Endoscopic submucosal dissection (ESD) has been utilized as an alternative treatment to endoscopic mucosal resection for superficial esophageal cancer. We aimed to evaluate the complications associated with esophageal ESD and elucidate predictive factors for post-ESD stenosis. The study enrolled a total of 42 lesions of superficial esophageal cancer in 33 consecutive patients who underwent ESD in our department. We retrospectively reviewed ESD-associated complications and comparatively analyzed regional and technical factors between cases with and without post-ESD stenosis. The regional factors included location, endoscopic appearance, longitudinal and circumferential tumor sizes, depth of invasion, and lymphatic and vessel invasion. The technical factors included longitudinal and circumferential sizes of mucosal defects, muscle disclosure and cleavage, perforation, and en bloc resection. Esophageal stenosis was defined when a standard endoscope (9.8 mm in diameter) failed to pass through the stenosis. The results showed no cases of delayed bleeding, three cases of insidious perforation (7.1%), two cases of endoscopically confirmed perforation followed by mediastinitis (4.8%), and seven cases of esophageal stenosis (16.7%). Monovalent analysis indicated that the longitudinal and circumferential sizes of the tumor and mucosal defect were significant predictive factors for post-ESD stenosis (P< 0.005). Receiver operating characteristic analysis showed the highest sensitivity and specificity for a circumferential mucosal defect size of more than 71% (100 and 97.1%, respectively), followed by a circumferential tumor size of more than 59% (85.7 and 97.1%, respectively). It is of note that the success rate of en bloc resection was 95.2%, and balloon dilatation was effective for clinical symptoms in all seven patients with post-ESD stenosis. In conclusion, the most frequent complication with ESD was esophageal stenosis, for which the sizes of the tumor and mucosal defect were significant predictive factors. Although ESD enables large en bloc resection of esophageal cancer, practically, in cases with a lesion more than half of the circumference, great care must be taken because of the high risk of post-ESD stenosis.

OBJECTIVES: Intraductal papillary mucinous neoplasms (IPMNs) are precursor lesions of pancreatic adenocarcinoma. Artificial intelligence (AI) is a mathematical concept whose implementation automates learning and recognizing data patterns. The aim of this study was to investigate whether AI via deep learning algorithms using endoscopic ultrasonography (EUS) images of IPMNs could predict malignancy. METHODS: This retrospective study involved the analysis of patients who underwent EUS before pancreatectomy and had pathologically confirmed IPMNs in a single cancer center. In total, 3,970 still images were collected and fed as input into the deep learning algorithm. AI value and AI malignant probability were calculated. RESULTS: The mean AI value of malignant IPMNs was significantly greater than benign IPMNs (0.808 vs 0.104, P < 0.001). The area under the receiver operating characteristic curve for the ability to diagnose malignancies of IPMNs via AI malignant probability was 0.98 ( P < 0.001). The sensitivity, specificity, and accuracy of AI malignant probability were 95.7%, 92.6%, and 94.0%, respectively; its accuracy was higher than human diagnosis (56.0%) and the mural nodule (68.0%). Multivariate logistic regression analysis showed AI malignant probability to be the only independent factor for IPMN-associated malignancy (odds ratio: 295.16, 95% confidence interval: 14.13–6,165.75, P < 0.001). DISCUSSION: AI via deep learning algorithm may be a more accurate and objective method to diagnose malignancies of IPMNs in comparison to human diagnosis and conventional EUS features.

Background and Aims Although the technique of endoscopic ultrasound‐guided choledochoduodenostomy (EUS‐CDS) is becoming standardized, its safety issues have not been sufficiently investigated. Therefore, we aimed to identify factors associated with adverse events and stent patency in EUS‐CDS. Methods Consecutive patients who underwent EUS‐CDS between September 2003 and July 2017 were included. Technical/clinical success, adverse events and stent dysfunctions were analyzed retrospectively. Results A total of 151 patients underwent EUS‐CDS. In nine patients, procedures were discontinued before puncture. Technical and clinical success rates were 96.5% (137/142) and 98.5% (135/137), respectively. The adverse event rate was 20.4% (29/142). As a risk factor for peritonitis, plastic stents (PS) showed a significantly high odds ratio (OR) compared with covered self‐expandable metal stents (CSEMS; OR, 4.31; P = 0.030). CSEMS cases showed a significantly longer patency period than PS cases (329 vs 89 days; HR, 0.35; P < 0.001). As a risk factor for early stent dysfunction (within 14 days), stent direction to the oral side showed a significantly high OR (OR, 43.47; P < 0.001). In cases with oblique‐viewing EUS, double penetration of the duodenum occurred at significantly higher frequency than in cases with forward‐viewing EUS (7.0 vs 0.0%; P = 0.024). Conclusions Plastic stents and stent direction to the oral side were risk factors for peritonitis and early stent dysfunction, respectively. Using covered self‐expandable metal stents and changing stent direction to the anal side seemed appropriate to prevent peritonitis and early stent dysfunction.

Summary of Event: Pneumoderma, mediastinal emphysema, and bilateral pneumothorax were developed in the patient who had undergone transesophageal endoscopic ultrasonography-guided rendezvous technique. Chest drainage was performed immediately. Teaching Point: Transesophageal approach carries the potential risks of severe complications such as mediastinal emphysema, mediastinitis, and pneumothorax. To prevent puncturing through the esophagus, clipping the esophagogastric junction using a forward-viewing scope before procedure is very useful. In cases of inadvertent transesophageal puncture, devices other than the needle should not be passed through the site.

It is important to distinguish between "primary" OBC in ALNs and "metastatic" OBC from micro-primary breast tumor. Further studies are required to determine if omission of mastectomy and WBRT is acceptable.