Abstract 4002: Peritumoral collagen microarchitecture is differently prognostic of biochemical recurrence across populations in prostate cancer
Rattachement africain : us. Niveau de preuve : code pays fourni par la source.
Le résumé fourni par la source
Abstract Background Extracellular matrix remodeling contributes to prostate cancer progression, yet race-associated differences in collagen architecture and their relationship to biochemical recurrence (BCR) remain poorly understood. We investigated whether collagen microstructure features differ between African American (AA) and European American(EA) men who had undergone radical prostatectomy, and whether population-specific collagen signatures carry prognostic value in the post-surgical setting. Methods We analyzed H&E-stained whole-slide images from radical prostatectomy specimens representing 101 AA and 101 EA men. Collagen fibers within the peri-tumoral region were computationally extracted using a deep-learning-based image-analysis pipeline, followed by quantification of 19 collagen morphology and texture features spanning structural (orientation, length, curvature, thickness, density, tortuosity, branching, intersection angles), spatial (inter-fiber distance), and textural (texture entropy and Haralick metrics) categories. The dataset was randomly divided into an 80% training set and a 20% independent test set, with equal representation of AA and EA patients in the test subset. Within the training set, population-specific prognostic features were identified using univariate median-threshold survival analysis (p<0.05). Multivariate Cox models were then trained separately using AA-derived and EA-derived feature subsets, and model performance was evaluated exclusively on the held-out test set to prevent data leakage. Results Nine collagen features were prognostic in AA patients, indicating more irregular collagen texture, greater fiber twisting, and higher variation in fiber orientation and curvature—suggesting a more disrupted stromal environment. In contrast, only five features were prognostic in EA patients, largely corresponding to differences in collagen density, fiber spacing, and overall stromal compactness. Only the collagen orientation variability feature was shared across both groups, underscoring the distinct tumor remodeling patterns in AA versus EA patients. The prognostic value was not retained when each population specific features was applied to the other population. In the full cohort, the AA-specific features demonstrated significant association with BCR (HR = 4.09, 95% CI: 1.09-15.34, p = 0.0243), whereas the EA-specific features did not (HR= 0.96, 95 % CI: 0.31-2.98, p = 0.9408). Conclusion Peri-tumoral collagen architecture differed markedly between AA and EA patients, influencing which features predicted BCR. Only AA-derived collagen patterns remained prognostic across groups, highlighting the need for population-aware prognostic approaches. Citation Format: Ajay Rajaraman Subramanian, Naoto Tokuyama, Tilak Pathak, Priti Lal, Anant Madabhushi. Peritumoral collagen microarchitecture is differently prognostic of biochemical recurrence across populations in prostate cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 4002.
Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.
Le contrôle bibliographique ouvert
DOI retrouvé dans Crossref DOI retrouvé ; titre concordant.
- Titre Crossref
- Abstract 4002: Peritumoral collagen microarchitecture is differently prognostic of biochemical recurrence across populations in prostate cancer
- Date Crossref
- 03/04/2026
- Éditeur
- American Association for Cancer Research (AACR)
- Type
- journal-article
Ce recoupement confirme des métadonnées liées au DOI. Il ne confirme ni la méthode ni les conclusions de l’étude, et il ne compte pas comme une seconde source scientifique indépendante.
Les institutions déclarées
Une affiliation ne permet pas de déduire la nationalité d’un auteur.