Higher than standard radiation doses (>= 72 Gy) with or without androgen deprivation in the treatment of localized prostate cancer

Purpose: To study the effect on biochemical relapse-free survival (bRFS) an d clinical disease-free survival of radiation doses delivered to the prosta te and periprostatic tissues for localized prostate cancer. Methods and Materials: A total of 1041 consecutive localized prostate cance r cases treated with external beam radiotherapy (RT) at our institution bet ween 7/86 and 2/99 were reviewed. All cases had available pretreatment para meters including pretreatment prostate-specific antigen (iPSA), biopsy Glea son score (bGS), and clinical T stage. The median age was 69 years. Twenty- three percent of cases (n = 238) were African-American. The distribution by clinical T stage was as follows: T1 in 365 cases (35%), T2 in 562 cases (5 4%), and T3 in 114 cases (11%). The median iPSA level was 10.1 ng/ml (range : 0.4-692.9). The distribution by biopsy Gleason score (bGS) was as follows : less than or equal to 6 in 580 cases (56%) and greater than or equal to 7 in 461 cases (44%). Androgen deprivation (AD) in the adjuvant or neoadjuva nt setting was given in 303 cases (29%). The mean RT dose was 71.9 Gy (rang e: 57.6-78.0 Gy). The median RT dose was 70.2 GS, with 458 cases (44%) rece iving at least 72.0 Gy. The average dose in patients receiving <72 Gy was 6 8.3 Gy (median 68.4) versus 765 Gy (median 78.0) for patients receiving gre ater than or equal to 72 Gy. The mean follow-up was 38 months (median 33 mo nths). The number of follow-up prostate-specific antigen (PSA) levels avail able was 5998. Results: The 5- and 8-gear bRFS rates were 61% (95% CI55-65%) and 58% (95% CI51-650/0), respectively. The 5-year bRFS rates for patients receiving rad iation doses greater than or equal to 72 Gy versus <72 Gy were 87% (95% CI 82-92%) and 55% (95 % CI 49-60%), respectively. The 8-year bRFS rates for p atients receiving radiation doses greater than or equal to 72 Gy versus <72 Gy were 87% (95% CI 82-92%) and 51% (95% CI 44-58%), respectively (p < 0.0 01). A multivariate analysis of factors affecting bRFS was performed using the following parameters: age (continuous variable), race, T-stage (T1-T2 v s. T3), iPSA (continuous variable), bGS (less than or equal to 6 vs. greate r than or equal to 7), use of AD (yes vs. no), radiation technique (conform al versus standard), and radiation dose (continuous variable). T-stage (p < 0.001), iPSA (p < 0.001),bGS (p < 0.001), and RT dose (p < 0.001) were ind ependent predictors of outcome. Age (p = 0.74), race (p = = 0.96), radiatio n technique (p = 0.15), and use of AD (p = 0.31) were not. We observed 11% clinical failures (local, distant, or both) at 5 years and 15% at 8 years f or the entire cohort. There was a statistically significant improvement wit h higher radiation doses (p = 0.032). The 5-year clinical relapse rates for patients receiving greater than or equal to 72 Gy versus <72 Gy were 5% an d 12%, respectively. The 8-year clinical relapse rates for patients receivi ng radiation doses greater than or equal to 72 Gy versus <72 Gy were 5% and 17%, respectively (p = 0.026). Conclusion: Patients receiving radiation doses exceeding 72 Gy had signific antly better bRFS and clinical disease-free survival rates. Although result s need to be confirmed with longer follow-up, these preliminary results are extremely encouraging. If these results are confirmed by other institution s and by longer follow-up, RT doses exceeding 72 Gy should be considered as standard of care. (C) 2000 Elsevier Science Inc.