Background: Nipple discharge is a common breast complaint. Because most nipple discharge is a result of benign processes, less-invasive, nonsurgical diagnostic modalities have been explored to reduce the need for surgical intervention.
Methods: Between September 1994 and December 2000, 108 female patients aged 25 to 77 years underwent duct excision for bloody or clear nipple discharge. Results of various preoperative studies were compared with surgical pathology to determine the sensitivity, specificity, positive predictive value, and negative predictive value of these tests in detecting malignant ductal pathology.
Results: Of the 108 surgical histopathology specimens, 90 of 108 patients were benign, 5 of 108 patients were atypical, and 13 of 108 patients were malignant. The sensitivity of mammography was 57.1%, specificity was 61.5%, positive predictive value was 16.7%, and negative predictive value was 91.4%. Hemoccult sensitivity was 50%, specificity was 0%, positive predictive value was 20%, and negative predictive value was 0%. The sensitivity of ductography was 0%, specificity was 90%, positive predictive value was 0%, and negative predictive value was 81.8%. The sensitivity of cytology was 11.1%, specificity was 96.3%, positive predictive value was 50%, and negative predictive value was 76.5%.
Conclusions: Preoperative evaluations by mammography, Hemoccult, ductography, and cytology are poor predictors of histological diagnosis. These data suggest that patients with pathologic nipple discharge should undergo surgical biopsy for accurate diagnosis.