Patterns and predictors of amelioration of genitourinary toxicity after high-dose intensity-modulated radiation therapy for localized prostate cancer: implications for defining postradiotherapy urinary toxicity.

Abstract

Treatment-related toxicity and quality of life (QoL) considerations are important when counseling patients with localized prostate cancer (PCa). To determine the incidence and longitudinal pattern of late genitourinary (GU) toxicity and QoL after high-dose, intensity-modulated radiotherapy (IMRT). A total of 268 patients with localized PCa were treated between June 2004 and December 2008 at a tertiary referral center. Median follow-up was 5 yr (range: 3-7.7 yr). Patients underwent IMRT to a total dose of 86.4Gy; 50% of patients underwent neoadjuvant and concurrent androgen-deprivation therapy. Patients were evaluated with the prospectively obtained International Prostate Symptom Score (IPSS) questionnaire. GU toxicity was also scored using the Common Terminology Criteria for Adverse Events (CTCAE) v.4.0; toxicity events were defined as increase over baseline. Differences in increases in IPSS sums and QoL index between baseline IPSS sum and QoL index groups were analyzed using the Kruskal-Wallis and Mann-Whitney tests. Univariate and multivariate Cox regression models were applied. The overall median IPSS sum increase during follow-up was 3 and was less pronounced among patients with severe baseline symptoms compared with those with mild baseline symptoms (median increase: 0 vs 4; p<0.0001). Overall QoL index was unchanged after IMRT but appeared to improve in patients with dissatisfied baseline QoL compared with satisfied baseline QoL (p<0.0001). Fifty-five (20%) and 2 (1%) patients developed grade 2 and 3 late GU toxicities, respectively; however, in 28 of 57 patients (49%), toxicity resolved during follow-up. Even though the IPSS data were prospectively obtained, most patients were not treated within a prospective protocol. Late GU toxicity after high-dose IMRT was mild; severe, late GU toxicity was rare. Changes in IPSS sum and QoL index were dependent on the baseline GU function, which might be useful for future patient counseling.