br Time from ADT therapy months br in LTADT patients
Time from ADT therapy (months)
38.2-69.6) in LTADT patients (P Z .004). Abbreviations: ADT Z androgen deprivation therapy; CI Z confidence interval; LTADT Z long-term ADT; STADT Z short-term ADT.
Testosterone recovery after ADT-RT 839
Table 4 Univariate and multivariate analyses
Characteristic HR 95% CI P HR 95% CI P
Duration of ADT
Black d d d d d d
d d d
Never d d d d d d
Abbreviations: ADT Z androgen deprivation therapy; BMI Z body mass index; CCI Z Charlson Comorbidity Index; CI Z confidence interval; HR Z hazard ratio; TR Z Sorafenib recovery.
* Indicates a significant result.
moderate-to-severe range, 27% had scores in the mild-to-moderate range, and the remainder had no self-reported erectile dysfunction. At the time of testosterone recovery, IIEF-5 scores were within the moderate-to-severe range for 74% of those within the training cohort and 77% of patients within the validation cohort. A total of 32% of patients within the training cohort and 60% of patients within the validation cohort had IIEF-5 data available at the time of testosterone recovery. At the most recent follow-up, 73% of patients within the training cohort had IIEF-5 data avail-able; 79% of those patients had IIEF-5 scores within the moderate-to-severe range, with median change in IIEF-5 score of e4 points (range, e24 to 17) from baseline IIEF-5 score. Within the validation cohort, 56% of patients had IIEF-5 data available at the most recent follow-up; 78% of those patients had IIEF-5 scores within the moderate to severe range with median change in score of e1 point (range, e23 to 5).
univariate analysis were then included into the multivariate analysis. Age was also included in the multivariate analysis because it was found to be a meaningful predictor of TR in previous studies.21
Predicting testosterone recovery
To our knowledge, this is the largest study reporting testosterone recovery after RT and ADT in patients with documented baseline testosterone levels. We found that
International Journal of Radiation Oncology Biology Physics
Baseline Testosterone (ng/dl)
1-year probability of recovery
2-year probability of recovery
Fig. 3. Shorter duration of ADT, higher pretreatment testosterone level, and lower BMI are associated with shorter time to TR. Older age and white race trended as longer TR predictors. Using these factors, a nomogram was generated to predict probability of TR at 1, 2, and 3 years. Abbreviations: ADT Z androgen deprivation therapy; BMI Z body mass index; TR Z testosterone recovery.
only 51.8% of patients had TR after ADT. When stratified by ADT duration, patients treated with shorter ADT were more likely to return to normal testosterone levels than patients treated with longer ADT. In addition, we identified factors associated with testosterone recovery in addition to duration of ADT, including pretreatment testosterone level, BMI, age, and race. Finally, we developed a nomogram that could be used as a tool by clinicians or patients to predict testosterone recovery based on individual patient charac-teristics. Our results quantified the likelihood and time course for TR beyond that done in prior studies.
Understanding testosterone recovery after STADT and RT is important because this standard therapy is often given with an implied understanding between clinicians and pa-tients that testosterone suppression will be temporary. Multiple smaller series with shorter follow-up suggest that with the use of STADT there is a reversible effect on serum testosterone. Shahidi et al22 studied the effect of 3 to 6 months of ADT in patients with prostate cancer and found that 88% of patients 228 (out of a total of 259 patients) had return to normal testosterone levels (290-860 ng/dL) at a median follow-up of 13.5 months. In the present study, with longer follow-up, we found a slightly lower rate of TR at 61.2% in men treated with STADT. These differences are likely a result of the characteristics of the individual pa-tients in our cohort.