| Abstract: | In the past 10 years, 7,495 cytological breast fine-needle aspirations (FNAs) were performed (4,756 FNAs of solid nodes and 2,739 of cystic nodes). Of these, 2,099 cases underwent surgery; 650 (31%) had histologically proven carcinoma. Sensitivity was 83.9%, specificity was 99.5%, the predictive value for negative results was 93.2% and for positive results was 98.6%, and the accuracy was 94.6%. Inadequate (13.3%) and doubtful samples (8.1%) were excluded from calculation. False-negative results (82 cases) mainly resulted from sampling errors. False suspicious results (six cases) lessened with increasing experience in breast pathology and with the application of strict diagnostic criteria, but most likely they will never reach zero. Frozen-section diagnosis could be bypassed only in selected cases. Guidelines on the role of FNA in management of solid breast lesions are given. FNA deserves further evaluation in diagnosing early stages of breast carcinoma: sensitivity was 7.5% in 57 carcinomas in situ, 67.5% in 55 minimally invasive carcinoma, and 92.7% in 538 nonminimally invasive carcinomas. |
