Evaluation of the Diagnostic Accuracy of Different Imaging Modalities in Patients with Pathologic Nipple Discharge
Riku Togawa, Dina Batarseh, Helena Dahm, Thomas M. Deutsch, André Hennigs, Andrea Lemke, André Pfob, Walid Shaalan, Anne Stieber, Oliver Zivanovic, Jörg Heil, Michael Golatta, Fabian RiedelBackground/Objectives: To evaluate the diagnostic accuracy of different imaging modalities in patients with pathologic nipple discharge. Methods: This is a single-centre retrospective analysis of patients with pathologic nipple discharge. All performed imaging modalities were reviewed for their sensitivity and specificity in correctly diagnosing malignancy, as well as lesions with uncertain malignant potential (B3). Results: Of 388 patients, pathology or follow-up confirmed 41 cases of malignancy (15 breast cancer, 26 ductal carcinoma in situ (DCIS)), 151 cases of B3 lesions, and 196 benign cases. The combination of ultrasound and mammography/tomosynthesis achieved a sensitivity of 82.5% (95% CI 68.1–91.3) and a specificity of 83.5% (77.4–88.2) for the detection of malignancy. For the classification of B3 lesions as suspicious (BI-RADS 4/5), the sensitivity and specificity were 33.8% (26.7–41.7) and 83.5% (77.4–88.2), respectively. Ductography demonstrated a sensitivity of 89.3% (72.8–96.3) and a specificity of 13.3% (8.2–20.8) for malignancy, while MRI showed a sensitivity of 71.4% (45.4–88.3) and a specificity of 90.0% (76.9–96.0). In classifying B3 lesions as suspicious, ductography reached a sensitivity of 94.8% (89.7–97.5) and a specificity of 13.3% (8.2–20.8), compared to 75.9% (57.9–87.8) and 90.0% (76.9–96.0), respectively, for MRI. Conclusions: While the combination of ultrasound and mammography/tomosynthesis correctly classified 82.5% of breast malignancies associated with pathologic nipple discharge as suspicious, their sensitivity for classifying B3 lesions as suspicious was only moderate (33.8%). In patients with negative or inconclusive conventional imaging but persistent clinical suspicion, MRI may be considered as an additional diagnostic tool, whereas ductography, given its low specificity, is of limited value for assessing lesion dignity and may primarily serve to localize the affected duct prior to surgery. Further research is needed to determine which diagnostic modality provides the greatest benefit in specific cases.