PURPOSE: Because of the increasing use of sonography to rule out cancer in women with palpable breast abnormalities, this study was performed to determine the rate of sonographically occult malignancy in this clinical setting. METHODS: Women who were recommended for biopsy based on mammographic and/or clinical findings underwent breast sonography. This study retrospectively analyzed the subset of patients with palpable malignant lesions. Lesions were classified as visible or occult on mammography and sonography. Patients without a tissue diagnosis of tumor were excluded. RESULTS: Of 1,346 masses that underwent biopsy or aspiration, 616 lesions were palpable, and of these, 293 were malignant. Sonography detected all 293 palpable malignant lesions (95% confidence interval for sensitivity, 99-100%). Eighteen lesions were mammographically occult. The median lesion size as determined by sonography was 1.8 cm; for the lesions that were mammographically occult, the median size was 1.6 cm. The most common histopathologic diagnosis for both groups of lesions was infiltrating ductal carcinoma. CONCLUSIONS: All palpable malignant breast lesions were visible by sonography in patients in whom a biopsy was recommended. However, we caution that until the false-negative rate of sonography for equivocal palpable abnormalities is determined prospectively, sonography cannot be accurately applied to rule out malignancy in this setting. Copyright 2000 John Wiley & Sons, Inc.
PURPOSE: Because of the increasing use of sonography to rule out cancer in women with palpable breast abnormalities, this study was performed to determine the rate of sonographically occult malignancy in this clinical setting. METHODS:Women who were recommended for biopsy based on mammographic and/or clinical findings underwent breast sonography. This study retrospectively analyzed the subset of patients with palpable malignant lesions. Lesions were classified as visible or occult on mammography and sonography. Patients without a tissue diagnosis of tumor were excluded. RESULTS: Of 1,346 masses that underwent biopsy or aspiration, 616 lesions were palpable, and of these, 293 were malignant. Sonography detected all 293 palpable malignant lesions (95% confidence interval for sensitivity, 99-100%). Eighteen lesions were mammographically occult. The median lesion size as determined by sonography was 1.8 cm; for the lesions that were mammographically occult, the median size was 1.6 cm. The most common histopathologic diagnosis for both groups of lesions was infiltrating ductal carcinoma. CONCLUSIONS: All palpable malignant breast lesions were visible by sonography in patients in whom a biopsy was recommended. However, we caution that until the false-negative rate of sonography for equivocal palpable abnormalities is determined prospectively, sonography cannot be accurately applied to rule out malignancy in this setting. Copyright 2000 John Wiley & Sons, Inc.