Implications of nipple discharge in Hong Kong Chinese women.

INTRODUCTION

There are no recent data on nipple discharge and its association with malignancy in Hong Kong Chinese women. This study reported our 5-year experience in the management of patients with nipple discharge, and our experience of mammography, ultrasonography, ductography, and nipple discharge cytology in an attempt to determine their role in the management of nipple discharge.

METHODS

Women who attended our Breast Clinic in a university-affiliated hospital in Hong Kong were identified by retrospective review of clinical data from January 2007 to December 2011. They were divided into benign and malignant subgroups. Background clinical variables and investigative results were compared between the two subgroups. We also reported the sensitivity, specificity, and positive and negative predictive values of the investigations that included mammography, ultrasonography, ductography, and cytology.

RESULTS

We identified 71 and 31 patients in the benign and malignant subgroups, respectively. The median age at presentation for the benign subgroup was younger than that of the malignant subgroup (48 vs 59 years; P=0.003). A higher proportion of patients in the malignant subgroup than the benign subgroup presented with blood-stained nipple discharge (87.1% vs 47.9%; P=0.002). Mammography had a specificity of 98.4% and positive predictive value of 66.7%; ultrasonography had a specificity of 87.0% and negative predictive value of 75.0%. Cytology and ductography were sensitive but lacked specificity. Ductography had a negative predictive value of 100% but a low positive predictive value (14.0%). Clinical variables including age at presentation, duration of discharge, colour of discharge, presence of an associated breast mass, and abnormal sonographic findings were important in suggesting the underlying pathology of nipple discharge. Multiple logistic regression showed that blood-stained discharge and an associated breast mass were statistically significantly more common in the malignant subgroup.

CONCLUSIONS

In patients with non-blood-stained nipple discharge, a negative clinical breast examination combined with negative imaging could reasonably infer a benign underlying pathology.