Fine-needle aspiration of the breast: a review of 1,995 cases with emphasis on diagnostic pitfalls.

Between 1985 to 1989, 1,995 fine-needle aspirations of palpable breast lesions were performed at our institution. In all cases, the aspirates were procured by cytopathologists using 22- or 23-gauge needles. Direct smears were immediately stained with Diff-Quik and Papanicolaou and assessed for specimen adequacy (criteria as followed in this institution). Tissue follow-up was available in 1,117 cases. The cytologic diagnoses rendered in these cases were: malignant, 690 cases (60.2%); suspicious for carcinoma, 49 cases (4.3%); benign, 343 cases (29.9%), and insufficient specimen, 35 cases (3.1%). There were 28 false-negative and 2 false-positive results. Considering only cases definitively diagnosed as benign or malignant, the sensitivity was 96%, specificity 99%, positive predictive value 99%, negative predictive value 94%, and overall efficiency 97%. Of these specimens considered suspicious, only 11 cases (22%) were proved not to be malignant after excisional biopsy. These were three fibroadenomas, three ductal hyperplasias, two adenosis tumors, two mucocele-like lesions, and one nipple adenoma. The two lesions that resulted in true false-positive diagnoses were an apocrine cyst with atypia and sclerosing adenosis with radial scar. The clinical and cytologic features of the benign conditions that resulted in false suspicious and positive diagnoses and those features that distinguish them from carcinoma are presented.