RESUMEN
PURPOSE: To evaluate patient characteristics and dosimetric parameters that predict biochemical failure (BCF) after real-time planned low-dose-rate prostate brachytherapy. METHODS: From 1998 to 2008, a low-risk cohort by National Comprehensive Cancer Network criteria of 341 men with a median followup of 41.6 months was analyzed. This cohort had a median age of 65.1 years, prostate volume of 35.8cc, and pretreatment prostate-specific antigen of 5.6ng/mL. Patients had predominately Gleason 6 (95.9%) and T1c (81.3%) disease. About 3.6% of the patients received androgen deprivation therapy. Kaplan-Meier and Cox proportional hazards survival analysis methods were used to analyze predictors of BCF (Phoenix definition). RESULTS: At 72 months, freedom from BCF was 91.1% (95% confidence interval=85.0-94.8). The median D(90) was 145.9Gy, and the median V(100) was 90.3%. Because of infrequent BCF, the following prostate volume groups were examined: 15-<25, 25-<35, 35-<45, and 45+cc. Of all possible predictors, only small prostate volume (15-<25cc group) was significantly associated with BCF (hazard ratio=8.44, 95% confidence interval=1.82-39.14, p=0.007). Using Kaplan-Meier analysis, time to BCF was also significantly increased in the lowest prostate volume 15-<25cc group with 24.1% failing at 48 months compared with 1.6-5.1% among the other groups. CONCLUSIONS: Real-time planned low-dose-rate prostate brachytherapy provides excellent biochemical control as a single-agent treatment for low-risk prostate cancer with 91.1% freedom from BCF at 72 months. Only prostate volume less than 25cc was an independent predictor of BCF.