Abstract 5839: Novel stroma-derived prostate cancer subtypes are associated with risk of clinical progression

Abstract

Abstract The tumor microenvironment influences prostate cancer aggressiveness, and by focusing on how different changes to the microenvironment can affect the cancer, we aimed to identify novel stromal subtypes in localized prostate cancer (LPC), characterize these subtypes, and utilize subtypes to increase accuracy in prostate cancer risk stratification. Subtype identification was conducted using non-negative matrix factorization with consensus clustering on a stroma-specific expression signature in RNA sequencing data. A discovery cohort (127 LPC patients) and two independent validation cohorts (406 LPC & 126 LPC patients) were used for subtype identification. All three cohorts were subsequently used for subtype characterization and prognostic evaluation. Identified subtypes were characterized by evaluation of clinical characteristics, by gene set enrichment analysis, and by cell-type deconvolution analysis. Survival differences between subtypes were assessed using Kaplan-Meier and univariate Cox regression analyses with biochemical recurrence (BCR) as endpoint. Lastly, a high-risk stromal subtype was investigated for its ability to improve risk stratification for intermediate risk prostate cancer using hazard ratio changes from univariate Cox regression analysis. We identified three stromal subtypes (S1-S3) in prostate cancer. Of these, subtype S3 showed characteristics of a highly reactive stromal environment with significantly (adj. p < 0.05) decreased cell polarity and stroma related functions, significantly (adj. p < 0.05) reduced fibroblast infiltration, and a significantly (adj. p < 0.05) increased immune cells infiltration compared to the other subtypes. Survival analysis also revealed significant (p < 0.05) association of S3 with BCR in all three cohorts (HR = 3.2, 2.1, and 3.4 in cohorts 1, 2 and 3 respectively). Lastly, S3 significantly (p < 0.05) improved the risk prediction for patients with intermediate recurrence risk according to the clinical nomogram CAPRA-S (cohort 1, delta HR = 1.4; cohort 2, delta HR = 2.3; cohort 3, delta HR = 4.2). In conclusion, by focusing on the stromal alteration in prostate cancer, this study sheds new light on the influence of tumor microenvironment on LPC aggressiveness. A stromal subtype characterized by dysregulated stromal functions and changes in stroma and immune cells infiltration predicted adverse outcome in patients with a high prognostic ambiguity. This utilization of stromal information to improve risk assessment for clinically intermediate risk PC indicates a potential use in future evaluation of PC aggressiveness. Citation Format: Martin Rasmussen, Jacob Fredsøe, Paul V. Salachan, Marcus P. Blanke, Benedicte P. Ulhøi, Michael Borre, Karina D. Sørensen. Novel stroma-derived prostate cancer subtypes are associated with risk of clinical progression. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5839.