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.

FOLFIRINOX (FX) and gemcitabine (GEM) plus nab-paclitaxel (GnP) have been reported as effective regimens for unresectable advanced pancreatic cancer (APC). FX may be more effective but is also associated with more adverse events (AEs). Therefore, first-line treatment with FX followed by second-line GnP may be appropriate. Aims: To assess the safety and efficacy of second-line GnP for patients with APC after first-line FX failure. Methods: This study was a multicenter prospective phase II study evaluating second-line GnP in patients with APC after failed first-line FX. The primary endpoint was response rate (RR), and the secondary endpoints were overall survival (OS), progression free survival (PFS), and the frequency and degree of adverse events (AEs). Results: Thirty patients (14 male; median age, 64 years) were enrolled. The RR was 13.3%, with a median follow-up time of 9.3 months. The median OS and PFS were 7.6 and 3.8 months, respectively. From the beginning of first-line treatment, the median OS and PFS were 14.2 and 9.3 months, respectively. Grade 3 or 4 AEs were seen in 70% of patients. Conclusion: Second-line GnP after FX failure for patients with APC could be more effective than GEM alone. Further comparison studies are warranted.

Background: Extracorporeal shock wave lithotripsy (ESWL) is a common treatment for pancreatic stones in chronic pancreatitis. In contrast, peroral pancreatoscopy-guided lithotripsy (POPS-L) remains underexplored, with limited comparative studies to ESWL. This study compared the treatment outcomes of disposable POPS-L tools and ESWL for pancreatic stones. Methods: A retrospective analysis was conducted on 66 patients who had undergone pancreatic stone treatment at three institutions between 2006 and 2022. The treatment outcomes of POPS-L and ESWL were compared. Results: This study included 19 and 47 patients who had undergone POPS-L and ESWL, respectively. In a comparison between POPS-L and ESWL, the stone clearance rates were 78.9% vs. 70.2% (p = 0.55), while the procedure-related complication rates were 21% vs. 6.3% (p = 0.09). The median total session counts were 1 vs. 5 (p < 0.01). The cumulative stone recurrence rates were comparable in both groups. Multivariate analysis revealed no significant factors influencing the stone clearance rates, and the choice between POPS-L and ESWL did not affect the stone clearance rates. Conclusions: POPS-L and ESWL exhibited comparable treatment outcomes in terms of stone clearance, complications, and recurrence rates. Furthermore, POPS-L is advantageous due to the need for fewer sessions to achieve pancreatic stone clearance.

Background: Contrast-enhanced endoscopic ultrasound (CE-EUS) is a promising diagnostic modality for differentiating malignant and benign lymph nodes. This study aimed to evaluate the diagnostic capability of CE-EUS in differentiating indolent non-Hodgkin’s lymphoma (NHL) from aggressive NHL. Methods: Patients who underwent CE-EUS and endoscopic ultrasound–guided fine needle aspiration (EUS-FNA) for lymphadenopathy and were diagnosed with NHL were included in this study. Echo features on B-mode endoscopic ultrasound (EUS) and vascular and enhancement patterns on CE-EUS were qualitatively evaluated. The enhancement intensity of the lymphadenopathy on CE-EUS over 60 s was also quantitatively evaluated using time–intensity curve (TIC) analysis. Results: A total of 62 patients who were diagnosed with NHL were enrolled in this study. Regarding qualitative evaluation using B-mode EUS, there were no significant differences in the echo features between aggressive NHL and indolent NHL. With regard to qualitative evaluation using CE-EUS, aggressive NHL showed a heterogeneous enhancement pattern that is significantly more frequent than indolent NHL (95% confidence interval: 0.57 to 0.79, p = 0.0089). When heterogeneous enhancement was defined as aggressive NHL, the sensitivity, specificity, and accuracy of the qualitative evaluation when using CE-EUS were 61%, 72%, and 66%, respectively. In TIC analysis, the velocity of reduction for homogeneous lesions was significantly higher in aggressive NHL than in indolent NHL (p < 0.0001). The sensitivity, specificity, and accuracy of CE-EUS in differentiating indolent NHL from aggressive NHL improved to 94%, 69%, and 82%, respectively, when combined with qualitative and quantitative evaluations. Conclusions: CE-EUS before EUS-FNA for mediastinal or abdominal lymphadenopathy may be useful for improving the diagnostic capability of differentiating between indolent NHL and aggressive NHL (clinical trial registration number: UMIN000047907).

The identification of anticancer therapies using next-generation sequencing (NGS) is necessary for the treatment of cholangiocarcinoma. NGS can be easily performed when cell blocks (CB) are obtained from bile stored overnight. We compared NGS results of paired CB and surgically resected specimens (SRS) from the same cholangiocarcinoma cases. Of the prospectively collected 64 bile CBs from 2018 to 2023, NGS was performed for three cases of cholangiocarcinoma that could be compared with the SRS results. The median numbers of DNA and RNA reads were 95,077,806 [CB] vs. 93,161,788 [SRS] and 22,101,328 [CB] vs. 24,806,180 [SRS], respectively. We evaluated 588 genes and found that almost all genetic alterations were attributed to single-nucleotide variants, insertions/deletions, and multi-nucleotide variants. The coverage rate of variants in SRS by those found in CB was 97.9–99.2%, and the coverage rate of SRS genes by CB genes was 99.6–99.7%. The NGS results of CB fully covered the variants and genetic alterations observed in paired SRS samples. As bile CB is easy to prepare in general hospitals, our results suggest the potential use of bile CB as a novel method for NGS-based evaluation of cholangiocarcinoma.