17-Year Cause-Specific Survival Following Permanent Prostate Brachytherapy
To determine the pre-treatment and treatment related factors that influence 17-year cause-specific survival (CSS) in men undergoing permanent seed implantation (PSI) for localized prostate cancer.
Of 1776 men followed a minimum of 6 years (mean 10.3, range 6-22.4) 705 (39.7%) had PSI alone, 123 (6.9%) had PSI plus external beam irradiation (EBRT), 434 (24.4%) had PSI plus neoadjuvant hormone therapy (NHT) and 514 (28.9%) had and PSI plus NHT/EBRT. The treatment protocol was based on NCCN status. 948 (53.4%) received NHT for a median of 6 months (75%ile 3-9). NHT was given in men with NCCN1 (798, 44.9%) for 3 months for prostate volume > 50 cc, in NCCN2 (413, 23.3%) for 6 months and for 9 months in men with NCCN3 (565, 31.8%). 123 men received EBRT boost instead of NHT (NCCN2) and 434 of the NCCN3 received both NHT and EBRT. Radiation dose was converted to the biological equivalent dose (BED). CSS was determined by the absence of clinical recurrence at the time of death. Means were compared by ANOVA. Survival estimates were calculated by Kaplan-Meier method with comparison of variables by log rank and hazard rates (HR) computed by Cox regression.
In the 1776 men median age was 66 years (range 39-81) and median PSA was 6.8 ng/ml (range 0.3-300). 41 (2.3%) had died of prostate cancer at last follow-up. The 10- and 17-year CSS rates were 98.5% (866 remaining) and 94.3% (60 remaining), respectively. NCCN status (figure), PSA, Gleason group, clinical stage all influenced CSS (table). The use of NHT was not significant (p=0.091). Significant variables in the Cox regression were stage (p<0.001), Gleason score (p=0.001) and BED (p=0.018). Men not dying of prostate cancer had higher radiation doses (BED 194.9 vs. 171.8 Gy, p<0.001). </p>
A prostate cancer program that utilizes PSI alone or with EBRT boost (for higher risk patients) yields excellent 17-year CSS results. Men with higher Gleason score, clinical stage and lower radiation dose have inferior CSS.