Journal of Hepato‐Biliary‐Pancreatic Sciences最新文献

Background: Despite advances in understanding the risk factors of post-endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP), this adverse event remains frequently unpredictable. This study aims to develop a machine learning (ML) model to predict PEP risk.

Methods: Data were collected from a prospective ERCP registry on patients with naïve papilla who underwent ERCP between 2022 and 2024. CatBoost and eXtreme Gradient Boosting algorithms were trained to estimate PEP risk and the performance of the resulting models was assessed using the area under the receiver operating characteristic (AUC) with 10-fold cross-validation.

Results: Of 1330 screened patients, 1190 met the inclusion criteria, and 170 (14.3%) developed PEP. The best-performing algorithm was CatBoost, which consisted of eight features: age, sex, normal papilla morphology, pancreatic duct cannulation, difficult cannulation, abnormal bilirubin levels, common bile duct diameter, and successful stone extraction. This model achieved an AUC of 68.8% (70.4% sensitivity, 67.2% specificity, 26.5% positive predictive value, and 92.0% negative predictive value). The CatBoost model effectively stratified patients into low-, intermediate-, and high-risk groups, with corresponding PEP incidences of 5.7%, 21.0%, and 40.0%, respectively.

Conclusions: ML is highly promising for prediction of PEP. Future studies should focus on multicenter data, inclusion of multimodal data, severity risk-stratification, and real-time application.

Introduction: Various guidelines recommend endoscopic stone removal for asymptomatic common bile duct stones (CBDS); however, previous studies have indicated that asymptomatic CBDS is associated with a higher incidence of post-ERCP pancreatitis (PEP). Our study aimed to compare the clinical outcomes of ERCP between elderly patients with asymptomatic and symptomatic CBDS.

Methods: Elderly patients with CBDS were divided into two groups: A (asymptomatic CBDS) and S (symptomatic CBDS). Propensity score matching was performed to reduce possible bias in the baseline characteristics between the two groups, yielding 221 matched patients. Primary outcomes included early adverse events (AEs) during ERCP, while secondary outcomes included technical success rate and late AEs.

Results: Regarding ERCP-related early AEs, the incidence of PEP was 6.3% in group A and 4.5% in group S, with no significant difference between the groups. The technical success rates were 87.8% and 84.6% in groups A and S, respectively, with no significant difference. The most common late AE was the recurrence of CBDS (11.3% in group A versus 18.6% in group S [p = 0.04]); Kaplan-Meier analysis also revealed a tendency toward a higher cumulative incidence in group S.

Conclusion: Endoscopic transpapillary stone extraction in asymptomatic elderly patients may be beneficial.

Background and aims: This study aimed to evaluate adverse events (AEs) for endoscopic ultrasound-guided biliary drainage (EUS-BD) and identify risk factors for early AEs and recurrent biliary obstruction (RBO).

Methods: A multicenter retrospective study was conducted using a common database of 21 Japanese referral centers.

Results: A total of 616 patients who underwent EUS-BD, including endoscopic ultrasound-guided choledochoduodenostomy (n = 107), hepaticogastrostomy (n = 487), and hepaticojejunostomy (n = 22), for malignant biliary obstruction were analyzed. Early AEs occurred in 13.6% of patients. Independent risk factors for all AEs included procedure time ≥ 32 min (odds ratio [OR] 1.82) and antiplatelet/anticoagulant use (OR 2.15). A risk factor for peritonitis included electrocautery use (OR 3.87), while bleeding risk was increased with antiplatelet/anticoagulant use (OR 7.19) and performance status > 2 (OR 5.26). The use of plastic stents was associated with a higher risk of a shorter time to RBO. AE and RBO rates did not significantly differ among the three EUS-BD approaches.

Conclusions: Patients on antiplatelet and/or anticoagulation therapy should be aware of the increased risk of AEs of EUS-BD. In addition, it is important to minimize procedure time, avoid the use of electrocautery, and use a metal stent to prevent early AEs and RBO.

