The safety of peroral endoscopic myotomy (POEM) is still debated since comprehensive analysis of adverse events (AEs) associated with the procedure in large multicenter cohort studies has not been performed. To study (1) the prevalence of AEs and (2) factors associated with occurrence of AEs in patients undergoing POEM. Patients who underwent POEM at 12 tertiary-care centers between 2009 and 2015 were included in this case–control study. Cases were defined by the occurrence of any AE related to the POEM procedure. Control patients were selected for each AE case by matching for age, gender, and disease classification (achalasia type I and II vs. type III/spastic esophageal disorders). A total of 1,826 patients underwent POEM. Overall, 156 AEs occurred in 137 patients (7.5%). A total of 51 (2.8%) inadvertent mucosotomies occurred. Mild, moderate, and severe AEs had a frequency of 116 (6.4%), 31 (1.7%), and 9 (0.5%), respectively. Multivariate analysis demonstrated that sigmoid-type esophagus (odds ratio (OR) 2.28, P=0.05), endoscopist experience <20 cases (OR 1.98, P=0.04), use of a triangular tip knife (OR 3.22, P=0.05), and use of an electrosurgical current different than spray coagulation (OR 3.09, P=0.02) were significantly associated with the occurrence of AEs. This large study comprehensively assessed the safety of POEM and highly suggests POEM as a relatively safe procedure when performed by experts at tertiary centers with an overall 7.5% prevalence of AEs. Severe AEs are rare. Sigmoid-type esophagus, endoscopist experience, type of knife, and current used can be considered as predictive factors of AE occurrence.

Background and study aims The variables associated with gastroesophageal reflux (GER) after peroral endoscopic myotomy (POEM) are largely unknown. This study aimed to: 1) identify the prevalence of reflux esophagitis and asymptomatic GER in patients who underwent POEM, and 2) evaluate patient and intraprocedural variables associated with post-POEM GER. Patients and methods All patients who underwent POEM and subsequent objective testing for GER (pH study with or without upper gastrointestinal [GI] endoscopy) at seven tertiary academic centers (one Asian, two US, four European) were included. Patients were divided into two groups: 1) DeMeester score ≥ 14.72 (cases) and 2) DeMeester score of < 14.72 (controls). Asymptomatic GER was defined as a patient with a DeMeester score ≥ 14.72 who was not consuming proton pump inhibitor (PPI). Results A total of 282 patients (female 48.2 %, Caucasian 84.8 %; mean body mass index 24.1 kg/m²) were included. Clinical success was achieved in 94.3 % of patients. GER evaluation was completed after a median follow-up of 12 months (interquartile range 10 – 24 months). A DeMeester score of ≥ 14.72 was seen in 57.8 % of patients. Multivariable analysis revealed female sex to be the only independent association (odds ratio 1.69, 95 % confidence interval 1.04 – 2.74) with post-POEM GER. No intraprocedural variables were associated with GER. Upper GI endoscopy was available in 233 patients, 54 (23.2 %) of whom were noted to have reflux esophagitis (majority Los Angeles Grade A or B). GER was asymptomatic in 60.1 %. Conclusion Post-POEM GER was seen in the majority of patients. No intraprocedural variables were identified to allow for potential alteration in procedural technique. ^* Co-first authors

Background and study aims Decisions concerning additional surgery after endoscopic resection of T1 colorectal cancer (CRC) are difficult because preoperative prediction of lymph node metastasis (LNM) is problematic. We investigated whether artificial intelligence can predict LNM presence, thus minimizing the need for additional surgery. Patients and methods Data on 690 consecutive patients with T1 CRCs that were surgically resected in 2001 – 2016 were retrospectively analyzed. We divided patients into two groups according to date: data from 590 patients were used for machine learning for the artificial intelligence model, and the remaining 100 patients were included for model validation. The artificial intelligence model analyzed 45 clinicopathological factors and then predicted positivity or negativity for LNM. Operative specimens were used as the gold standard for the presence of LNM. The artificial intelligence model was validated by calculating the sensitivity, specificity, and accuracy for predicting LNM, and comparing these data with those of the American, European, and Japanese guidelines. Results Sensitivity was 100 % (95 % confidence interval [CI] 72 % to 100 %) in all models. Specificity of the artificial intelligence model and the American, European, and Japanese guidelines was 66 % (95 %CI 56 % to 76 %), 44 % (95 %CI 34 % to 55 %), 0 % (95 %CI 0 % to 3 %), and 0 % (95 %CI 0 % to 3 %), respectively; and accuracy was 69 % (95 %CI 59 % to 78 %), 49 % (95 %CI 39 % to 59 %), 9 % (95 %CI 4 % to 16 %), and 9 % (95 %CI 4 % – 16 %), respectively. The rates of unnecessary additional surgery attributable to misdiagnosing LNM-negative patients as having LNM were: 77 % (95 %CI 62 % to 89 %) for the artificial intelligence model, and 85 % (95 %CI 73 % to 93 %; P < 0.001), 91 % (95 %CI 84 % to 96 %; P < 0.001), and 91 % (95 %CI 84 % to 96 %; P < 0.001) for the American, European, and Japanese guidelines, respectively. Conclusions Compared with current guidelines, artificial intelligence significantly reduced unnecessary additional surgery after endoscopic resection of T1 CRC without missing LNM positivity.

A retrospective analysis was conducted comparing open and percutaneous repair of ruptured Achilles tendon, over a 14-year period. One hundred and eight patients were identified; 70 underwent traditional open repair and 38 had a modified Ma & Griffith repair. The mean operating time with percutaneous repair was 28.5 minutes compared to 45.9 in the open group (p = < .0001). Comparable results were obtained in a subjective analysis of range of movement, stiffness, and power in both groups at time of discharge from clinic. In the open group there were four cases of rerupture (5.7%), four deep infections (4.7%), two palpable suture knots (2.9%), and one sural nerve lesion (1.4%). Complications with percutaneous repair included one rerupture (2.6%), five palpable suture knots (13.2%), four transient sural nerve lesions (10.5%), and no wound infections. There was no statistical significance between the two groups. The authors advocate percutaneous repair, in experienced hands, as a suitable alternative to traditional open techniques.