Abu Dayyeh, Barham K

Background and Aims: Early identification of malignant biliary strictures (MBSs) is challenging, with up to 20% classified as indeterminants after preliminary testing and tissue sampling with endoscopic retrograde cholangiopancreatography. We aimed to evaluate the use of methylated DNA markers (MDMs) from biliary brushings to enhance MBS detection in a prospective cohort. Approach: Candidate MDMs were evaluated for their utility in MBS diagnosis through a series of discovery and validation phases. DNA was extracted from biliary brushing samples, quantified, bisulfite-converted, and then subjected to methylation-specific droplet digital polymerase chain reaction. Patients were considered to have no malignancy if the sampling was negative and there was no evidence of malignancy after 1 year or definitive negative surgical histopathology. Results: Fourteen candidate MDMs were evaluated in the discovery phase, with top-performing and new markers evaluated in the technical validation phase. The top 4 MDMs were TWIST1, HOXA1, VSTM2B, and CLEC11A, which individually achieved AUC values of 0.82, 0.81, 0.83, and 0.78, respectively, with sensitivities of 59.4%, 53.1%, 62.5%, and 50.0%, respectively, at high specificities for malignancy of 95.2%–95.3% for the final biologic validation phase. When combined as a panel, the AUC was 0.86, achieving 73.4% sensitivity and 92.9% specificity, which outperformed cytology and fluorescence in situ hybridization (FISH). Conclusions: The selected MDMs demonstrated improved performance characteristics for the detection of MBS compared to cytology and FISH. Therefore, MDMs should be considered viable candidates for inclusion in diagnostic testing algorithms.

Objective Endoscopic sleeve gastroplasty (ESG) has gained global adoption but our understanding of its mechanism(s) of action and durability of efficacy is limited. We sought to determine changes in gastric emptying (GE), gastric motility (GM), hormones and eating behaviours after ESG. Design A priori-designed single-centre substudy of a large US randomised clinical trial, adults with obesity were randomised to ESG or lifestyle interventions (LS) alone. We measured GE, hormones and weight loss and assessed eating behaviours. In a subset of ESG patients, we assessed GM. The primary outcome was the change in T1/2 (min) at 3 months, and secondary outcomes were changes in weight, GE, GM, hormones and eating behaviours. We used t-test analyses and regression to determine the association between GE and weight loss. Results 36 (ESG=18; LS=18) participated in this substudy. Baseline characteristics were similar between the two groups. At 3 months, T1/2 was delayed in the ESG group (n=17) compared with the LS group (n=17) (152.3±47.3 vs 89.1±27.9; p<0.001). At 12 months, T1/2 remained delayed in the ESG group (n=16) vs control group (n=14) (137±37.4 vs 90.1±23.4; p<0.001). Greater delays in GE at 3 months were associated with greater weight loss. GM was preserved and fasting ghrelin, glucagon-like peptide 1 and polypeptide YY significantly increased 18 months after ESG. Conclusion ESG promotes weight loss through several key mechanistic pathways involving GE and hormones while preserving GM. These findings further support clinical adoption of this technique for the management of obesity. Trial registration number NCT03406975.

Objective The diagnosis of autoimmune pancreatitis (AIP) is challenging. Sonographic and cross-sectional imaging findings of AIP closely mimic pancreatic ductal adenocarcinoma (PDAC) and techniques for tissue sampling of AIP are suboptimal. These limitations often result in delayed or failed diagnosis, which negatively impact patient management and outcomes. This study aimed to create an endoscopic ultrasound (EUS)-based convolutional neural network (CNN) model trained to differentiate AIP from PDAC, chronic pancreatitis (CP) and normal pancreas (NP), with sufficient performance to analyse EUS video in real time. Design A database of still image and video data obtained from EUS examinations of cases of AIP, PDAC, CP and NP was used to develop a CNN. Occlusion heatmap analysis was used to identify sonographic features the CNN valued when differentiating AIP from PDAC. Results From 583 patients (146 AIP, 292 PDAC, 72 CP and 73 NP), a total of 1 174 461 unique EUS images were extracted. For video data, the CNN processed 955 EUS frames per second and was: 99% sensitive, 98% specific for distinguishing AIP from NP; 94% sensitive, 71% specific for distinguishing AIP from CP; 90% sensitive, 93% specific for distinguishing AIP from PDAC; and 90% sensitive, 85% specific for distinguishing AIP from all studied conditions (ie, PDAC, CP and NP). Conclusion The developed EUS-CNN model accurately differentiated AIP from PDAC and benign pancreatic conditions, thereby offering the capability of earlier and more accurate diagnosis. Use of this model offers the potential for more timely and appropriate patient care and improved outcome.