Touch-preparation cytologic examination of breast core biopsy specimens: Accuracy in predicting benign or malignant core histologic results

Abstract

The purpose of this study was to determine the accuracy of touch-preparation cytologic examination of breast core biopsy specimens in predicting benign or malignant core histologic results. One hundred two core biopsies were performed on 88 women with stereotactic or ultrasonographic (US) guidance. Slides were prepared by smearing one core sample on each slide, spraying the slides with fixative, and staining them with the Papanicolaou technique. Slides were blindly reviewed by a cytopathologist. Cytologic results were categorized as positive for malignancy, not diagnostic for malignancy, or insufficient for diagnosis. Results were correlated with histologic results from all specimens obtained during the core biopsy. Imaging depicted the lesions sampled for biopsy as masses (n = 70), clustered calcifications (n = 29), focal asymmetries (n = 2), or architectural distortion (n = 1). Touch-preparation slides of 87 (85%) lesions contained sufficient material for diagnosis. Cytologic results correctly identified 12 of 16 (three of five intraductal and nine of 11 invasive) malignancies in 10 of 13 masses and two of three clusters of calcifications. Two false-positive results occurred, both with fibroadenomas. Overall, touch-preparation studies produced 69 true-negative and four false-negative results. Excluding slides with insufficient material, the sensitivity, specificity, and accuracy of touch-preparation results were 75%, 97%, and 93%, respectively. Including insufficient samples, accuracy was 79%. Although touch-preparation cytologic examination of breast core biopsy specimens is fairly accurate in prediction of benign or malignant core histologic results, its correlation with histologic results is not sufficient to justify routine use in immediate counseling and treatment planning.