Does Timing of Androgen Deprivation Influence Radiation-Induced Toxicity? A Secondary Analysis of Radiation Therapy Oncology Group Protocol 9413

Abstract

We conducted a secondary analysis of the Radiation Therapy Oncology Group protocol 9413 to compare whether the timing of antiandrogen therapy (concomitant versus adjuvant) and testosterone levels influences the incidence of rectal and urinary toxicity in whole pelvic radiotherapy. We analyzed two of four study arms, in which all patients received radiotherapy to the whole pelvis followed by a boost to the prostate. The two arms differed solely in the timing of the total of 4 months of total androgen deprivation: arm I (320 patients, given concomitantly) and arm III (319 patients, given as adjuvant therapy). The influence of testosterone levels (measured at baseline and yearly thereafter) and its recovery on late rectal and urinary toxicity were modeled by multivariate logistic regression analysis and Fine and Gray's regression models. The occurrence of late rectal toxicity (grade 0-1 versus 2-5, P = 0.16) and late urinary toxicity (grade 0-1 versus 2-5, P = 0.52) was not significantly different statistically between the two arms. The time to testosterone recovery was significantly lower in the adjuvant arm (mean difference of 3.2 months, P = 0.0103). Age and radiation field size are statistically significant risk factors for late urinary toxicity (P = 0.01 and P = 0.02). Baseline testosterone levels, before beginning total androgen deprivation, were a statistically significant predictive factor for late urinary toxicity (P = 0.03). Older patients and patients with low testosterone levels at baseline are risk factors for late urinary and rectal toxicities, possibly through impaired tissue repair.