Impact of biochemical failure on overall survival after radiation therapy for localized prostate cancer in the PSA era

Purpose: To study the impact of biochemical failure on overall survival rates during the first 10 years after definitive radiotherapy for localized prostate cancer. Methods and Materials: The analysis was performed on 936 cases treated at a single institution between 1986 and 1998 with definitive radiotherapy. The median age of treatment was 69 years (range: 46 - 86 years). Pretreatment PSA levels (iPSA) and biopsy Gleason scores (bGS) were available for all cases. The clinical stage was Tl/T2A in 63%, T2B/C in 27%, and T3 in 10%. The median iPSA level was 9.6 ng/mL (range: 0.4-692.9 ng/ml.). The iPSA was less than or equal to10 in 53% and >10 in 47%. The bGS was less than or equal to6 in 59% and greater than or equal to7 in 41%. Androgen deprivation (AD) was administered in 181 cases (19%) for a median duration of 6 months (range: 1-6 months). All 181 cases received AD neoadjuvantly, i.e., before and/or during the radiotherapy. No AD was delivered after the completion of radiation. The median radiation dose was 70 Gy (range: 60-78 Gy). The radiotherapy technique was conformal in 376 (40%) cases. The American Society of Therapeutic Radiology and Oncology definition of biochemical failure (bF) was used; 316 cases (34%) had failed biochemically, and 620 (66%) had not. The end point was overall survival (OS). Time to death was determined from the time of definitive radiotherapy. The median PSA follow-up was 58 months. The median follow-up times for bF vs. no-bF cases were 77 and 49 months, respectively. A multivariate analysis of factors affecting OS using the proportional hazards model was performed for all cases using the following variables: age (>65 vs. less than or equal to65 years), race (African-American vs. Caucasian), clinical T stage (T1-2A vs. T2B-C vs. T3), bGS (less than or equal to6 vs. 7 vs. greater than or equal to8), iPSA (continuous variable), use of AD (yes vs. no), year of therapy (continuous variable), radiation dose (continuous variable), radiation technique (conformal vs. standard), and biochemical failure (yes vs. no). Results: The 5-year OS rate for the entire group was 89% (95% CI [confidence interval]: 86-91 %). The 5-year OS rates for bF vs. no-bF patients were 89% (95% Cl: 86-93%) and 89% (95% Cl: 86-92%), respectively. The 10-year OS rate for the entire group was 68% (95% CI: 61-75%). The 10-year OS rates for bF vs. no-bF patients were 65% (95% CI: 56-74%) and 77% (95% CI: 69-84%), respectively. The difference between bF and no bF was not significant in predicting overall survival in univariate analysis (log-rank test, p = 0.68). On multivariate analysis, bGS (p < 0.001), T stage (p = 0.003), radiation dose (p = 0.017), year of therapy (p = 0.031), and age (p = 0.020) were independent predictors of death. iPSA levels (p = 0.33), race (p = 0.80), radiation technique (p = 0.16), and use of AD (p = 0.09) were not predictive of OS. Biochemical failure (p = 0.052) showed only a trend for independently predicting overall survival on multivariate analysis. Conclusion: Biochemical failure after definitive radiotherapy for localized prostate cancer is not associated with increased mortality within the first 10 years after initial therapy, although a trend toward worse outcome was observed at 10 years. Longer follow-up from initial therapy is needed to fully understand the impact of biochemical failure on overall survival. With longer follow-up, significant differences might be observed at 15 or 20 years after therapy. (C) 2002 Elsevier Science Inc.