Experimental low colorectal anastomosis.

The authors evaluated a side-to-end colorectal anastomosis carried out by the abdominoperineal approach, in which opening of the mucosal layer of the colon, and therefore the lumen, is delayed for 4 days. This technique was compared to transabdominal end-to-end anastomosis performed as low as possible, using the circular stapler and hand-sewing with a one-layer technique. Experimental end-to-end stapled and hand-sewn anastomoses were performed randomly in 24 female mongrel dogs. The operating time was similar for each group. Anastomoses were examined for leaks at 7 to 10 days digitally and by roentgenography after barium enema. At sacrifice 14 days after anastomosis, measurements were made of bursting strength and the distance of anastomoses from the anus. The degree of inflammation was evaluated histologically. There were three leaks in the hand-sewn anastomoses and none in the experimental end-to-end stapled anastomoses. The mean bursting strength was similar for all three techniques. The mean distance from the anus was lowest for the experimental anastomoses. Inflammatory changes were least in the end-to-end stapled anastomoses. Two dogs in this group and one in the experimental group were incontinent postoperatively. The experimental anastomosis compares favourably to end-to-end stapled and hand-sewn anastomoses and may have some advantages for low anastomosis in man.