Endoscopic ultrasound-guided fine-needle aspiration has become an indispensable tool for the diagnostic workup and biological characterization of lesions arising within or adjacent to the gastrointestinal tract. However, in certain clinical scenarios, the presence of interposed vascular structures may impede direct access to the target lesion, limiting the feasibility of conventional approaches. In recent years, transvascular EUS-guided puncture has emerged as a technically feasible and potentially safe diagnostic alternative.

The aim of this study was to evaluate the feasibility, safety profile, and diagnostic performance of EUS-guided transvascular fine-needle aspiration, including portal thrombus puncture, in patients undergoing workup for suspected malignant disease in a tertiary referral centre.

A retrospective case series was conducted including consecutive patients who underwent EUS-guided transvascular aspiration of suspicious lesions between 2014 and 2025. Demographic characteristics, comorbidities, procedural variables (including needle gauge, type of access, target anatomy, and number of needle passes), and peri-procedural management were recorded. Diagnostic performance (sensitivity, specificity, and accuracy) was assessed based on cytopathological results, including immunohistochemical profiling when available. Procedural safety was evaluated by documenting adverse events and classifying them according to severity.

Sixteen patients were included, with a median age of 69.5 years; 75% were male. Most patients were classified as ASA II, and the mean Charlson Comorbidity Index was 4.75. Anticoagulant or antiplatelet therapy had been prescribed in 25% of cases and was temporarily discontinued before the procedure. Targets included suspected malignant portal thromb (50%), pancreatic cystic lesions (25%), lymphadenopathy (12.5%), and solid masses (12.5%). Access routes comprised transportal puncture (31.25%), non-portal transvascular routes (18.75%), and direct puncture of portal thrombi (50%). A 22G needle was used in 62.5% of procedures, employing a slow-pull technique with two to three passes. A diagnostic specimen was successfully obtained in 81% of cases, sensitivity of 81.8% (95% CI 52.3–94.9), specificity of 100% (95% CI 56.6–100), and overall accuracy of 87.5% (95% CI 63.0–96.8). Malignancy was confirmed in 56.25% of patients, and immunohistochemistry was feasible in all cytologically malignant specimens.

A total of 43.75% of procedures were performed in hospitalized patients and 56.25% on an outpatient basis. Only one patient required admission due to acute pancreatitis, which resolved with conservative management. The composite adverse event rate was 12.5%, and no statistically significant association was found between adverse events and age, sex, ASA score, Charlson Index, or anticoagulant use (p > 0.15).

EUS-guided transvascular puncture appears to be a feasible and safe technique with high diagnostic performance in selected clinical scenarios involving difficult-to-access lesions or malignant thrombi. The favourable outcomes and low complication rate observed in this cohort support the incorporation of transvascular access as a complementary diagnostic strategy in advanced endoscopic practice. Furthermore, these findings raise the possibility that transvascular access may eventually extend beyond diagnostic applications and into therapeutic endoscopic interventions.