A retrospective study of the diagnostic accuracy of fine‐needle aspiration for breast lesions and implications for future use

Abstract

In recent years, the use of fine-needle aspiration (FNA) in the diagnosis of breast lesions has declined in many institutions. We sought to evaluate the role of FNA for breast lesions and the annual rate of the procedure at our institution over a 4(1/2) year period (May 2002-October 2006). A total of 831 FNAs were performed, with 258 (31%) having histologic follow-up. The number of FNAs obtained was 159 from 5/02 to 4/03, 192 from 5/03 to 4/04, 194 from 5/04 to 4/05, 191 from 5/05 to 4/06, and 95 from 5/06 to 10/06. Each case was placed into one of four categories: nondiagnostic (9%), benign (77.5%), atypical/suspicious (5.5%), or malignant (8%). Surgical tissue was available for 37% of nondiagnostic cases, 22% of benign cases, 80% of atypical/suspicious cases, and 72% of malignant cases. The overall sensitivity and specificity for FNA was 83 and 92% respectively. The overall positive and negative predictive values were 83 and 92% respectively. There were no false-positive cases, indicating a positive predictive value of 100% for a Dx of malignancy. For cases with surgical follow-up, the false-negative rate was 5.4%. Although there is a national trend away from FNAs of breast lesion, this has not been the experience at our institution. Although FNA may not be ideal in the initial evaluation of suspicious lesions, we argue that FNA for clinically benign palpable lesions and recurrent carcinomas has significant value.