Accuracy of aspiration cytology in detecting breast cancer.

To evaluate the role of needle placement in determining accuracy and to establish criteria for clinical decision making, we used aspiration cytology to diagnose 49 palpable breast masses in patients undergoing excisional biopsy. Closed cytology, obtained by percutaneous aspiration, and open cytology, obtained by aspiration of the excised mass, were compared with histology. Nineteen (39%) of the breast masses were carcinoma. Unsatisfactory cytologic findings were frequent after both closed (37%) and open (24%) aspiration. Most of these aspirates were from mammary dysplasia, suggesting that acellularity rather than needle placement was responsible. Closed aspiration was falsely negative in one patient and falsely suggestive of cancer in five. If atypical, suspicious or malignant cytologic findings were considered to be carcinoma; closed cytology had high sensitivity (94%), low specificity (64%), and a false positive fraction of 36%. The positive predictive value was 76% and the negative predictive value 90%. Accuracy was 81%. If only suspicious or malignant results were considered positive, specificity and positive predictive value increased at the expense of sensitivity and negative predictive value but accuracy did not change. Even with optimal needle placement (open aspiration), accuracy of aspiration cytology was less than 90%.