Combined Risk Factors and Outcomes in High-Risk Prostate Cancer: Secondary Analysis of a Phase III Trial

Abstract

<h3>Purpose/Objective(s)</h3> In patients with high-risk prostate cancer (HRPC) treated with androgen deprivation therapy (ADT) and pelvic radiotherapy (RT), we used known prognostic risk factors in a secondary analyses of a phase 3 trial, aiming to identify favorable/unfavorable strata. Biochemical failure (BF), distant metastases (DM), prostate cancer specific mortality (PCSM), overall survival (OS) and distant metastases-free survival (DMFS) rates were compared according to high-risk factors (HRF). <h3>Materials/Methods</h3> 630 patients were randomized (PCS4 trial) to ADT of 36 vs. 18 months (ms) plus RT. We examined the impact of risk factor (RF) on outcomes reviewing 553 patients who completed the prescribed dose of ADT (206 patients in the 36ms duration and 347 in the 18ms). Analyses were computed by total ADT received and not by intention to treat. Outcomes were based on known prognostic RF: T stage (T1c-T2 vs. T3-T4), PSA level (≤20 vs. > 20) and Gleason score (<8 vs. ≥8). Because of the limited number of patients with 3 RF, patients with only one RF were compared to patients with 2 or 3 RFs. BF, DM and PCSM were analyzed with competing risks methods, while OS and DMFS with Kaplan Meier method and the log rank test. <h3>Results</h3> Of the 553 patients, 416 (75%) had only one RF, as follows: 226 (54%) Gleason ≥ 8, 140 (34%) PSA ≥ 20, 50 (12%) T stage T3-T4. There was no significant difference between patients receiving 36ms or 18ms in the frequency of one RF (77 vs. 72%, p=0.2). A total of 120 patients had 2 RFs and 17 had all 3 RFs. Overall, 172 (31.1%) patients developed BF, 68 (12.3%) DM, 65 (11.8%) PCSM. In patients with 2-3 RFs vs. 1 RF, the cumulative incidence was significantly higher for BF (sub distribution Hazard Ratio (sHR) = 1.65 (1.19-2.29), p=0.002), DM (sHR= 1.79 (1.09-2.95), p=0.02) and PCSM (sHR= 2.21 (1.35-3.64), p=0.002). Overall, 222/553 patients have survived, 215 of them without metastasis. The 10-year DMFS rate was lower (56% vs. 66%) and significantly worse [Hazard ratio = 1.28 (95% CI, 1.00-1.63), p=0.04] for patients with 2-3 RFs compared to those with one RF. The 10-year OS rate was also lower for 2-3 RFs (59% vs. 68%), though not statistically significant (HR=1.26 (0.99-1.61), p=0.064). Multivariable models for BF, DM, PCSM, OS and DMFS were computed including RFs (2-3 vs. 1 RF), age, Zubrod performance score and ADT duration (18 vs. 36ms). In all models, except for OS, number of RFs and age were significant. Except for BF, longer ADT duration did not significantly improve any of the outcomes. <h3>Conclusion</h3> HRPC patients harboring 2-3 RFs presented worst outcomes compared with those with one RF. Therapeutic strategies of optimized treatments should be considered in these patients.