ENDOSONOGRAPHY-GUIDED REAL-TIME FINE-NEEDLE ASPIRATION BIOPSY

Twenty-six patients were prospectively evaluated with endosonography-g uided real-time fine-needle-aspiration biopsy. This cohort comprised 1 4 patients with a pancreatic mass revealed by CT or a stricture of the main pancreatic duct seen at ERCP, 7 patients with mediastinal lympha denopathy, 3 patients with extrapancreatic abdominal masses, and 2 pat ients with subepithelial or infiltrative lesions. Endosonography-guide d real-time fine-needle-aspiration biopsy was diagnostic in 18 of 20 p atients in whom surgical confirmation was available or in whom maligna ncy was found and confirmed by clinical follow-up (accuracy of 90%). I n the subgroup of patients with pancreatic lesions, 3 had previously u ndergone nondiagnostic CT-guided fine-needle-aspiration biopsy and 2 d id not have evidence of a mass by CT. Real-time fine-needle-aspiration biopsy was diagnostic for malignancy in 4 of these individuals. In th e 7 patients with mediastinal lymph nodes, 2 had nondiagnostic transbr onchial biopsy and 2 had no evidence of mediastinal lymphadenopathy by CT scan. Endosonography-guided real-time fine-needle-aspiration biops y diagnosed malignancy in both individuals with nondiagnostic transbro nchial studies and was able to identify mediastinal lymphadenopathy in the 2 patients with negative CT scans (malignancy confirmed with real -time fine-needle-aspiration biopsy in 1). Overall, in 9 of 10 lesions in which visualization by CT was not possible (5), CT-guided fine-nee dle aspiration was unsuccessful (3), or prior nonsurgical biopsy techn iques were unsuccessful (2), real-time fine-needle-aspiration biopsy w as diagnostic. No complications were experienced in this series. Endos onography-guided real-time fine-needle-aspiration biopsy should be con sidered as a primary or secondary method for establishing a tissue dia gnosis in patients with mediastinal and abdominal masses.