Rectal sequelae after conformal radiotherapy of prostate cancer: dose-volume histograms as predictive factors.

PURPOSE: To identify clinically relevant parameters predictive of late rectal bleeding derived from cumulative dose-volume histograms (DVHs) of the rectum after conformal radiotherapy of prostate cancer. MATERIALS AND METHODS: One hundred and nine patients treated with 3D conformal radiotherapy between 1/1994 and 1/1996 for localized prostate cancer (clinical stage T1-T3) were available for analysis. All patients received a total dose of 66 Gy/2 Gy per fraction (specified at the International Commission on Radiation Units and Measurements ICRU reference point). DVHs of the contoured rectum were analyzed by defining the absolute (aV) and relative (rV) rectum volume that received more than 30% (V30), 50% (V50), 70% (V70), 80% (V80), 90% (V90) and 100% (V100) of the prescribed dose. Additionally, a new aspect of DVH analysis was investigated by calculation of the area under the DVH-curve between several dose levels (area under the curve (AUC)-DVH). DVH-variables were correlated with radiation side effects evaluated in 3-6 months intervals and graded according to the EORTC/RTOG score. The median follow-up was 30 months (12-60 months). RESULTS: Univariate and multivariate stepwise Cox-Regression analysis including age, PTV, rectum size, rV100, rV90, rV80, rV70, rV50 rV30 and aV30 to aV100 were calculated. Late rectal bleeding (EORTC/RTOG grade 2) was significantly correlated with the percentage of rectum volume receiving > or = 90% of the prescribed dose (rV90) (P = 0.007) and inversely correlated in a significant way with the size of contoured rectum (P = 0.006) in multivariate analysis. In our series, a proportion of the rectum volume > or = 57% were included in the 90%-isodose (rV90 > or = 57%) in one half of the patients, with an actuarial incidence of 31% of late rectal bleeding at 3 years. In the other half of the patients, when rV90 < 57%, the 3-year actuarial incidence was 11% (P < 0.03). CONCLUSION: Our data demonstrate a dose-volume relationship at the reference dose of 60 Gy ( approximately 90% of the prescribed dose) with respect to late rectal toxicity. The rV90 seems to be the most useful and easily obtained parameter when comparing treatment plans to evaluate the risk of rectal morbidity.