Background: Contrast-enhanced endoscopic ultrasound (CE-EUS) is useful for the differentiation of pancreatic tumors. Using deep learning for the segmentation and classification of pancreatic tumors might further improve the diagnostic capability of CE-EUS. Aims: The aim of this study was to evaluate the capability of deep learning for the automatic segmentation of pancreatic tumors on CE-EUS video images and possible factors affecting the automatic segmentation. Methods: This retrospective study included 100 patients who underwent CE-EUS for pancreatic tumors. The CE-EUS video images were converted from the originals to 90-s segments with six frames per second. Manual segmentation of pancreatic tumors from B-mode images was performed as ground truth. Automatic segmentation was performed using U-Net with 100 epochs and was evaluated with 4-fold cross-validation. The degree of respiratory movement (RM) and tumor boundary (TB) were divided into 3-degree intervals in each patient and evaluated as possible factors affecting the segmentation. The concordance rate was calculated using the intersection over union (IoU). Results: The median IoU of all cases was 0.77. The median IoUs in TB-1 (clear around), TB-2, and TB-3 (unclear more than half) were 0.80, 0.76, and 0.69, respectively. The IoU for TB-1 was significantly higher than that of TB-3 (p < 0.01). However, there was no significant difference between the degrees of RM. Conclusions: Automatic segmentation of pancreatic tumors using U-Net on CE-EUS video images showed a decent concordance rate. The concordance rate was lowered by an unclear TB but was not affected by RM.

Pancreatic cancer is one of the most lethal cancers. To improve its prognosis, conversion surgery for initially unresectable advanced pancreatic cancer (UAPC) after chemotherapy has been reported in recent years. Methods: A retrospective analysis of the patients with initially UAPC underwent conversion surgery after the first-line modified FOLFIRINOX (mFX) was conducted at a single tertiary care center between January 2014 and March 2020. Results: Among 79 patients with UAPC who had mFX, 8 patients with a median age of 63 years, including 5 males (3 with locally advanced and 5 metastatic lesions), underwent conversion surgery after a median of 20 cycles of mFX. Conversion surgery was performed in 10.1% of patients (8/79) and surgical resection was successful in all with R0 resection. Postoperative major adverse events were seen in 2 patients, but no perioperative deaths were recognized. Recurrence was confirmed in 3 patients, and these 3 patients died due to cancer recurrence in 17.7, 30.6 and 57.8 months after mFX initiation. 5 patients were still alive without recurrence. The median OS in the patients who underwent conversion surgery was estimated as 65.9 months and was significantly longer than that of the patients without conversion surgery or that in the patients who had a partial response for mFX but did not have conversion surgery. The median follow-up period for the patients who had conversion surgery was 35.2 months. Conclusion: Conversion surgery achieved long-term survival in patients with UAPC who were treated with the first-line mFX, although controversy still remained.

Background/Objectives: The presence of a mural nodule (MN) is one of the findings indicating malignant transformation of an intraductal papillary mucinous neoplasm (IPMN). It is difficult to distinguish true MNs from mucus clots (MCs) by endoscopic ultrasound (EUS) alone. This study aimed to evaluate the efficacy of contrast-enhanced (CE)-EUS for differentiating true MNs from MCs and carcinoma from adenoma. Methods: A total of 104 patients who were diagnosed as having branch duct-type IPMNs with MN-like structures by EUS and underwent CE-EUS between January 2016 and August 2022 were included. MN-like structures without perfusion on CE-EUS were defined as MCs and those with perfusion were defined as true MNs. This was a retrospective study with limited pathological confirmation, and diagnoses in non-surgical cases were based on imaging and follow-up. Results: CE-EUS showed MN-like structures with perfusion in 35 patients and without perfusion in 69 patients. Surgical resection was eventually performed in a total of 28 patients and the diagnostic sensitivity, specificity and accuracy of MNs among them were 100%, 66.7% and 96.4% in CE-EUS; 48%, 66.7% and 50% in CE-CT; and 61.9%, 33.3% and 58.3% in MRCP, respectively. Possible risk factors indicating malignancy were statistically evaluated and presence of an MN was the only significant factor. Among the 35 true MNs, the height of an MN in carcinoma was significantly higher than that of an adenoma. The ROC analysis for detecting carcinoma in true MNs showed an area under the curve of 0.92 with the optimal cut-off value of 7 mm. When this cut-off value was used for diagnosing carcinoma, the sensitivity, specificity, and accuracy were 94.1%, 83.3% and 88.6%, respectively. Conclusions: CE-EUS may be useful for differentiating true MNs from MCs, although diagnostic performance should be interpreted cautiously because most non-surgical cases lacked pathological confirmation.

Background/Objectives: The optimal timing for biliary drainage in elderly patients remains controversial. This study compared the clinical outcomes of early versus late biliary drainage in elderly patients diagnosed with acute cholangitis associated with common bile duct stones (CBDS). Methods: This single-center, retrospective cohort study compared early (within 24 h of admission) versus late (>24 h after admission) biliary drainage using endoscopic retrograde cholangiopancreatography (ERCP) in 143 elderly patients diagnosed with acute cholangitis, who were divided into early and late groups, respectively. Results: There were no statistical differences in patient characteristics between the early and late drainage groups. Although the time to clinical success was significantly shorter in the early drainage group, there were no significant differences in persistent organ failure, length of hospital stay, in-hospital and 30-day mortality, and ERCP-related adverse events (AEs) between the early and late groups. In the late group, the clinically worsened group tended to exhibit higher white blood cell counts and lower platelet counts on admission. Conclusions: Time to clinical success was significantly shorter in the early drainage group than in the late group. No statistically significant differences in persistent organ failure, length of hospital stay, in-hospital and 30-day mortality, and ERCP-related adverse AEs were detected between early and late biliary drainage in elderly patients with acute cholangitis associated with CBDS.

Background: Steroids are the first‐line treatment for autoimmune pancreatitis (AIP); however, the impact of lower initial doses on relapse remains unclear. Methods: Ninety patients were divided into high‐dose (HD; > 0.4 mg/kg/day) and low‐dose (LD; ≤ 0.4 mg/kg/day) induction groups. The treatment response, relapse rate, and predictors of relapse were evaluated. Results: Sixty‐two and 28 patients received HD and LD steroid induction. The treatment response (100% vs. 100%, p > 0.99), relapse rates (33.9% vs. 32.1%, p > 0.99) and relapse‐free survival (hazard ratio [HR]: 0.86; log‐rank test, p = 0.99) did not differ significantly between the groups. Multivariable analysis identified discontinuation of maintenance treatment within 3 years (HR: 5.32, p = 0.0004), larger pancreatic volume (HR: 1.02, p = 0.01), and a smaller IgG4 reduction rate at 1 month after initiation of steroid treatment (HR: 0.99, p = 0.02) as independent predictors of relapse. In contrast, LD steroid induction was not identified as a risk factor for relapse (HR: 0.99, p = 0.99). Conclusions: There were no significant differences between LD and HD steroid induction in terms of treatment efficacy or long‐term outcomes in patients with AIP.