Value of EUS-FNA cytological preparations compared with cell block sections in the diagnosis of pancreatic solid tumours.

摘要

OBJECTIVE: Endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) is performed in order to achieve a definite tissue diagnosis of pancreatic lesions. This in turn is a guide to the appropriate treatment for the patient. Tissue samples collected by the same needle for cytological preparations and cell block histological sections (often referred to as FNA-cytology and FNA-biopsy, respectively) are handled differently. The specific contribution of each of these tests was evaluated. METHODS: One hundred and two consecutive patients underwent EUS-FNA while being investigated for pancreatic solid lesions. Diagnosis was made by cytology, cell block sections or both. The diagnosis was confirmed by clinical outcome. RESULTS: Male/female ratio was 61/41. Mean age was 65+/-12 years (range, 22-94). Mean lesion size was 3.1+/-1.8 cm (range, 0.6-10 cm); 68% were >2 cm and 75% were located in the pancreatic head. The average number of needle passes was two (range, 1-4 passes). Final tissue diagnosis was malignant in 66 (65%) patients. Sensitivity, specificity and accuracy were 73%, 94% and 81%, respectively, for cytology alone, and 63%, 100% and 78%, for cell blocks alone. Eighty-two patients (80%) had cytology and cell blocks, which matched in 64 (78%) patients. EUS-FNA results that relied on both techniques had 84% sensitivity, 94% specificity and 88% accuracy. Cytology revealed 13 malignancies not diagnosed on cell blocks, while cell blocks revealed five malignancies not diagnosed by cytology. Malignant lesions were more common in men; they were larger in size and located in the pancreatic head. CONCLUSION: EUS-FNA cytology was more sensitive than cell blocks but less specific for the diagnosis of solid pancreatic lesions. The two methods are complementary and implementing both improves the diagnostic value of EUS-FNA.