INTRODUCTION: Immune checkpoint inhibitors (ICIs) are being increasingly used to treat advanced malignancies. ICI-induced pancreatic injury (ICI-PI), which is an immune-related adverse event that may be a risk factor of ICI-associated pancreatitis, is not well documented in the literature. METHODS: Consecutive patients who received ICIs for advanced malignancies from August 2015 through October 2022 were analyzed for the incidence of ICI-PI based on the Common Terminology Criteria for Adverse Events and ICI-associated pancreatitis. The imaging, clinical, and pathological findings of ICI-associated pancreatitis were also assessed. RESULTS: This study enrolled 843 patients. In multivariable analyses, dual or simultaneous immunotherapy and ≥10 cycles of ICI administration were significant predictive factors for all grades of pancreatic injury, including grade ≥3. Notably, patients who received simultaneous immunotherapy exhibited a higher incidence of grade ≥3 pancreatic injuries compared with those receiving asynchronous immunotherapy in univariable analysis (P = 0.032). One-fifth of the patients (16/70) with grade ≥3 pancreatic injuries had imaging evidence of pancreatitis similar to mild acute pancreatitis. ICI-associated pancreatitis was observed in 5.7% (48/843) of patients, including 1.8% (15/843) with moderate-to-severe pancreatitis (grade ≥2). Symptomatic cases (0.36%, 3/843) were treated with steroids with favorable outcomes. Immunohistochemistry for CD4 and CD8 revealed greater infiltration of CD8⁺ than CD4⁺ lymphocytes. DISCUSSION: Simultaneous immunotherapy and dual immunotherapy are risk factors of ICI-PI. Although most patients diagnosed with ICI-PI and ICI-associated pancreatitis were asymptomatic and had a low mortality likelihood, long-term outcomes, including endocrine and exocrine function, should be carefully monitored.

Pancreatic cancer is a fatal disease, and thus its chemoprevention is an important issue. Based on the recent report that patients with allergic diseases have a low risk for pancreatic cancer, we examined the potential chemopreventive effect of anti-allergic agents using a hamster pancreatic carcinogenesis model. Among the three anti-allergic drugs administered, montelukast showed a tendency to suppress the incidence of pancreatic cancer. Further animal study revealed a significantly decreased incidence of pancreatic cancer in the high-dose montelukast group compared with controls. The development of the pancreatic intraepithelial neoplasia lesions was also significantly suppressed. The Ki-67 labeling index was significantly lower in pancreatic carcinomas in the high-dose montelukast group than in controls. In vitro experiments revealed that montelukast suppressed proliferation of pancreatic cancer cells in a dose-dependent manner with decreased expression of phospho-ERK1/2. Montelukast induced G1 phase arrest. Conversely, leukotriene D₄ (LTD₄), an agonist of CYSLTR1, increased cellular proliferation of pancreatic cancer cells with an accumulation of phospho-ERK1/2. In our cohort, pancreatic ductal adenocarcinoma patients with high CYSLTR1 expression showed a significantly unfavorable clinical outcome compared with those with low expression. Our results indicate that montelukast exerts a chemopreventive effect on pancreatic cancer via the LTD₄–CYSLTR1 axis and has potential for treatment of pancreatic carcinogenesis.

Histological evidence is essential for diagnosing malignant biliary strictures. However, conventional brush cytology remains the primary method used worldwide, despite its low diagnostic sensitivity and accuracy, as it is technically easy, rapid, and cost-effective. The aim of this study was to evaluate the diagnostic performance of a recently introduced endoscopic scraper, the simplicity of which is comparable to that of a conventional brush, by comparing diagnostic yields and the number of collected cells. The sensitivity of the endoscopic scraper when using the cell block technique was significantly higher than when using brush cytology or a brush with the cell block technique (53.6% vs. 30.9%, p < 0.001; 53.6% vs. 31.6%, p = 0.024, respectively). Quantitative digital image analysis of cell block sections revealed that the median number of cells obtained with the endoscopic scraper was significantly higher than when using the brush (1917 vs. 1014 cells, p = 0.042). Furthermore, seven cases (8.3%; 7/84) were diagnosed by immunohistochemical analysis of the cell block section obtained from the endoscopic scraper. Given its simplicity and greater capacity for sample acquisition, use of the endoscopic scraper in conjunction with the cell block technique could replace brush cytology for the histological diagnosis of malignant biliary strictures.