Postoperative Pneumomediastinum Detected By Computed Tomography After Endoscopic Submucosal Dissection for Esophageal Neoplasms in Consecutive Cases

Abstract

Background: Endoscopic submucosal dissection (ESD) has the advantage over conventional endoscopic mucosal resection. ESD are now being increasingly used for the treatment of gastric cancers. However, esophageal wall is very thin and there is no serous membrane in esophageal wall. If there is small injury of muscularis propria, easily pneumomediasinum is occurred. The procedure needs a higher quality of skill for early esophageal cancer. Aim: We evaluated the clinical outcome of ESD for early esophageal cancer and esophageal tumor, and the frequency of post operative pneumomediasinum detected by CT. Patients and Methods: From January 2003 to September 2008, 42 consecutive patients underwent endoscopic submucosal dissection for early esophageal cancer and esophageal tumor, and their clinical outcomes were evaluated in our institution. In these patients, from March 2005 to September 2008, 30 consecutive patients evaluated post operative pneumomediasinum by CT. The using device for ESD were Flex Knife alone (25 cases), Flex and IT Knife (9 cases), Hook knife and IT knife (9 cases). And ESD procedures performed by using sodium hyaluronate. Indications were cquamous cell carcinoma (40cases), adenocarcinoma (3cases), granular cell tumor (1case). Setting: Gifu university hospital department of gastroenterology, Gifu, Japan. Main Outcome Measurements: The en bloc resection rate, complications, total operation time, recurrence rate and frequency of postoperative pneumomediasinum detected by CT were also evaluated. And leukocyte counts, C-reactive protein value within 3 days postopration and hospitalization days were assessed between the postoperative pneumomediasinum groups and the non- pneumomediasinum groups. Results: The en bloc resection rate of ESD was 93% (40/43). The curative resection rate of ESD was 81.4% (35/43). Mean operation time ware 75.7minutes. No post operative bleeding and no perforation occurred during ESD. A case of recurrence was observed, and was treated by surgery. 8cases had postoperative pneumomediasinum (26.7%): groupA. 22cases had no pneumomediasinum (73.3%): groupB. Mean leukocyte count (9120: groupA vs. 7731: groupB), mean C-reactive protein value (3.1 mg/dl: groupA vs. 2.2 mg/dl: groupB) and hospitalization were not significantly different between the two groups. Conclusion: ESD can be safely performed for early esophageal cancer and esophageal tumor, resulting in a high en bloc resection rate. But ESD for esophageal lesion has a relatively high rate of postoperative pneumomediasinum, so should only be performed by experienced endoscopists.