Purpose: to evaluate core biopsy (CB) for the diagnosis of subclinical breast lesions, comparing with surgical biopsy previously identified by stereotaxic mammography. Methods: this is a cross-sectional study of 41 subclinical lesion cases over 35 years of age, between January 1995 and February 1997 at the Instituto de Ginecologia da Universidade Federal do Rio de Janeiro. The cases were classified mammographically as benign, probably benign, suspicious and malignant. Benign and probably benign lesions were studied together for statistical purposes. The histopathologic diagnosis of CB was classified as inadequate for diagnosis, absence of malignancy, suspicious and malignant. The histopathologic diagnosis of the surgical biopsy was classified as absence of malignancy, pre-malignant and malignant. The sensitivity, especificity and predictive values of CB were evaluated. Mammography likelihood ratio and core biopsy likelihood ratio were evaluated to predict breast cancer. Results: CB coincided with surgical biopsy in 86.2% of the 29 cases of absence of malignancy. All cases suspected by CB were malignant by surgical biopsy. All cases malignant by CB were also malignant by surgical biopsy. CB sensitivity and specificity were 36.4% and 100%, respectively. Positive predictive value was 100% and negative predictive value was 78.1%. In the group classified mammographically as malignant the likelihood ratio was 9.7; for suspicious lesions it was 1.3 and for probably benign lesions it was 0.1. Core biopsy likelihood ratio was infinite (¥) for suspicious and malignant lesions, 0.4 for cases classified as absence of malignancy and 1.4 for inadequate for diagnosis cases. Conclusions: after analysis of the results, with the use of the likelihood ratio, we conclude that CB report of absence of malignancy did not allow to rule out the diagnosis of malignancy. In these cases, we should correlate the result with mammography. If CB shows absence of malignancy and there is no correlation with mammography, the research must be continued. When the report of CB was suspicious, the probability of a breast carcinoma was very high. In these cases, we should perform a surgical biopsy to establish a definitive diagnosis, because an infiltrating carcinoma needs a different therapy when compared with carcinoma in situ and atypical hyperplasia. In the cases of histopathologic report of malignancy, the probability of breast cancer was high, since we did not observe any false positive CB. In these cases, CB allowed a quick diagnosis without the need of surgical biopsy.
Breast neoplasm; Cancer; Mammography
