In 418 men from an event-enriched radical prostatectomy (RP) cohort (1987–2004), this study evaluated whether adding Genomic Prostate Score (GPS), cribriform size, and Grade Group to quantitative unfavorable histology burden (UH%) improves 15-year metastasis prediction.
UH% alone yielded a c-index of ~0.85 for 15-year metastasis prediction; adding GPS and cribriform size raised this to 0.901. UH ≥10% carried an HR of 50.95 (95% CI 10.4–249), and all 102 metastatic events (24%) occurred exclusively in UH-positive tumors.
Retrospective, single-institution cohort with centralized pathology review limits generalizability; event-enriched design inflates event rate relative to typical clinical populations; long accrual window (1987–2004) may not reflect contemporary surgical or pathological practice.
Quantitative unfavorable histology burden (UH%) should anchor post-RP risk stratification — it outperforms Grade Group, stage, and genomic classifiers alone. GPS and cribriform size offer marginal refinement but are unlikely to change management once UH burden is known.
Explore related topics
