Diagnostic performance of digital breast tomosynthesis in female patients with nipple discharge

Abstract Background Nipple discharge is one of the most common symptoms related to the breast, but it is a presenting feature of breast cancer in 5%–12% of women. Aims The purpose of this study was to determine the diagnostic performance of digital breast tomosynthesis (DBT) in the evaluation of patients with nipple discharge and to compare it with mammography (MMG), ultrasound (US), and magnetic resonance imaging (MRI). Methods and Results This retrospective study included 53 patients with nipple discharge. All patients underwent DBT, and results were compared to MMG, breast US, and MRI. Radiological findings for each method were categorized according to BI‐RADS classification: categories 1–2 were considered negative and categories 3–5 positive. If a tissue specimen was obtained, the final diagnosis was established based on the results of histopathological analysis; otherwise, a clinical follow‐up was required for at least 2 years to confirm benign radiological findings. Measures of diagnostic accuracy of DBT, MMG, US, and MRI were calculated and compared. Results Final histopathological analysis revealed six malignant breast lesions, all of which were detected in patients with pathologic nipple discharge. DBT and MRI exhibited high sensitivity (100%) and high negative predictive value (100%) for the detection of breast cancer in patients with nipple discharge. DBT showed higher specificity compared to MRI (82.9% vs. 61.9%). Sensitivity and specificity of MMG were 83.3% and 76.6%, respectively. Breast US was determined to have a sensitivity of 66.7% and specificity of 57.5%. Conclusion DBT exhibited higher specificity than MRI at the same level of sensitivity and negative predictive value. Therefore, the use of DBT should be considered as an alternative to MRI in the assessment of patients with nipple discharge.


| INTRODUCTION
Nipple discharge is the third most reported complaint related to the breast after breast pain and palpable breast mass. 1 Up to 80% of women in their reproductive years will experience at least one episode of nipple discharge. 2 Although majority of these cases are of benign origin, nipple discharge can be a source of anxiety for patients and can cause concern in physicians. 3 It accounts for 2%-5% of medical visits among women, but most importantly, it is a presenting feature of breast cancer in 5%-12% of women. 4 Clinically, nipple discharge can be categorized as physiologic or pathologic. Physiologic discharge is usually bilateral. It can be transparent or colored (but never contains blood. The most common causes of physiologic nipple discharge are pregnancy, lactation, nipple stimulation, endocrine abnormalities, and medications. Pathologic nipple discharge is usually unilateral. It can be bloody or serous. Although pathologic nipple discharge can indicate the presence of breast cancer, it can also be caused by intraductal papilloma, duct ectasia, or mastitis. 5 Evaluation of non-lactating patient with nipple discharge should begin with thorough history and physical examination. Cytologic analysis of nipple discharge is not routinely recommended in diagnostic workup of nipple discharge. 4,6 If initial evaluation suggests physiologic nipple discharge, imaging examination is not indicated. Additional imaging examination is required for patients with pathologic nipple discharge due to the associated increased risk of malignancy. Initial diagnostic approach in the evaluation of patients with pathologic nipple discharge includes mammography (MMG) and ultrasound (US). 1,6 Magnetic resonance imaging (MRI) is appropriate for further evaluation of patients with negative MMG and US. 6,7 There are several studies evaluating digital breast tomosynthesis (DBT) galactography (ductography) for nipple discharge workup, comparing it with traditional 2D digital mammography ductography, but there were no reports published on the diagnostic performance of DBT in the evaluation of nipple discharge. 8,9 The aim of this study is to assess the value of DBT in evaluating patients with nipple discharge and to compare it with MMG, US, and MRI.

| METHODS
This retrospective study was granted approval by the institutional review board and all data had been fully anonymized before they were accessed. All procedures performed in this case series involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
The electronic radiology information system was reviewed, and 2361 female patients underwent DBT and mammography between July 2017 and May 2019 were identified as shown in Figure 1.
Patients who did not have nipple discharge (N = 2302), were lost to 2 years follow up (N = 3) and had poor quality DBT (N = 2) were not included in the study. Finally, 53 patients met the eligibility criteria and were included in the study. Imaging characteristics of radiologically detected suspicious lesions are given in Table 1.   14,20 In this study MRI showed the same level of sensitivity and NPV, but lower specificity compared to DBT.
These results suggest DBT as a suitable alternative imaging method to MRI in evaluation of patients with nipple discharge.
In this study all patients with an underlying malignancy presented with pathologic nipple discharge as reported in previous study. 16 The incidence of breast cancer (in situ or invasive) among patients with pathologic nipple discharge was 15.4%. These results are in accordance with previous studies, in which reported incidence ranged from 5% to 23%. 5,6,21 The most common malignant lesion detected in patients with pathologic nipple discharge was DCIS as it was reported in a previous study. 22 In this study sensitivity and specificity of cytology were 75.0% and 79.4%, respectively. In different studies sensitivity varies between 11.1% and 74.5% and specificity varies from 30.0% to 99.5%. 23,24 Cytologic analysis is considered inappropriate as the only modality for evaluation of nipple discharge due to its variability in sensitivity and specificity and therefore should always be supplemented with imaging methods. 6,23,24 There are several limitations to this study, most obvious being its retrospective design. Lack of statistical significance in sensitivity between DBT and other imaging method (MMG, US, MRI) may be due to a relatively sample size. Considering these limitations, prospective studies with larger sample size are warranted to establish accuracy of DBT in patients with nipple discharge.

| CONCLUSION
DBT is a valuable addition to the evaluation of patients with nipple discharge. In the case of a positive DBT finding, US has value as a method to rule out the malignancy. Since DBT exhibited greater specificity than MRI at the same level of sensitivity and negative predictive value, it should be considered as an alternative to MRI in diagnostic workup of patients with nipple discharge.