELIM predicts treatment response, its utility is limited by the need for post-treatment data. Tumor–stroma proportion (TSP) has emerged as a prognostic biomarker across several malignancies. Studies from our group have shown that high TSP (≥50% stroma content assessed by pathologist evaluation) predicts platinum resistance and poor survival in EOC at diagnosis and, thus, before treatment. Methods: We compared the prognostic value of TSP and KELIM by analyzing manual pathologist assessments (TSPmanual) and artificial intelligence (AI)–derived evaluations (TSPauto) on digitized images from a curated cohort of EOC specimens. Results: In this cohort, we validated the prognostic significance of manual TSP, confirming prior findings. Furthermore, TSPauto and TSPmanual assessments were highly concordant (94% agreement, Cohen’s Kappa 0.89, p<0.001), providing a highly reproducible, automated, scalable approach. Unlike KELIM, which was associated with platinum resistance but not overall survival, high TSP auto was significantly associated with poor survival (HR 1.99, p=0.02). Conclusion: These findings support AI-derived tumor–stroma proportion as a pre-treatment prognostic biomarker for EOC that complements KELIM by offering additional clinically actionable survival information. Citation Format: Morgann Madill, Arpit Aggarwal, Mayukhmala Jana, Tilak Pathak, Boris Winterhoff, Timothy Starr, Katelyn Tessier, Britt Erickson, Andrew Nelson, Emil Lou, Anant Madabhushi, Martina Bazzaro. STAR (Stroma-Tumor AI Risk) Assessment: Association of AI-derived tumor-stroma proportion with patient survival provides added prognostic value beyond KELIM in epithelial ovarian cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Ovarian Cancer Research; 2025 Sep 19-21; Denver, CO. Philadelphia (PA): AACR; Cancer Res 2025;85(18_Suppl):Abstract nr B042.