Nipple discharge is a frequent presenting complaint at breast clinics. Bloody nipple discharge (BND) has the highest risk of malignancy, albeit low. If mammogram and ultrasound are unrevealing, central duct excision (CDE) has been considered the gold standard in its management. Magnetic resonance imaging (MRI) has been widely confirmed as a highly sensitive test for detection of breast cancer, with an accompanying high negative predictive value. This article presents a retrospective review of patients with BND and negative conventional imaging, comparing outcome of patients who went directly to CDE without MRI to those patients who underwent preoperative MRI. Of 115 patients who underwent mammography and US alone prior to CDE, eight cancers were detected (seven ductal carcinoma in situ [DCIS] and 1 IDC, 7 mm [T1b]; incidence: 7%). Of 85 patients who underwent conventional imaging followed by MRI prior to surgery, eight cancers were detected (all DCIS; incidence: 9.4%), seven of which were identified by MRI. The one false-negative MRI had subtle findings which, in retrospect, were misinterpreted; however, a clinically apparent nipple lesion prompted surgical biopsy. Of 56 patients with a negative or benign MRI, CDE was negative for malignancy in all but that one patient. Sensitivity and specificity were 87.5%/71.4%. Positive predictive value and negative predictive value (NPV) were 24.1%/98.2%. MRI should be performed in all patients with BND and negative conventional imaging. The extremely high NPV of MRI suggests that a negative study could obviate CDE in most patients unless overriding clinical factors prevail.
Keywords: bloody nipple discharge; breast MRI; central duct excision.
© 2015 Wiley Periodicals, Inc.