Search

Background: Fine-needle aspiration cytology (FNAC) has been used extensively in the diagnosis of breast lesions, but false-negative rates are a matter of concern. Immediate onsite evaluation of breast lesions, combining FNAC results with clinical and radiological data allows improving its diagnostic accuracy. The objective of this study was to evaluate the diagnostic accuracy of a large series of FNAC of breast lesions during the first 3-year period of a dedicated one-stop clinic, by comparing the FNAC results with the corresponding definitive histological examination outcome or the results of the radiological follow-up at 18 months. Method: Data of all consecutive patients (pts) whose lesions were prospectively characterized by ultrasonography (US) and diagnosed by FNAC in the one stop clinic of the Institute Gustave Roussy between May 2004 and March 2007. Histological verification by core-needle biopsy or surgery was systematically performed for lesions classified malignant or suspicious by cytopathology, and for benign lesions, when no perfect concordance between clinico-radiologic features and FNAC results (such as benign FNAC but BI-RAD 5) was found. For non-operated patients, follow-up consisted on biannual US and/or mammography; the results at 18 months were considered. Pts characteristics, radiological findings, cytopathological and histological results were extracted from the hospital computerized prospectively registered medical records. Complete sensitivity (number of carcinomas that were malignant or suspicious by cytopathology) and full specificity (benign lesions at cytology of the total number of benign lesions) were calculated by comparing the cytopathological results to the histological or the long-term follow-up status. Results: A total of 1822 nodular breast lesions (mean size: 20 mm, BI-RAD ACR 1/2/3/4/5/unknown: 10/96/471/459/777/9) in 1739 pts (mean age: 56 years) were studied. FNAC was US-guided in 1115 lesions (61 %). Lesions were classified by FNAC as malignant in 842 (46%), benign in 771 (42%), suspicious in 154 (9%), and unsatisfactory in 55 (3%) cases. Complete sensitivity (number of carcinomas that were malignant or suspicious by cytopathology) was estimated at 98.5%, whereas full specificity (benign lesions at cytology of the total number of benign lesions) was equal to 94.4%. There were 4 false-positive lesions and 32 false-negative lesions, but all these lesions except one were operated as radiological results were suspicious of malignity. Conclusions: Breast FNAC performed in a dedicated one-stop clinic with immediate on-site diagnosis presented a low rate of unsatisfactory specimens and a very efficient tool for triaging patients candidates for histological evaluation conversely for follow-up by imaging. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P6-01-05.