Prediction of survival outcomes following postoperative radiotherapy after radical prostatectomy for prostate cancer.

To evaluate predictive factors for survival outcomes after post-prostatectomy radiotherapy. In the years 2003-2008, 324 patients have received postoperative radiotherapy a median time of 14 months after radical prostatectomy. All patients have been treated up to 66.0-66.6 Gy in 1.8-2.0 Gy fractions. Predictive factors were analyzed at two stages, using a multivariable Cox regression analysis: (1) based on factors known before radiotherapy and (2) based on prostate-specific antigen response after radiotherapy. Median follow-up after radiotherapy was 121 months. Prostate-specific antigen before radiotherapy, pN1 and Gleason score remained predictive factors for disease-free (hazard ratio, HR of 6.0, 2.3 and 2.5) and overall survival (HR of 2.8, 2.0 and 1.6) in multivariable analysis. Prostate-specific antigen levels increased despite radiotherapy in 27% of patients in the first six months. Failed response following salvage radiotherapy and prostate-specific antigen doubling time at the time of biochemical recurrence were predictive factors for disease-free (HR of 2.8 and 7.3;  < .01) and overall survival (HR of 2.2 and 2.6;  < .01). To reach the best survival outcomes following prostatectomy, salvage radiotherapy should be initiated early with low prostate-specific antigen levels, especially in patients with higher Gleason scores. Patients not responding to radiotherapy and/or patients with a short prostate-specific antigen doubling time after radiotherapy are candidates for early additional treatments.