1005 High-dose (70–78 GY) conformal radiotherapy for prostate cancer; the relation between observed late bladder and rectum complications and parameters derived from the dose volume histograms

Abstract

Purpose: To determine the incidence of late gastrointestinal (GI) and genitourinary (GU) complications after conformal radiotherapy for prostate cancer, and to investigate the relation between these observed incidences and parameters derived from the Dose Volume Histograms (DVHs) of rectum and bladder wall. Patients and Methods: Hundred and thirty patients with T{sub 2-4}G{sub 1-3}N{sub 0}M{sub 0} prostate cancer were treated with conformal radiotherapy with the simultaneous boost technique in a dose-escalating protocol; 78 patients received a total dose of 70 Gy, 11 patients 74 - 76 Gy and 41 patients 78 Gy, each with a dose of 2 Gy per fraction. DVHs of the rectal wall were used to calculate NTCPs according to the model of Kutcher et al. with the estimated parameter values (n = 0.12, m = 0.15, TD{sub 50} = 80 Gy) according to Burman et al. The median follow-up was 17 months (range 6 - 72 months). The crude and actuarial incidence of late (> 6 months) GI and GU complications were determined using the RTOG/EORTC morbidity scoring system (Grade I to IV). Results: Neither for late GI nor for GU complaints, a grade IV complication was observed. GU complaints occurred in 90 patients (69%): 54 patients (42%) only experienced grade I toxicity, 26 patients (20%) had grade II toxicity, and 10 patients (8%) had grade III complications, of which 8 patients (6%) developed a urethral (7 pts) or ureteric stenosis (1 pt). The actuarial incidence of grade III GU complications was 10% at 2 years. Since bladder wall DVHs are unreliable and most grade III complications were not related to the bladder, the grade II and/or III complications were analyzed in terms of the total prescribed dose only, but no correlation could be demonstrated. GI complications occurred in 71 patients (55%): 59 patients (45%) developed a grade I complication, 11 a grade II complication and only 1 patient required laser treatment twice and blood transfusion because of rectal bleeding (grade III). The actuarial incidence of GI complications ({>=} grade II) was 13% at 2 years, both for doses to the anterior rectal wall { 75 Gy (Figure). This is much lower than after conventional radiotherapy for prostate cancer (Smit et al.), where these figures were 21% and 60%, respectively. In univariate analyses, we did not find a significant relation between GI complications rates (grade II and/or III) and irradiated rectal wall volumes, total prescribed radiation dose or maximum radiation dose to the anterior rectal wall. The estimated NTCPs (on average 11%) were much higher than the incidence of grade III GI toxicity in our study ( =} grade II. In patients with an estimated NTCP between 0% and 10%, the incidence of complications {>=} grade II was 9.5%; for NCTPs of 10%-20%, and > 20% these figures were 11.4% and 0%, respectively. We are currently investigating whether dose surface histograms are more reliable to predict the observed late GI complications {>=} grade II. Conclusions: Conformal radiotherapy reduces late rectal complications significantly with respect to conventional radiotherapy. Observed late GI complications grade III were much lower than the NTCP estimations based on the model and parameter values of Kutcher et al. and Burman et al., respectively. Observed late GI complications {>=} grade II were overall in the range of the NTCP estimations, but did not correlate specifically with DVH parameters and NTCP estimations. Consequently, adaptation of the NTCP model and/or its parameter values for rectum complications is highly necessary. Grade III GU complications were mostly related to the prostatic urethra and consequently DVHs of the bladder have only limited value.