Transanal endoscopic microsurgery: a systematic review.

PURPOSE: The aim of this study was to systematically review the evidence relating to the safety and efficacy of transanal endoscopic microsurgery, a relatively new technique used to locally excise rectal tumors, compared with existing techniques such as anterior resections and abdominoperineal resections or local excisions.
METHODS: We conducted a systematic review of comparative studies and case series of transanal endoscopic microsurgery from 1980 to August 2002.
RESULTS: Three comparative studies (including one randomized, controlled trial) and 55 case series were included. The first area of study was the safety and efficacy of adenomas. In the randomized, controlled trial, no difference could be detected in the rate of early complications between transanal endoscopic microsurgery (10.3 percent) and direct local excision (17 percent) (relative risk, 0.61; 95 percent confidence interval, 0.29-1.29). Transanal endoscopic microsurgery resulted in less local recurrence (6/98; 6 percent) than direct local excision (20/90; 22 percent) (relative risk, 0.28; 95 percent confidence interval, 0.12-0.66). The 6 percent rate of local recurrence for transanal endoscopic microsurgery in this trial is consistent with the rates found in case series of transanal endoscopic microsurgery (median, 5 percent). The second area of study was the safety and efficacy of carcinomas. In the randomized, controlled trial, no difference could be detected in the rate of complications between transanal endoscopic microsurgery and direct local excision (relative risk for overall early complication rates, 0.56; 95 percent confidence interval, 0.22-1.42). No differences in survival or local recurrence rate between transanal endoscopic microsurgery and anterior resection could be detected in either the randomized, controlled trial (hazard ratio,1.02 for survival) or the nonrandomized, comparative study. There were 2 of 25 (8 percent) transanal endoscopic microsurgery recurrences in the randomized, controlled trial, but no figures were given for recurrence after anterior resection. In the case series, the median local recurrence rate for transanal endoscopic microsurgery was 8.4 percent, ranging from 0 percent to 50 percent. The third comparison was cost of the procedures. Transanal endoscopic microsurgery had both a lower recurrence rate and a lower cost than local excision or anterior resection for adenomas. Although the effectiveness of transanal endoscopic microsurgery could not be established for carcinomas, costs were lower than those for either anterior resection or abdominoperineal resection.
CONCLUSIONS: The evidence regarding transanal endoscopic microsurgery is very limited, being largely based on a single relatively small randomized, controlled trial. However, transanal endoscopic microsurgery does appear to result in fewer recurrences than those with direct local excision in adenomas and thus may be a useful procedure for several small niches of patient types--e.g., for large benign lesions of the middle to upper third of the rectum, for T1 low-risk rectal cancers, and for palliative, not curative, use in more advanced tumors.