Abstract
PURPOSE/OBJECTIVE(S)Testosterone (T) kinetics and its relationship with clinical outcomes has not been studied in trials using salvage radiotherapy and androgen deprivation therapy (ADT). We performed a secondary analysis of the NRG Oncology/RTOG 0534 SPPORT trial, which compared prostate bed radiotherapy (PBRT) (arm 1), PBRT + short-term androgen deprivation therapy (ADT) (arm 2), or PBRT + pelvic lymph node radiotherapy (PLNRT) + short-term ADT (arm 3). We assessed longitudinal serum T levels and the impact of testosterone recovery (TR) on clinical outcomes. MATERIALS/METHODSADT was given for 4-6 months in arms 2 and 3, starting 2 months prior to radiotherapy. The trial excluded patients with baseline T < 40% of the lower limit of normal. TR was defined in 3 ways: 1) return to non-castrate level (>50 ng/dL), 2) return to normal level (>300 ng/dL), and 3) return to baseline level. Time to TR was estimated using cumulative incidence and death without an event considered a competing risk. Unadjusted and adjusted hazard ratios and 95% confidence intervals (CIs) were calculated using Cox proportional hazards model. Freedom from progression (FFP) was defined as biochemical failure according to the Phoenix definition (PSA ≥2 ng/mL over the nadir PSA), clinical failure (local, regional, or distant), or death from any cause. RESULTSA total of 1699 patients with T at baseline and at least 1 follow-up assessment were included. The median age was 64 years (IQR 59 - 69), 12.8% were black, 14.9% had diabetes, and 54.1% were former or current smokers. Median baseline T in arms 1, 2 and 3 was 320 ng/dL (IQR 239 - 424), 319 ng/dL (IQR 237 - 438) and 330 ng/dL (IQR 252 - 446), respectively. At 6 months, median T in arms 1, 2 and 3 was 290 ng/dL (IQR 210 - 390), 190.4 ng/dL (IQR 66 - 296) and 191 ng/dL (IQR 40.5 - 313). At 2 years, in arms 2 and 3, TR to non-castrate, normal and baseline levels were 95%, 55% and 23%, respectively. At 5 years, in arms 2 and 3, TR to non-castrate, normal and baseline levels were 98%, 73% and 42%, respectively. FFP was superior in arms 2 and 3 vs. arm 1 in patients with TR by all three definitions. In patients with recovered T to normal levels by 2 years (n = 904), the 5-year FFP rates were 71.8% (95% CI 66.9-76.6) in arm 1, 77.2% (72.1-82.2) in arm 2, and 86.3% (82.3-90.3) in arm 3 (arm 2 vs arm 1: HR 0.74, 95% CI 0.56-0.98, p = 0.034; arm 3 vs arm 1: HR 0.54, 95% CI 0.40-0.72, p<.0001). CONCLUSIONThis work represents the largest study of T kinetics in patients treated with salvage radiation and ADT. Approximately half of patients did not normalize their T levels by 2 years. Our data validate an incremental and meaningful FFP benefit of adding short-term ADT and PLNRT to PBRT independent of T recovery.