Enhancing the diagnostic performance of repeated endoscopic ultrasound-guided tissue acquisition combined with surrogate repeated endoscopic retrograde pancreatography for small pancreatic cancer Running title: EUS-TA and ERP for pancreatic cancer

Abstract Purpose Diagnosing pancreatic tumors ≤10 mm is challenging due to limited visualization and low sampling sensitivity. This study aimed to evaluate the cumulative diagnostic performance of repeated endoscopic ultrasound-guided tissue acquisition (EUS-TA) and surrogate repeated endoscopic retrograde pancreatography (ERP). Methods This study analyzed 40 patients with suspected pancreatic tumors ≤10 mm who underwent EUS-TA and/or ERP retrospectively. When a diagnosis could not be determined based on the initial EUS-TA or ERP procedure, EUS-TA or ERP was repeated as necessary. The cumulative diagnostic performance of EUS-TA and ERP for pancreatic tumors was evaluated. Results EUS-TA was performed once for 35 cases, twice for seven cases, and three times for one case. ERP was performed for 15 cases, and the median number of ERP attempts was two (range, 1-8). The cumulative sensitivity of EUS-TA increased from 56.7% to 70.0% after three attempts, while ERP sensitivity rose from 54.5% to 72.7% after two attempts. The cumulative diagnostic performance of repeated EUS-TA and ERP combined by case included sensitivity and accuracy rates of 87.9% and 90.0%. When limited to pancreatic cancer, the sensitivity and accuracy rates were 95.8% and 96.8%, respectively.No severe adverse events occurred. Conclusions Repeated EUS-TA and ERP significantly improved diagnostic sensitivity for small pancreatic tumors. When a single EUS-TA or ERP procedure fails to establish a diagnosis in suspected malignant cases, repeating the procedures may be warranted. In cases where initial procedures are inconclusive, repeated application of both methods may enable accurate pathological diagnosis and inform optimal treatment strategies.