Histologic predictors of clinically significant biochemical recurrence and PSA kinetics after radical prostatectomy.

Abstract

398 Background: Radical prostatectomy (RP) is a key treatment for localized prostate cancer. However, 20–40% of patients experience biochemical recurrence (BCR)—defined as a PSA rise >0.2 ng/mL after surgery—indicating disease progression. Unfavorable histologic features, such as cribriform and intraductal carcinoma, have been shown to predict adverse oncologic outcomes and long-term disease progression following RP. We investigated whether these features can provide insight into the risk of BCR, PSA kinetics, and the risk of metastasis or cancer-related death after BCR. Methods: We retrospectively analyzed a post-RP group from one institution with a median follow-up of 13.5 years (7.9–17.8). Pathology specimens were re-reviewed by a blinded GU pathologist and classified as containing favorable (FH) or unfavorable (UH) histology. Outcomes included BCR, PSA doubling time (PSADT), and development of metastasis or cancer-specific death following BCR. Group differences were assessed using Wilcoxon rank-sum tests and Chi-squared or Fisher’s exact. Multivariable Cox regression or Logistic regression evaluated associations between histology type, BCR risk, and PSADT. Results: A total of 844 subjects met the inclusion criteria. BCR occurred in 33.8% (285/844), and 61.8% (123/199) had a PSADT ≤1 year. Patients with UH had higher rates of BCR (52% vs. 10%, p < 0.001), earlier BCR (26 vs. 60 months, p = 0.005), and more frequently had a PSADT ≤1 year (65% vs. 40%, p = 0.016). Among UH patients with BCR, metastatic progression (39% vs. 0%, p < 0.001) and cancer-specific death (20% vs. 0%, p < 0.001) were more common. On multivariable analysis, UH (HR 3.47, p < 0.001), initial PSA (HR 1.02, p <0.001), positive margins (HR 1.48, p = 0.002), and grade group (HR 1.56, p < 0.001) were independently associated with BCR, while age and pathologic T stage were not. UH (OR 2.77, p = 0.044) and positive margins (OR 1.86, p = 0.047) were also associated with PSADT ≤1 year, while age, initial PSA, and pathologic T stage were not. Conclusions: UH features were associated with higher rates of clinically-significant BCR resulting in adverse cancer outcomes, underscoring their value as predictors of prostate cancer progression. Notably, those with UH experienced earlier BCR and had a faster PSADT, which raises possible differences in post-RP follow-up and management. Clinical characteristics by histology type. Characteristic Overall n = 844 Favorable Histology n = 361 Unfavorable Histology n = 483 p-value Age at surgery (yrs) 61 (57, 66) 60 (56, 64) 63 (58, 67) <0.001 Initial PSA before surgery (ng/mL) 6.0 (4.5, 9.2) 5.2 (4.2, 6.7) 6.9 (4.9, 11.0) <0.001 Margin Status <0.001 Positive 301 (37%) 96 (28%) 205 (44%) Pathological T Stage <0.001 T2 434 (51%) 267 (74%) 167 (35%) T3 410 (49%) 94 (26%) 316 (65%) Grade Group <0.001 0-1 15 (1.8%) 15 (4.2%) 0 (0%) 2 545 (65%) 337 (93%) 208 (43%) 3 162 (19%) 9 (2.5%) 153 (32%) 4-5 122 (14%) 0 (0%) 122 (25%)