Hooft, Jeanin E. van

AbstractBackgroundSurveillance of pancreatic cysts focuses on the detection of (mostly morphologic) features warranting surgery. European guidelines consider elevated CA19.9 as a relative indication for surgery. We aimed to evaluate the role of CA19.9 monitoring for early detection and management in a cyst surveillance population.MethodsThe PACYFIC‐registry is a prospective collaboration that investigates the yield of pancreatic cyst surveillance performed at the discretion of the treating physician. We included participants for whom at least one serum CA19.9 value was determined by a minimum follow‐up of 12 months.ResultsOf 1865 PACYFIC participants, 685 met the inclusion criteria for this study (mean age 67 years, SD 10; 61% female). During a median follow‐up of 25 months (IQR 24, 1966 visits), 29 participants developed high‐grade dysplasia (HGD) or pancreatic cancer. At baseline, CA19.9 ranged from 1 to 591 kU/L (median 10 kU/L [IQR 14]), and was elevated (≥37 kU/L) in 64 participants (9%). During 191 of 1966 visits (10%), an elevated CA19.9 was detected, and these visits more often led to an intensified follow‐up (42%) than those without an elevated CA19.9 (27%; p < 0.001). An elevated CA19.9 was the sole reason for surgery in five participants with benign disease (10%). The baseline CA19.9 value was (as continuous or dichotomous variable at the 37 kU/L threshold) not independently associated with HGD or pancreatic cancer development, whilst a CA19.9 of ≥ 133 kU/L was (HR 3.8, 95% CI 1.1–13, p = 0.03).ConclusionsIn this pancreatic cyst surveillance cohort, CA19.9 monitoring caused substantial harm by shortening surveillance intervals (and performance of unnecessary surgery). The current CA19.9 cutoff was not predictive of HGD and pancreatic cancer, whereas a higher cutoff may decrease false‐positive values. The role of CA19.9 monitoring should be critically appraised prior to implementation in surveillance programs and guidelines.

Main Recommendations 1 ESGE recommends a prolonged course of a prophylactic broad-spectrum antibiotic in patients with ascites who are undergoing therapeutic endoscopic ultrasound (EUS) procedures.Strong recommendation, low quality evidence. 2 ESGE recommends placement of partially or fully covered self-expandable metal stents during EUS-guided hepaticogastrostomy for biliary drainage in malignant disease.Strong recommendation, moderate quality evidence. 3 ESGE recommends EUS-guided pancreatic duct (PD) drainage should only be performed in high volume expert centers, owing to the complexity of this technique and the high risk of adverse events.Strong recommendation, low quality evidence. 4 ESGE recommends a stepwise approach to EUS-guided PD drainage in patients with favorable anatomy, starting with rendezvous-assisted endoscopic retrograde pancreatography (RV-ERP), followed by antegrade or transmural drainage only when RV-ERP fails or is not feasible.Strong recommendation, low quality evidence. 5 ESGE suggests performing transduodenal EUS-guided gallbladder drainage with a lumen-apposing metal stent (LAMS), rather than using the transgastric route, as this may reduce the risk of stent dysfunction.Weak recommendation, low quality evidence. 6 ESGE recommends using saline instillation for small-bowel distension during EUS-guided gastroenterostomy.Strong recommendation, low quality evidence. 7 ESGE recommends the use of saline instillation with a 19G needle and an electrocautery-enhanced LAMS for EUS-directed transgastric endoscopic retrograde cholangiopancreatography (EDGE) procedures.Strong recommendation, low quality evidence. 8 ESGE recommends the use of either 15- or 20-mm LAMSs for EDGE, with a preference for 20-mm LAMSs when considering a same-session ERCP.Strong recommendation, low quality evidence.

Main Recommendations 1 ESGE recommends the use of endoscopic ultrasound-guided biliary drainage (EUS-BD) over percutaneous transhepatic biliary drainage (PTBD) after failed endoscopic retrograde cholangiopancreatography (ERCP) in malignant distal biliary obstruction when local expertise is available. Strong recommendation, moderate quality evidence. 2 ESGE suggests EUS-BD with hepaticogastrostomy only for malignant inoperable hilar biliary obstruction with a dilated left hepatic duct when inadequately drained by ERCP and/or PTBD in high volume expert centers. Weak recommendation, moderate quality evidence. 3 ESGE recommends that EUS-guided pancreatic duct (PD) drainage should only be considered in symptomatic patients with an obstructed PD when retrograde endoscopic intervention fails or is not possible. Strong recommendation, low quality evidence. 4 ESGE recommends rendezvous EUS techniques over transmural PD drainage in patients with favorable anatomy owing to its lower rate of adverse events. Strong recommendation, low quality evidence. 5 ESGE recommends that, in patients at high surgical risk, EUS-guided gallbladder drainage (GBD) should be favored over percutaneous gallbladder drainage where both techniques are available, owing to the lower rates of adverse events and need for re-interventions in EUS-GBD. Strong recommendation, high quality of evidence. 6 ESGE recommends EUS-guided gastroenterostomy (EUS-GE), in an expert setting, for malignant gastric outlet obstruction, as an alternative to enteral stenting or surgery. Strong recommendation, low quality evidence. 7 ESGE recommends that EUS-GE may be considered in the management of afferent loop syndrome, especially in the setting of malignancy or in poor surgical candidates. Strong recommendation, low quality evidence. 8 ESGE suggests that endoscopic ultrasound-directed transgastric ERCP (EDGE) can be offered, in expert centers, to patients with a Roux-en-Y gastric bypass following multidisciplinary decision-making, with the aim of overcoming the invasiveness of laparoscopy-assisted ERCP and the limitations of enteroscopy-assisted ERCP. Weak recommendation, low quality evidence.

Novel artificial intelligence techniques are emerging in all fields of healthcare, including gastroenterology. The aim of this review is to give an overview of artificial intelligence applications in the management of pancreatic diseases. We performed a systematic literature search in PubMed and Medline up to May 2020 to identify relevant articles. Our results showed that the development of machine‐learning based applications is rapidly evolving in the management of pancreatic diseases, guiding precision medicine in clinical, endoscopic and radiologic settings. Before implementation into clinical practice, further research should focus on the external validation of novel techniques, clarifying the accuracy and robustness of these models.