Unique role of proximal rectal dose in late rectal complications for patients with cervical cancer undergoing high-dose-rate intracavitary brachytherapy

Purpose: To investigate the correlation of the radiation dose to the upper rectum, proximal to the International Commission of Radiation Units and Measurements (ICRU) rectal point, with late rectal complications in patients treated with external beam radiotherapy (EBRT) and high-dose-rate (HDR) intracavitary brachytherapy (ICRT) for carcinoma of the uterine cervix. Methods and Materials: Between June 1997 and February 2001, 75 patients with cervical carcinoma completed definitive or preoperative RT and were retrospectively reviewed. Of the 75 patients, 62 with complete dosimetric data and a minimal follow-up of at least 1 year were included in this analysis. Of the 62 patients, 36 (58%) also received concurrent chemotherapy, mainly with cisplatin during EBRT. EBRT consisted of a mean of 50.1 +/- 1.3 Gy of 18-MV photons to the pelvis. A parametrial boost was given to 55 patients. Central shielding was used after 40-45 Gy of pelvic RT. HDR ICRT followed EBRT, with a median dose of 5 Gy/fraction given twice weekly for a median of four fractions. The mean dose to point A from HDR ICRT was 23.9 +/- 3.0 Gy. In addition to the placement of a rectal tube with a lead wire during ICRT, 30-40 mL of contrast medium was instilled into the rectum to demonstrate the anterior rectal wall up to the rectosigmoid junction. Late rectal complications were recorded according to the Radiation Therapy Oncology Group grading system. The maximal rectal dose taken along the rectum from the anal verge to the rectosigmoid junction and the ICRU rectal dose were calculated. Statistical tests were used for the correlation of Grade 2 or greater rectal complications with patient-related variables and dosimetric factors. Correlations among the point A dose, ICRU rectal dose, and maximal proximal rectal dose were analyzed. Results: Fourteen patients (23%) developed Grade 2 or greater rectal complications. Patient-related factors, definitive or preoperative RT, and the use of concurrent chemotherapy were not associated with the occurrence of rectal complications. The maximal rectal dose during ICRT was at the proximal rectum rather than at the ICRU rectal point in 55 (89%) of 62 patients. Patients with Grade 2 or greater rectal complications had received a significantly greater total maximal proximal rectal dose from ICRT (25.6 Gy vs. 19.2 Gy, p = 0.019) and had a greater maximal proximal rectal dose/point A dose ratio (1.025 vs. 0.813, p = 0.024). In contrast, patients with and without rectal complications had a similar dose at point A (25.0 Gy vs.23.6 Gy, p = 0.107). The differences in the ICRU rectal dose (17.8 Gy vs. 15.4 Gy,p = 0.065) and the ICRU rectal dose/point A dose ratio (0.71 vs. 0.66, p = 0.210) did not reach statistical significance. Patients with >62 Gy of a direct dose sum from EBRT and ICRT to the proximal rectum (12 of 29 vs. 2 of 33, p = 0.001) and >110 Gy of a total maximal proximal rectal biologic effective dose (13 of 40 vs. 1 of 22, p = 0.012) presented with a significantly increased frequency of Grade 2 or greater rectal complications. The correlations between the maximal proximal rectal dose and the ICRU rectal dose were less satisfactory (Pearson coefficient 0.375). Moreover, 11 of the 14 patients with rectal complications had colonoscopic findings of radiation colitis at the proximal rectum, the area with the maximal rectal dose. Conclusion: Eighty-nine percent of our patients had a maximal rectal dose from ICRT at the proximal rectum instead of the ICRU rectal point. The difference between patients with and without late rectal complications was more prominent for the proximal rectal dose than for the ICRU rectal dose. It is important and useful to contrast the whole rectal wall up to the rectosigmoid junction and to calculate the dose at the proximal rectum for patients undergoing HDR ICRT. (C) 2003 Elsevier Inc.