Purpose: Testosterone recovery (TR) after androgen deprivation therapy (ADT) and radiation therapy (RT) is not well characterized. We studied TR in men who received RT and either short-term ADT (STADT) or long-term ADT (LTADT) and aimed to create a nomogram to predict TR.
Methods and materials: We identified consecutive localized prostate cancer patients treated with ADT-RT at 2 academic medical centers from January 2011 to October 2016 with documented baseline testosterone. TR was time from last ADT injection to testosterone normalization. The Kaplan-Meier method was used to estimate time to TR. Cox proportional hazards models identified TR predictors. A nomogram was trained with site 1 and externally validated with site 2.
Results: A total of 340 patients were included; 69.7% received STADT for a median duration of 6 months; 30.3% received LTADT for a median duration of 24.3 months. Median follow-up was 26.7 months. Median time for TR was 17.2 months for STADT and 24.0 months for LTADT patients (P = .004). The 2-year cumulative incidence of TR was 53.1% after LTADT versus 65.7% after STADT (P = .004). On multivariate analysis, shorter duration ADT (hazard ratio [HR], 0.96; P = .004), higher pretreatment testosterone (HR, 1.004; P < .001), and lower body mass index (HR, 0.95; P = .002) were associated with shorter time to TR. Older age (HR, 0.97; P = .09) and white race (HR, 0.67; P = .06) trended as longer TR predictors. A nomogram was generated to predict probability of TR at 1, 2, and 3 years. The concordance index was 0.71 (95% confidence interval, 0.64-0.78) for the validation cohort.
Conclusions: In this population of localized prostate cancer patients, TR after ADT-RT was variable. Using baseline testosterone, ADT duration, body mass index, age, and race, a predictive nomogram can estimate the likelihood of TR.
Copyright © 2018 Elsevier Inc. All rights reserved.