To describe the clinical characteristics, primary tumor sites, endosonographic features, and diagnostic performance of endoscopic ultrasound-guided tissue acquisition (EUS-TA) in patients with confirmed pancreatic metastases evaluated over a 12-year period in a tertiary oncology center in Brazil.

Retrospective observational study including all consecutive patients with histologically or clinically confirmed pancreatic metastases who underwent EUS-guided fine-needle aspiration or biopsy between April 2013 and February 2025. Data collected from electronic medical records comprised demographics, primary tumor origin, lesion size and pancreatic location, and final histopathological/immunohistochemical diagnosis. Diagnostic accuracy was calculated using definitive histology of the pancreatic lesion or unequivocal clinical/imaging progression of known metastatic disease in other organs as the reference standard.

Among 433 patients who underwent 491 pancreatic EUS-TA procedures during the study period, 21 procedures (4.8%) in 20 patients were confirmed as pancreatic metastases. Patients were predominantly female (80%) with a median age of 65 years (IQR 60–70). Mean lesion size was 30.2 mm (range 8–76 mm; median 30 mm). Lesions were located in the pancreatic head in 47.6%, body 33.3%, tail 14.3%, uncinate process 4.8%, and neck 28.6% (some lesions involving more than one segment).

The most frequent primary tumors were colorectal adenocarcinoma (23.8%), breast carcinoma (23.8%), melanoma (14.3%), renal cell carcinoma (14.3%), ovarian carcinoma (9.5%), lung carcinoma (9.5%), and thyroid carcinoma (4.8%).

EUS-TA yielded 18 true-positive diagnoses (85.7%) and 3 false-negative results (14.3%), with no false positives, resulting in an overall diagnostic accuracy of 85.7% (95% CI 64.9–95.4%). Immunohistochemical studies were crucial for lineage confirmation in all true-positive cases.

Pancreatic metastases are uncommon (4.8% of all pancreatic EUS-TA procedures in a high-volume oncology center).  Although the literature classically describes pancreatic metastases as predominantly originating from renal cell carcinoma and melanoma, in this 12-year Latin-American cohort colorectal and breast primaries were jointly the most frequent origins (each 23.8%), highlighting potential regional differences in referral patterns and oncological epidemiology. Lesions are typically solitary, average 3 cm in diameter, and predominate in the pancreatic head.

In this 12-year cohort, EUS-guided tissue acquisition achieved a diagnostic accuracy of 85.7%, driven largely by the systematic application of immunohistochemical panels that allowed precise diagnosis. Accurate distinction between primary pancreatic adenocarcinoma and metastatic disease carries major therapeutic and prognostic implications. Therefore, EUS-TA with adequate sampling and routine immunohistochemical work-up remains the cornerstone for correct diagnosis and appropriate oncological management in patients with suspected pancreatic metastases. These data reinforce the importance of maintaining a high index of suspicion for secondary pancreatic involvement in oncology patients and support EUS-TA as the diagnostic modality of choice in this setting.