Randomized comparative long-term survival study of endoscopic and thoracoscopic esophageal wall repair after NOTES mediastinoscopy in healthy and compromised animals

BACKGROUND AND STUDY AIMS Natural orifice transluminal endoscopic surgery (NOTES) has not yet been widely adopted because of lack of suitable equipment and fear of possible serious complications, especially in the mediastinum. We compared endoscopic with thoracoscopic esophageal wall repair after full-thickness esophageal wall incision (FTEI) and NOTES mediastinoscopy in healthy versus compromised animals. METHODS After FTEI for mediastinoscopy, 24 pigs (12 healthy, 12 compromised) were randomly allocated to endoscopic or thoracoscopic repair (each arm of each group, n = 6). They were kept alive for 3 months after endoscopic closure with prototype T-anchor suturing or thoracoscopic repair. RESULTS FTEI and mediastinoscopy were uneventful in all as was the initial repair of the incision (mean repair times: thoracoscopic 65 +/- 3.2 minutes, endoscopic 52 +/- 5.1 minutes; P < 0.0005). Post procedure, all 12 healthy pigs thrived with no complications or deaths. Two compromised animals died during the preparation period, and had to be replaced. In the compromised group, during endoscopic repair, 2 / 6 pigs suffered from gastric reflux into esophagus and mediastinum; the repair was completed and the pigs kept alive; one subsequently died of mediastinitis, and in the other, autopsy showed a gastric abscess in the lower mediastinum. Regarding the compromised thoracoscopic subgroup, one animal died from mediastinitis and all had abscesses at or near the incision sites. CONCLUSION Transesophageal mediastinoscopy could be performed equally well as the transthoracic procedure, both in healthy and compromised animals. However, on follow-up, the compromised animals had worse outcomes, with more complications and two deaths (17 %), one in each arm.