Background: The recent improvement of peroral cholangioscopy (POCS) maneuverability has enabled the precise, targeted biopsy of bile duct lesions under direct cholangioscopic vision. However, as only small-cup biopsy forceps can pass through the scope channel, the resulting small sample size may limit the pathological diagnosis of biopsy specimens. This study compared the diagnostic abilities of POCS-guided biopsy and conventional fluoroscopy-guided biopsy for bile duct cancer. Method: This multicenter, retrospective cohort study included patients exhibiting bile duct stricture with suspected cholangiocarcinoma in whom POCS-guided and fluoroscopy-guided biopsies were performed in the same session. The primary endpoint was the diagnostic sensitivity for malignancy. The size and quality of the biopsy specimens were also compared. Result: A total of 59 patients were enrolled. The sensitivity of POCS-guided biopsy was similar to that of fluoroscopy-guided biopsy (54.0% and 64.0%, respectively). However, when the modalities were combined, the sensitivity increased to 80.0%. The mean specimen size from POCS-guided biopsy was significantly smaller than that from fluoroscopy-guided biopsy. The specimen quality using fluoroscopy-guided biopsy was also better than that using POCS-guided biopsy. Conclusions: The diagnostic sensitivity of POCS-guided biopsy is still insufficient, mainly because of the limited specimen quantity and quality. Therefore, conventional fluoroscopy-guided biopsy would be helpful to improve diagnostic sensitivity.

Backgrounds: Few data are available for surveillance decisions focusing on factors related to mortality, as the primary outcome, in intraductal papillary mucinous neoplasm (IPMN) patients.Aims: We aimed to identify imaging features and patient backgrounds associated with mortality risks by comparing pancreatic cancer (PC) and comorbidities.Methods: We retrospectively conducted a multicenter long‐term follow‐up of 1864 IPMN patients. Competing risk analysis was performed for PC‐ and comorbidity‐related mortality.Results: During the median follow‐up period of 5.5 years, 14.0% (261/1864) of patients died. Main pancreatic duct ≥5 mm and mural nodules were significantly related to all‐cause and PC‐related mortality, whereas cyst ≥30 mm did not relate. In 1730 patients without high‐risk imaging features, 48 and 180 patients died of PC and comorbidity. In the derivation cohort, a prediction model for comorbidity‐related mortality was created, comprising age, cancer history, diabetes mellitus complications, chronic heart failure, stroke, paralysis, peripheral artery disease, liver cirrhosis, and collagen disease in multivariate analysis. If a patient had a 5 score, 5‐ and 10‐year comorbidity‐related mortality is estimated at 18.9% and 50.2%, respectively, more than 7 times higher than PC‐related mortality. The model score was also significantly associated with comorbidity‐related mortality in a validation cohort.Conclusions: This study demonstrates main pancreatic duct dilation and mural nodules indicate risk of PC‐related mortality, identifying patients who need periodic examination. A comorbidity‐related mortality prediction model based on the patient's age and comorbidities can stratify patients who do not require regular tests, especially beyond 5 years, among IPMN patients without high‐risk features.Clinical trial registration: T2022‐0046.