Introduction

Preoperative biliary drainage (PBD) is commonly used in patients with malignant distal biliary obstruction (MDBO) to maintain performance status, prevent cholangitis and allow subsequent surgical or neoadjuvant treatment. Endoscopic retrograde cholangiopancreatography (ERCP) is the first-line approach, but in cases of failure, percutaneous transhepatic biliary drainage (PTBD) or endoscopic ultrasound-guided biliary drainage (EUS-BD) are considered. While EUS-BD has shown favourable outcomes compared to PTBD in the palliative setting, evidence comparing these modalities in the preoperative setting is limited.

Aims & Methods

The aim of this study was to compare the efficacy, safety, and surgical outcomes of PTBD and EUS-BD in patients with MDBO after failed ERCP. This retrospective multicenter study included patients who underwent PTBD or EUS-BD followed by pancreaticoduodenectomy (PD) for periampullary cancer at 15 referral centers in Japan between January 2012 and December 2021. Only patients who had failed or were not amenable to ERCP were included. The primary outcome was surgery-related adverse events (AEs), while secondary outcomes included drainage efficacy, recurrence patterns, disease-free survival (DFS), and overall survival (OS). Propensity score matching (PSM) was used to adjust for baseline differences.

Results

In total, 2350 patients received PBD before PD. Of the 79 patients in whom ERCP failed, 65 underwent PTBD and 14 underwent EUS-BD. After PSM, 13 matched pairs were analyzed. Technical and clinical success rates were comparable between groups. However, the EUS-BD group had a significantly higher internalization success rate (100% vs. 46%, P=0.005) and fewer drainage sessions required (median: 1 vs. 2, P=0.015). The EUS-BD group also had a significantly shorter hospital stay for drainage (11 vs. 17 days, P=0.042). Rates of drainage-related AEs tended to be lower in the EUS-BD group, although not statistically significant. Surgery-related outcomes, including operative time, intraoperative blood loss, R0 resection rate, postoperative hospital stay, and surgery-related AEs, were not significantly different between groups. No cases of needle tract seeding were observed. Median DFS and OS were comparable between groups (DFS: 14.2 vs. 16.8 months, P=0.22; OS: 17.9 vs. 27.9 months, P=0.167).

Conclusion

EUS-BD showed comparable surgical outcomes and long-term oncological efficacy to PTBD, while offering superior internalization success, fewer interventions, and a shorter preoperative hospital stay. These advantages make EUS-BD a favorable alternative to PTBD in the preoperative management of MDBO after failed ERCP, particularly in patients scheduled for neoadjuvant therapy.

Disclosure

Kitano M has received grants from Boston Scientific, Zeon Medical Inc., and Medicos Hirata Inc. Takenaka M, Shiomi H, Kitagawa K, Shintani S, Maruyama H, Sagami R, Ikeura T, Ogura T, Ishida Y, Mandai K, Sugimori S, Imamura Y, Masuda A, Ikezawa K,Shimizu A, Nakai A, Nagai M and Nakano R have no conflicts of interest or financial ties to disclose.

Introduction

Endoscopic ultrasound (EUS) is the gold standard for evaluating pancreatic lesions, allowing for real-time imaging and biopsy sampling. Nevertheless, the differentiation between solid pancreatic lesions — such as pancreatic ductal adenocarcinoma (PDAC) and pancreatic neuroendocrine tumors (PNETs) — as well as cystic lesions is often troublesome. Specifically, identifying intraductal papillary mucinous neoplasms (IPMNs) with varying degrees of dysplasia is crucial for guiding clinical management. However, the diagnostic accuracy of current methods remains suboptimal. This multicenter study aimed to develop and validate convolutional neural networks (CNNs) to characterize solid and cystic pancreatic lesions and stratify IPMNs into high-grade dysplasia/carcinoma (HGD/C) and low-grade dysplasia (LGD).

Aims & Methods

This multicenter study included EUS images from 12 centers in Spain, Brazil, the United Kingdom and the United States of America. Solid lesion differentiation (PDAC vs. PNET) was achieved using a CNN trained on 49034 images from 95 patients (n=69 for PDAC and n=26 for PNET), with diagnoses confirmed via EUS-guided fine-needle biopsy or surgical specimen analysis. A second CNN was trained on 79395 images from 82 patients (n=64 for mucinous cystic lesions and n=18 for non-mucinous cystic lesions). Finally, a third model used the subset of IPMNs (n=51046 frames, 30 patients) to distinguish between HGD/C (n=5 patients) and LGD (n=25 patients). Histological confirmation was used as the gold standard. Performance metrics included sensitivity, specificity, accuracy, and area under the precision-recall curve (AUC-PR). To comprehensively determine the model's performance, we evaluated multiple thresholds for the confidence score and the IoU, between 0.45 and 0.55 with a step of 0.02 ([0.45, 0.47, 0.50, 0.52, 0.55]). The datasets were split into training (70%), validation (20%), and testing (10%) subsets.

Results

For solid lesions, the model differentiated PDAC from PNETs with an accuracy of 91.8%, with sensitivities of 96.2% and 97.8%, respectively. The AUC-PR was 0.982 for PDAC and 0.977 for PNET. Regarding cystic lesions the model detected mucinous cystic lesions with a sensitivity of 97.5%, an overall accuracy of 89.7% and an AUC-PR of 0.977. When the subset of IPMNs was specifically analyzed, the model distinguished IPMNs with HGD/C from those with LGD with a sensitivity of 97.5%, specificity of 94.3%, and an overall accuracy of 89.8%. The AUC was 0.988.

Conclusion

This is the first multicenter, transatlantic AI study aiming to provide a holistic approach to both solid and cystic pancreatic lesions, focusing on those with a significantly higher malignant potential. Our models achieved high accuracy in identifying and differentiating these conditions. Furthermore, to our knowledge this is the first integrative model to include the ability to stratify IPMNs into LGD and HGD/C, which has significant clinical implications, potentially reducing unnecessary surgeries and ensuring timely intervention for high-risk patients. The inclusion of data from diverse regions reduces demographic bias and enhances model generalizability. Real-time implementation of these solutions could significantly improve decision-making in EUS and optimize patient outcomes.