Dose/volume relationship of late rectal bleeding after external beam radiotherapy for localized prostate cancer: Absolute or relative rectal volume?

PURPOSE The purpose of this study was to analyze predictors of late rectal bleeding after external-beam radiotherapy for localized prostate cancer, with a focus on the volume of rectum irradiated. MATERIALS AND METHODS One hundred twenty-eight patients were treated with external-beam radiotherapy at the Cleveland Clinic Foundation between January 1998 and June 1999. Conformal radiotherapy (CRT) was used to deliver 78 Gy at 2 Gy per fraction in 76 cases, and shortcourse intensity-modulated radiotherapy (SCIM-RT) was used to deliver 70 Gy at 2.5 Gy per fraction in 52 cases. All contours were determined by one physician. The rectum was outlined from 1 cm above the target structures to 1 cm below the target structures. The entire volume of the rectum, along with the outer rectal wall, was included. All cases had detailed planning parameters that specifically determined the rectal volume receiving the prescription dose (VrPr), that is, 78 Gy for CRT and 70 Gy for SCIM-RT, and the percent of rectal volume receiving the prescription dose (%VrPr). The RTOG scales were used to evaluate late toxicity. The median follow-up was 24 months for all cases (range, 3-34 months), 21 months for SCIM-RT cases (range, 11-26 months), and 28 months for CRT cases (range, 3-34 months). RESULTS To date, five patients have had grade 1 late rectal toxicity (one CRT case and four SCIM-RT cases), one patient had grade 2 late rectal toxicity (CRT), and three patients had grade 3 late rectal toxicity (all CRT cases). Because of the low number of events, the analysis was performed with all patients experiencing rectal bleeding grouped together. The actuarial rectal bleeding rates at 18 and 24 months were 6% and 8%, respectively. The actuarial rectal bleeding rates at 24 months were identical (8%) for both SCIM-RT and CRT. A multivariate analysis of the following parameters was performed to determine independent predictors of rectal bleeding; age (continuous variable), race (Caucasian vs African American), coverage of seminal vesicles (yes vs no), adjuvant androgen deprivation (yes vs no), technique (CRT vs SCIM-RT), Radiation Therapy Oncology Group acute rectal toxicity score (continuous variable), VrPr (continuous variable in cubic centimeters), and %VrPr (continuous variable). Only the VrPr (cubic centimeter) was an independent predictor of rectal bleeding; %VrPr was not. With different cut-off levels being used, a VrPr of 15 cm(3) was significant on univariate analysis; the actuarial rectal bleeding rates at 24 months for patients with a VrPr :5 15 cm3 versus a VrPr > 15 cm(3) were 5% versus 22%, respectively. CONCLUSION In our study sample, which included both conformal and intensity-modulated radiotherapy patients, the volume of rectum receiving the prescribed radiation dose (the equivalent of 78 Gy) was an independent predictor of late rectal bleeding. The percent of rectal volume receiving the full dose was not. Using actual volume rather than percent volume also avoids the dependence on the extent of rectal volume contours. We recommend 15 cm(3) as the cut-off of the rectal volume not to exceed the prescription dose. The rectal bleeding rate at 2 years for cases with < 15 cm(3) receiving the full dose was only 5%.