Use of individual fraction size data from 3756 patients to directly determine the alpha/beta ratio of prostate cancer

PURPOSE: To examine the effect of fraction size and total dose of radiation on recurrence of localized prostate cancer. METHODS AND MATERIALS: A total of 3756 patients treated with radiation monotherapy at three institutions were analyzed, including 185 high-dose-rate brachytherapy (HDRB) boost patients. The 5th to 95th centiles of external beam radiotherapy (EBRT) fraction sizes and doses were 1.8 to 2.86 Gy, and 57.4 to 77.4 Gy, respectively, and HDRB fractional doses were between 5.5 and 12 Gy, totaling 147 unique fractionation schedules. Failure was defined by one biochemical (nadir + 2 ng/ml) and two advanced disease endpoints. The alpha/beta ratios were estimated via a proportional hazards model stratified by risk severity and institution. RESULTS: The alpha/beta ratio using biochemical recurrence was 3.7 Gy (95% confidence interval 95% CI], 1.1, infinity Gy) for EBRT-only cases and 2.6 Gy (95% CI, 0.9, 4.8 Gy) after the addition of HDRB data. This estimate was highly dependent on an HDRB homogeneity correction factor (120% HDRB dose increase; alpha/beta ratio 4.5 Gy, 95% CI 1.6, 8.7 Gy). A 5-Gy increase in total dose reduced the hazard of failure by 16% (95% CI 11, 21%, p < 0.0001), and had more impact as follow-up matured (p < 0.0003). The clinically advanced endpoints concurred with the biochemical failure results, albeit with less precision. CONCLUSIONS: This study supports the concept that the alpha/beta ratio of prostate cancer is low, although considerable uncertainty remains in the estimated value. Outcome data from EBRT studies using substantially higher doses per fraction are needed to show increased precision in these estimates.