Conventional vs Traction Endoscopic Submucosal Dissection for Colorectal Tumors: A Meta-Analysis of Randomized Controlled Trials

Introduction: Endoscopic submucosal dissection (ESD) was developed for en bloc removal of pre-malignant and early malignant gastrointestinal (GI) lesions. Compared to the conventional ESD (C-ESD), the traction-assisted method (T-ESD) allows the lesion to be stabilized with more facile dissection. Several studies demonstrated the potential benefit of traction (T-ESD) compared to conventional (C-ESD) technique. We conducted a meta-analysis of all currently reported randomized controlled trials (RCTs) comparing C-ESD to T-ESD for colorectal tumors. Methods: Embase, Cochrane and MEDLINE databases were searched for RCTs evaluating C-ESD vs T-ESD in patients with colorectal tumors in April 2023. The endpoints of interest were procedure time (minutes), resection speed (mm2/min), R0 and en bloc resection rates, and incidence of adverse events. Standard meta-analysis methods were employed using a random-effects model. Results: Six RCTs with a total of 566 patients (284 C-ESD vs 282 T-ESD) were included in the final analysis. The mean age of the patients was 67 years and 60% were men. In the C-ESD and T-ESD groups respectively, the average type of lesions were 53.7% and 54.8% protruding or laterally spreading tumor granular type (LST-G), and 46.3% and 45.2% laterally spreading tumor nongranular type (LST-NG); the location of the lesions was 34.1% and 32.4% rectum, and 65.9% and 67.6% rest of the colon; and the average lesion size was 31.5mm and 30.1mm (Table 1). As compared with T-ESD technique, the C-ESD group was associated with longer procedure time (SMD 0.91, 95% CI [0.58 to 1.23, P < 0.00001, I2 = 65%]) and slower resection speed (SMD -1.03, [-2.01 to -0.06, P = 0.04, I2 = 93%]). No significant difference was found in the 2 groups with respect to R0 resection rate (RR 1.00, [0.94 to 1.06, P = 0.87, I2 = 42%]), en bloc resection (RR 0.99, [0.97 to 1.01, P = 0.35, I2 = 0%]), incidence of delayed bleeding (RR 0.66, [0.17 to 2.59, P = 0.55, I2 = 0%]) and perforation (RR 2.16, [0.75 to 6.27, P = 0.16, I2 = 0%]). Conclusion: Based on this meta-analysis of 6 RCTs, the clinical outcomes are similar with T-ESD and C-ESD strategies for colorectal tumors with no significant difference in R0 and en bloc resection. T-ESD was found to be significantly faster. These findings have significant practical implications, and given the multitude of traction tools, may warrant a cost analysis in the future. Table 1. - Clinical Outcome Results Outcome Results (Risk ratio/standardized mean difference, 95% confidence interval, P value) Procedure time SMD 0.91, 95% CI [0.58 to 1.23], P < 0.00001 Resection speed SMD -1.03, 95% CI [-2.01 to -0.06], P = 0.04 R0 resection rate RR 1.00, 95% CI [0.94 to 1.06], P = 0.87 En-bloc resection RR 0.99, 95% CI [0.97 to 1.01], P = 0.35 Delayed bleeding RR 0.66, 95% CI [0.17 to 2.59], P = 0.55 Perforation RR 2.16, 95% CI [0.75 to 6.27], P = 0.16