Predictive Factors for Late Genitourinary and Gastrointestinal Toxicity in Patients with Prostate Cancer Treated with Adjuvant or Salvage Radiotherapy

Abstract

Purpose/Objective: Radiotherapy (RT) is frequently used following radical prostatectomy (RP) in the adjuvant (ART) or salvage (SRT) setting. The use of ART has been somewhat limited due to concerns regarding cumulative toxicity. However, with mounting evidence suggesting that ART may lead to improved outcome for specific operative findings, use of ART may increase. More information is needed to better characterize toxicity. A large multi-institutional database with men who received ART or SRT after RP was analyzed for the magnitude of toxicity and predictors of late genitourinary (GU) and gastrointestinal (GI) toxicity. Materials/Methods: The characteristics of 959 men with a median follow up of 55 mo were analyzed. Pathologic stage was T1 in 6%, T2 in 30%, T3-4 in 64%. Gleason score was 2–6 in 29%, 7 in 49%, and 8–10 in 22%. 26% had seminal vesicle involvement and 60% had positive surgical margins. 15% received androgen deprivation (AD) for a median of 6 mo (range 1–80). 19% received ART, as defined by RT within 1 year of RP and PSA<0.2 ng/ml, with a median time from RP to RT of 4 mo (range 2–12). 81% received SRT, with a median time from RP to RT of 20 mo (range 2–155). 78% were treated to the prostate bed only (P), while 22% received radiation to the pelvis (WP). Dose ranged from 50–78 Gy, with a median of 64 Gy. GI and GU toxicity were graded using standard RTOG criteria. Stepwise Cox multivariate models for grade 2 or higher GU and GI toxicity were constructed with covariates including type of RT (SRT vs. ART), use of AD, RT dose and RT target (P vs. WP). Results: 12% of patients (13% P, 7%WP) had grade 2 or higher GU toxicity. 4% of patients (5% P, 2% WP) had grade 2 or higher GI toxicity. The distribution for all patients with grade 2 or higher GU toxicity was grade 2 in 86%, grade 3 in 12%, and grade 4 in 1%. For GI toxicity, the distribution was 84%, 9%, and 7%, respectively. In P patients, median dose was 64.8 Gy (range 54–78). In WP patients, median dose was 62 Gy (range 50–70.3). This difference was significant (p<0.0001) using the non-parametric Wilcoxon statistic. Both univariate and multivariate analysis demonstrated that ART (p=0.03), AD (p=0.0003), and prostate RT (p=0.004) significantly predicted for grade 2 or higher late GU toxicity, with hazard ratios of 1.58, 2.27, and 2.19, respectively. For GI toxicity, although ART and dose were significant in univariate analysis, there were no significant factors found in multivariate analysis. Conclusions: Overall, the number of high grade toxicities for postoperative RT is low. ART, AD, and P RT predicted for grade 2 or higher late GU toxicity. There were no predictors for GI toxicity found in this analysis. This suggests that radiotherapy before maximal recovery of urinary function, which can take up to 2 years, may be associated with more GU morbidity. The finding of increased GU toxicity with P RT was unexpected and suggests that other patient or treatment technique factors influencing toxicity have not yet been characterized.