Comparison of late rectal toxicity from conventional versus three-dimensional conformal radiotherapy for prostate cancer: Analysis of clinical and dosimetric factors

Objectives. To compare late rectal toxicity (LRT) after definitive radiotherapy (DR) and salvage radiotherapy (SR) in prostate cancer using conventional (CONV) or three-dimensional conformal (3-D) techniques. Methods. The outcomes and clinical factors of 212 patients with Stage T1a-T4 prostate cancer were evaluated (separated into DR and SR groups). The median prescribed dose was 66, 74, 66, and 70 Gy, for the CONV-DR, 3-D-DR, CONV-SR, and 3-D-SR groups, respectively. LRT was scored using both Radiation Therapy Oncology Group (RTOG) and modified RTOG and Late Effects Normal Tissue (mRTOG/LENT) scales. Results. The 4-year biochemical relapse-free survival rate was 83 % for all patients, with a trend toward improvement in the 3-D groups (78 % CONV and 85 % 3-D, P = 0.12). One patient (1 %) in the CONV group and 24 (24 %) in the 3-D group experienced grade 2 or worse LRT by the mRTOG/LENT scale. Patients undergoing DR experienced grade 2 or worse LRT of 1 % versus 21 % (P = 0.003) for the CONV and 3-D groups, respectively. Patients undergoing SR experienced grade 2 or worse LRT of 0 % versus 40 % for the CONV and 3-D groups, respectively. The following variables correlated significantly with LRT on both univariate and multivariate analyses: prescribed radiation dose (P < 0.0001), percentage of rectal volume receiving 60 Gy (P < 0.005), and percentage of rectal volume receiving 70 Gy (P < 0.001). The pretreatment clinical factors, when added to the dosimetric data, were not statistically significant on multivariate analysis (P > 0.05). Conclusions. The prescribed radiation dose and percentage of rectal volume treated to 60 or 70 Gy had statistically significant correlations with increased LRT. The rate of grade 2 or worse LRT was greater for patients undergoing SR than for those undergoing DR. We believe that continued close attention to dosimetric variables is imperative for future studies of dose escalation. (C) 2005 Elsevier Inc.