LUMINAL PLENARY: OR-LUM-02: Endoscopic ultrasound-guided fine needle aspiration for liver lesions: Comparative results in a large series of more than 150 patients with primary and secondary tumors

BACKGROUND AND OBJECTIVES: A large part of the liver is now accessible for endoscopic ultrasound-guided fine needle aspiration (EUS-FNA), but there are limited data in the literature evaluating efficacy and safety of the technique. PATIENTS AND METHODS: We performed a 5-year review of all EUS-FNAs performed in primitive and secondary hepatic lesions. Results were shown as mean (range) and statistics were calculated as per-protocol or intention-to-treat (ITT) analyses. RESULTS: A total of 152 patients (81 males/71 females), with a mean age of 63 years (35–86), with 158 lesions were examined either for primitive masses (n = 50) or for metastases (n = 102). Sensitivity for malignancy was 96% and 91%, specificity 100 and 100%, positive predictive value 100 and 100%, and negative predictive value 40% and 22% in the per-protocol and ITT analyses, respectively. Results were significantly better (P < 0.05) in secondary versus primitive masses. Liver EUS-FNA provided diagnosis in ten patients without evident primary location and in four pancreatic adenocarcinomas without contributive pancreatic FNA. The 25-gauge needles were significantly less sensitive (P < 0.01) than 22- or 19-gauge needles. Lesions with a smaller size and hilar location (P < 0.01) were more frequently associated with false negative results. Bleeding was the sole complication observed in 3% of patients, mainly in primitive lesions, and statistically associated with histology needles (P < 0.001). CONCLUSIONS: EUS-FNA in the liver is highly sensitive for the diagnosis of malignancy in primitive and secondary masses and is helpful in determining the primary origin of liver metastases. Histology needles might be preferred for diagnosis, but their used was associated with a small but significant increase of bleeding.