Abstract:Objective To evaluate the clinical value of endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) in early esophageal cancer and precancerous lesions of different sizes.Methods Retrospectively analyzed 579 cases of early esophageal cancer and precancerous lesions which received ESD and EMR. According to the size of lesions, the cases were divided into large-lesion group (n = 277) and small-lesion group (n = 302). Then compared the efficacy and complications of EMR and ESD in different sizes lesions.Results In small-lesion group: the complete resection rate of EMR was 93.2%, the curative resection rate of EMR was 89.0%, and the complete resection rate of ESD was 92.6%, the curative resection rate of ESD was 90.4%. Both of them had not statistically significant (P > 0.05). They were not statistically significant between EMR and ESD in bleeding and recurrence rates (P > 0.05). In large-lesion group: the complete resection rate of EMR was 84.2%, the curative resection rate of EMR was 84.2%, and the complete resection rate of ESD was 87.9%, the curative resection rate of ESD was 85.4%, which had not statistically significant (P > 0.05). The bleeding and recurrence rates in EMR were not statistically significant compared with ESD (P > 0.05). The results of multivariate Logistic regression analysis showed that the upper esophageal segment (OR^ = 5.098, 95%CI: 1.614 ~ 16.105, P = 0.006) was risk factors for endoscopic treatment of postoperative stenosis; longitudinal axis length < 5 cm (OR^ = 0.164, 95%CI: 0.064 ~ 0.420, P = 0.000) and circumferential ≤ 1/2 (OR^ = 0.302, 95%CI: 0.124 ~ 0.734, P = 0.008) were protective factors for endoscopic treatment of postoperative stenosis.Conclusion Both EMR and ESD are effective and safe treatments for early esophageal cancer and precancerous lesions. Lesions located in the upper esophagus, ≥ 5 cm in the longitudinal axis length, and circumferential > 1/2 are risk factors for postoperative stenosis for endoscopic treatment.