Abdominoperineal endoanal pull-through resection. A comparative study between immediate and delayed colorectal anastomosis.

Abdominoperineal endoanal pull-through resection with colorectal anastomosis was performed on 728 patients--primarily those with chagasic megacolon and cancer of the rectum. Intestinal continuity was reestablished through immediate anastomosis (Swenson procedure) in 229 patients and through delayed anastomosis (Cutait-Turnbull procedure) in 499. Comparative studies showed: that the incidence of leakage was 31.9 percent in immediate and only 2.2 percent in delayed anastomosis; that presacral infection occurred in 27.9 percent in immediate and in 6.8 percent in delayed anastomosis; that stenosis was observed in 4.4 percent in immediate and 1.8 percent in delayed anastomosis; that mortality was 6.1 percent in immediate and 2.2 percent in delayed anastomosis; that anal continence was good in both procedures and that sexual disturbances were rare in benign and frequent in malignant lesions in both procedures. The final conclusion is that, in abdominoperineal endoanal pull-through resection with colorectal anastomosis, complications and mortality are less frequent in delayed than in immediate anastomosis and that continence and sexual behavior are identical in both procedures.