Background: Application of artificial intelligence (AI) in intraoperative imaging has been expanding rapidly. The surgeon's gazing point indicates the exact site of surgical procedures and concentrates critical information for AI applications. This study aimed to develop a machine learning-based system to automatically detect the surgeon's gazing point from surgical video data.

Methods: Surgical instruments were detected using a deep-learning model applied to images extracted from pancreaticoduodenectomy videos. Gazing points were estimated through a mathematical algorithm based on the axes and intersections of detected instruments, and time-averaging was applied to enhance stability in real-time analysis. After validation using pancreaticoduodenectomy cases, the system was subsequently applied to extended cholecystectomy and distal pancreatectomy cases to evaluate its applicability to other procedures.

Results: Surgical instrument detection yielded AP50 of 60.5%. Gaze points detection achieved accuracies of 82.7% and 93.9% within 216- and 324-pixel radii (9.42% and 21.2% of a 1440 × 1080 screen) in pancreaticoduodenectomy. When applied to extended cholecystectomy and pancreaticoduodenectomy distal pancreatectomy, our system demonstrated comparable performance, with an accuracy of 85.5% within the 324-pixel radius. Time averaging improved accuracy, particularly with a 5-s average.

Conclusions: Our system successfully detected the surgeon's gaze point across procedures, suggesting potential utility in future AI-assisted surgery.

Background: Continuous glucose monitoring (CGM) is increasingly used for glycemic control, but its role after total pancreatectomy (TP) in pancreatogenic diabetes mellitus (PDM) is understudied. This study evaluated postoperative CGM utility in this population.

Methods: Thirty-seven TP patients (2022-2024) at Seoul National University Hospital were grouped by CGM use (CGM users = 11) and analyzed for baseline and perioperative diabetes-related factors. In a preoperative non-diabetic subgroup (n = 10, CGM users = 6), pre- and postoperative hemoglobin A1c (HbA1c) was compared by CGM use. For consenting CGM users, changes in time in range (TIR), time above range (TAR), and time below range (TBR) from 1 to 3 months postoperatively were analyzed.

Results: CGM users had lower postoperative HbA1c than non-users overall (6.5% vs. 8.0%; p < 0.001) and in non-diabetics (6.0% vs. 7.4%; p = 0.002). Among 80% sharing CGM data, TIR increased from 62.6% at 1 month to 71.8% at 3 months (p = 0.035), TAR decreased from 36.3% to 26.6% (p = 0.037), and TBR was unchanged.

Conclusions: Proactive CGM use may improve postoperative glycemic management in PDM after TP by increasing awareness of glycemic patterns, resulting in lower HbA1c and greater time in the target range.

Background: Extrahepatic endoscopic ultrasound-guided rendezvous (EUS-RV) with narrow common bile duct (CBD), as encountered in benign biliary diseases, can be technically challenging. We evaluated its modification (no aspiration, no contrast) in this difficult clinical scenario.

Methods: EUS-RV using conventional versus modified technique in patients with benign diseases and narrow CBD (≤ 6 mm) with difficult biliary cannulation (European Society of Gastrointestinal Endoscopy definition) was evaluated. The primary outcome was technical success, and secondary outcomes were overall EUS-RV core time, radiation dose, and adverse events.

Results: Overall, 113 patients were evaluated (conventional 49; modified 64). The modified EUS-RV had numerically higher technical success (83.7% vs. 95.3%; p = 0.05), which was significantly higher when the CBD access diameter was ≤ 3 mm (61.1% vs. 96.9%; p = 0.002). Moreover, the modified EUS-RV had shorter total core EUS-RV time (median 10 vs. 6.12 min; p < 0.001) and overall radiation exposure (median 345 vs. 208.93 mGy; p < 0.001), with comparable overall adverse events. On multivariate analysis, the predictors for technical success were the use of the modified technique (aOR: 6.26; p = 0.02) and CBD access diameter (aOR: 5.56; p = 0.04).

Conclusions: Modified EUS-RV technique enables efficient and rapid biliary access in cases of narrowed CBD without the need for contrast injection and bile aspiration.