Adverse Features and Competing Risk Mortality in Patients With High-Risk Prostate Cancer

PURPOSE: To assess survival and competing causes of mortality in prostate cancer (PCa) patients referred to radical prostatectomy through a combination of unfavorable characteristics. PATIENTS AND METHODS: We evaluated 615 PCa patients referred to radical prostatectomy and pelvic lymph node dissection at single tertiary-care center with at least one adverse feature (AF): preoperative prostate-specific antigen ≥ 20 ng/mL, pathologic Gleason score 8 to 10, and no organ-confined disease at final pathology (seminal vesicle involvement, positive surgical margins, and/or lymph node invasion). Kaplan-Meier analyses were used to assess cancer-specific mortality (CSM)-free survival rates by stratifying patients into 3 risk categories according to the number of AFs (namely, 1, 2, and 3 AFs). Multivariable competing risk Cox regression analyses were used to assess CSM and other cause of mortality. RESULTS: Significant differences were found in terms of preoperative and pathologic tumor characteristics, adjuvant therapies, and biochemical recurrence (BCR). Men with 1 AF had higher CSM-free survival estimates compared to those with 2 and 3 AFs (92.8% vs. 84.2% vs. 27.7% at 10 years' follow-up, P < .001). Moreover, the presence of 3 AFs (hazard ratio [HR], 2.96), postoperative adjuvant treatment status (HR, 2.44), and time to BCR (HR, 0.96) were all independent predictors of CSM (P ≤ .04). Age at surgery and time to BCR were the only independent predictors of other causes of mortality (P ≤ .0009). CONCLUSION: The risk group stratification according to the number of AFs could help physicians to accurately predict oncologic outcomes and to select PCa patients for the most appropriate postoperative strategies.