Late rectal toxicity: dose-volume effects of conformal radiotherapy for prostate cancer

Abstract

Purpose/Objective: To identify dosimetric, anatomic, and clinical factors that correlate with late rectal complications after three-dimensional conformal radiotherapy (3D-CRT) for prostate cancer. Materials/Methods: We retrospectively analyzed the dose volume histograms (DVH) and clinical records of 163 stage T1b-T3c prostate cancer patients who received definitive 3D-CRT without hormone therapy between 1992-1999. The patients were initially treated to 46 Gy using a conventional 4-field technique. A 6-field 3D-CRT approach was then used to boost the total isocenter dose to 74-78 Gy at 2 Gy per fraction using 18 MV photons. All late rectal complications were scored using modified RTOG and LENT criteria. Median follow-up was 62 (24-102) months. The DVHs generated from the pretreatment planning scan were analyzed to provide specific information on several dosimetric and anatomic variables. For the calculation of the DVH, the entire rectal volume was outlined. The 6-year toxicity rate was assessed using Kaplan-Meier analysis and log-rank tests. A univariate proportional hazard regression model was used to test the correlation between Grade 2 or higher toxicity and dosimetric, anatomic, and clinical factors. For those variables found to be significant, classification and regression tree (CART) analysis was used to identify cut-points that best discriminated those patients at high risk of late toxicity. Using a multivariate proportional hazard regression model, various clinical factors were added one at a time to each dosimetric variable to determine whether the hazard would be altered by the presence of the clinical factor. Results: The median time to developing Grade 2 or higher complications was 12 (6-72 months). The 6-year rates of Grade 2 and 3 complications were 21% and 6%, respectively. There were no Grade 4 complications. Univariate regression analysis showed that several dosimetric variables were highly significant with respect to developing Grade 2 or higher complications. The risk of rectal toxicity increased exponentially as a function of these dosimetric variables rather than linearly. These variables included maximum dose to CTV, maximum dose to rectum, and maximum dose to rectum as a percent of the prescribed dose (p<0.003 for all comparisons). The percent volume of rectum irradiated to 60 Gy, 70 Gy, 75.6 Gy, and 78 Gy was found to be highly significant (p<0.0001 for all comparisons). The absolute volume of rectum irradiated to the higher doses of 75.6 Gy and 78 Gy was also associated with late complications (p=0.0016 and 0.0021, respectively). Variables that were not significant were volume of CTV, volume of rectum, and maximum dose to CTV as a percent of the prescribed dose. For the dose-volume variables, CART analysis identified the optimal cut-points that best differentiated patients at high risk of late toxicity from those at low risk. The cut-points for the percent volume of rectum irradiated to 60 Gy, 70 Gy, 75.6 Gy, and 78 Gy were 40.6%, 26.2%, 15.8%, and 5.1%, respectively. The cut-points for the absolute volumes of rectum irradiated to 75.6 Gy and 78 Gy were 3.8 cc and 1.4 cc, respectively. The Kaplan-Meier 6-year rate of rectal complications was 54% for patients who had greater than 26.2% of rectum irradiated to 70 Gy versus 13% for those who had 26.2% or less irradiated to 70 Gy (p<0.0001). Of the clinical variables tested using univariate analysis, only hemorrhoids was found to increase the risk of Grade 2 or higher toxicity (p=0.003). Clinical variables that were not significant were diabetes, diverticulitis, inflammatory bowel disease, and abdominal surgery. In multivariate analysis each of the clinical variables was added one at a time to the dosimetric variables. Again, only the addition of hemorrhoids was found to significantly increase the risk of late toxicity (p<0.05 for all comparisons). Conclusions: DVH analyses clearly indicate a volume effect on the risk of developing late rectal complications. Therefore, dose escalation may be safely achieved by adherence to DVH constraints during treatment planning.