Late complications of postoperative radiation therapy for cancer of the rectum and rectosigmoid

Abstract

Purpose: We retrospectively examined the surgical, medical, radiotherapeutic and technical factors associated with late small bowel and nonsmall bowel morbidity.

Methods and materials: The medical records of 224 patients with cancer of the rectum and rectosigmoid treated mainly with abdominoperineal resection or anterior resection and postoperative radiotherapy at the University of Texas M.D. Anderson Cancer Center from 1973 to 1990 were reviewed. The median dose was 54 Gy (range 34-66 Gy) at 1.8-2 Gy per fraction using various techniques (23 had extended fields to L1 or L2; pelvic fields were treated with anterior-posterior in 85, 83 had a 3-field plan and 33 had a 4-field "box"). A positioning technique that treats patients on an open table-top device was used in 78 patients to move the small intestine out of the pelvis. Bladder distension was used in eight. Forty-seven patients received concomitant 5-fluorouracil. Small bowel series were performed in 122 patients to assess the volume of small bowel inside the pelvis below the conjugate line.

Results: In 29 patients, the median time to the development of small bowel obstruction was 7 months (range 0-69 months); 18 patients required reoperations. The small bowel obstruction rate was 30% in patients treated with daily extended field radiotherapy, 21% in those with a single pelvic field and 9% with multiple pelvic fields. Small bowel obstruction was positively correlated with postsurgical adhesions prior to radiotherapy and absence of reperitonealization at the time of initial surgery (p < 0.05). There was no correlation of small bowel obstruction with a history of hypertension, diabetes, prior surgery, history of abdominal infections, postoperative infections, wound healing, pathologic tumor stage, types of surgical procedures, sites of primary tumor, age, or sex. Patients developing small bowel obstruction had larger amounts of small bowel assessed radiologically below the conjugate line than those without complications. With the open table-top device, the small bowel obstruction rate was 3%. In 47 patients treated with radiation and chemotherapy on the open table-top device, the small bowel obstruction rate was 15%, but these patients had more small bowel inside the pelvis than those without the complication. The median time to the development of nonsmall bowel obstruction in 29 patients was 8 months (range 0-85 months), and the nonsmall bowel obstruction complications were significantly correlated with postoperative infection. Most nonsmall bowel obstruction complications were in the genitourinary tract and occurred in patients who had abdominoperineal resection.

Conclusion: The open table-top device, by moving the small bowel out of the treatment field, reduces small bowel obstruction in patients treated with radical surgery and postoperative radiotherapy for cancer of the rectum and rectosigmoid. This technique is facile, reproducible, and does not require patient compliance.