Endoscopic submucosal dissection for treatment of rectal carcinoid tumors

Abstract

Background: Conventional EMR has been the endoscopic treatment of choice for rectal carcinoid tumors. However, histologically complete resection often cannot be achieved because the carcinoid tumors are located mainly in the submucosal layer. Endoscopic submucosal dissection (ESD), a new method for large colorectal neoplasm resection, may overcome this problem.

Objective: To compare ESD with EMR for the endoscopic treatment of rectal carcinoid tumors.

Design: A prospective case series with comparison to retrospective controls.

Setting: Tertiary-care center.

Patients: From January 2007 to January 2009 we prospectively enrolled consecutive patients with rectal carcinoid tumors less than 16 mm in diameter and with no regional lymph node enlargement shown by CT or EUS. For comparison, we retrospectively randomly selected patients who had undergone EMR for treatment of rectal carcinoid tumors between March 2000 and December 2006.

Intervention: We performed ESD of rectal carcinoid tumors in prospectively enrolled patients.

Main outcome measurements: Rate of en bloc resection, rate of histologically complete resection, incidence of complications, and length of procedures.

Results: The ESD group contained 31 patients (18 male, 13 female; age range 34-65 years), and the EMR group contained 62 patients (42 male, 20 female; age range 22-77 years). Both groups had similar mean rectal carcinoid tumor diameters (ESD 6.8 +/- 2.4 mm, EMR 7.3 +/- 2.2 mm; P = .106). Resection time was longer in the ESD group than in the EMR group (11.4 +/- 3.7 minutes vs 4.2 +/- 3.2 minutes, P < .001). The en bloc resection rate was 100% (31 of 31) in the ESD group and 95.2% (59 of 62) in the EMR group (P = .213). The histologically complete resection rate was 90.3% (28 of 31) in the ESD group and 71.0% (44 of 62) in the EMR group (P = .035). Suspected perforation occurred in 1 ESD patient (3.2%) and in 1 EMR patient (1.6%), and both patients were successfully managed by conservative measures. Immediate bleeding occurred in 1 ESD patient (3.2%) and in 4 EMR patients (6.5%); all instances of bleeding were controlled endoscopically.

Limitations: Retrospective control study and limited experience at a single center.

Conclusion: Compared with EMR, ESD resulted in a higher histologically complete resection rate, had a similar complication rate, and took slightly longer to perform. Given the advantages of complete resection, these findings indicate that ESD may be considered for treatment of rectal carcinoid tumors.