Abdominosacral resection for primary irresectable and locally recurrent rectal cancer

PURPOSE: The purpose of this study was to present a technique of abdominosa cral resection and its results in patients with locally advanced primary or locally recurrent rectal cancer with dorsolateral fixation. METHODS: Betwe en 1994 and 1999, 13 patients with locally advanced primary rectal cancer a nd 37 patients with locally recurrent rectal cancer underwent abdominosacra l resection as part of a multimodality treatment, i.e., preoperative irradi ation, surgery, and Intraoperative irradiation. After the abdominal phase, the patient was turned from supine to prone position to perform the transsa cral phase of the resection. RESULTS: Margins were microscopically negative in 26 patients (52 percent), microscopically positive in 18 (36 percent), and positive with gross residual disease in 6 patients. Operation time rang ed from 210 to 590 (median, 390) minutes, and blood loss ranged from 400 to 10,000 (median, 3,500) ml. No operative or hospital deaths occurred. Posto perative complications occurred in 41 patients (82 percent); most notable w ere perineal wound infections or dehiscence (n = 24, 48 percent). Other com plications were postoperative urinary retention or incontinence (n = 9, 18 percent), peritonitis (n = 4), grade II neuropathy (n = 1), and fistula for mation (n = 3). Kaplan-Meier 3-year overall survival, disease-free survival , and local control rates were, respectively, 41 percent, 31 percent, and 6 1 percent. Completeness of the resection (negative vs. positive margins) wa s a significant factor influencing survival (P = 0.04), disease-free surviv al (P = 0.0006), and local control (P = 0.0002). CONCLUSION: The abdominosa cral resection provides wide access and may be the therapeutic solution for the accomplishment of a radical resection for distally situated, dorsally or dorsolaterally fixed primary or locally recurrent rectal cancers.