Pancreatic cancer represents a growing global health burden, making early and reliable diagnosis crucial. Solid pancreatic lesions require histological sampling to confirm diagnosis and guide treatment. Endoscopic ultrasound-guided fine needle biopsy (EUS-FNB) is now essential for this purpose.

We aimed to evaluate the technical success, histological yield, and diagnostic performance of EUS-FNB for solid pancreatic masses in a single tertiary center.

We conducted a retrospective study of all patients who underwent EUS-FNB for suspected solid pancreatic lesions between 2018 and 2025. Cystic lesions and chronic pancreatitis were excluded. Malignancy was defined by histological identification of malignant cells (EUS-FNB or surgical specimens) or clinico-radiological follow-up demonstrating tumor progression and/or metastases. Technical failure was defined as obtaining insufficient sample or other normal tissue.

Overall, 380 patients were included (236 men, 144 women; mean age 62.8 years [29–88]). Mean lesion size was 34.2 mm [6–88]. Lesions were located in the pancreatic head (67.9%), uncinate process (11.8%), body (11.7%), neck (6.4%), and tail (2.2%). Vascular contact and frank invasion were observed in 55% and 11.8% respectively. The approach was predominantly transduodenal, with transgastric approach in 24 patients (6.3%). A 22-gauge needle was used in nearly all procedures.

Technical success was achieved in 367/380 cases (96.6%) and significantly correlated with lesion size (p = 0.02). Malignancy was confirmed in 353 patients (93%). Pancreatic adenocarcinoma accounted for 92% of malignant tumors, predominantly ductal subtype (88%). Signet-ring cell and mucinous components were identified in 17% and 4.6% respectively. Other malignancies included 16 pancreatic neuroendocrine tumors (4.2%), two mucinous cystadenocarcinomas, two undifferentiated carcinomas, one sarcomatoid carcinoma, one solid pseudopapillary neoplasm, one degenerated intraductal papillary mucinous neoplasm, and ten pancreatic metastases from other primary sites. Thirteen lesions (3.4%) represented mass-forming chronic pancreatitis.

EUS-FNB demonstrated excellent diagnostic performance: sensitivity 97% (95% CI 94.6–98.5), positive predictive value 99.3%, and diagnostic accuracy 96.6% (95% CI 94.2–98.2). Second biopsy was performed in 6 patients (1.6%), yielding neoplasia in 5 cases (four adenocarcinomas, one neuroendocrine tumor) and one benign lesion, with cumulated sensitivity and accuracy remaining essentially unchanged. A single hemorrhagic complication (0.3%) occurred in a patient on antiplatelet therapy.

EUS-FNB demonstrates excellent technical success (96.6%) and diagnostic accuracy (96.6%) for solid pancreatic lesions in routine practice. It provides high-quality histological cores with minimal needle passes and a low complication rate.