Accurate preoperative evaluation of the nipple–areola complex (NAC) is essential for surgical planning in breast cancer, as its involvement may preclude nipple-sparing procedures. Imaging plays a central role in identifying infiltration and guiding clinical decisions. Magnetic resonance imaging (MRI) is widely used for this purpose, while contrast-enhanced mammography (CEM) has emerged as an alternative approach with increasing clinical use. Comparative performance in routine practice remains central to determining its role in preoperative assessment.
Study Design and Imaging Approach
A retrospective cohort included 195 women with biopsy-proven breast cancer who underwent preoperative imaging at a single institution between January and December 2022. Ninety-one women received CEM and 104 underwent MRI. Allocation to each modality followed institutional clinical practice: CEM was generally used for women aged 60 years or older or for those with contraindications to MRI, while MRI was used for other patients. Histopathological examination of surgical specimens served as the reference standard for determining NAC involvement.
CEM examinations were performed using a dedicated mammography system following intravenous injection of iodinated contrast medium. The imaging protocol included bilateral cranio-caudal and medio-lateral oblique projections shortly after contrast administration, followed by additional targeted views of the affected breast. MRI examinations used 1.5-T scanners with patients positioned prone in a bilateral coil. The protocol incorporated diffusion-weighted imaging, T2-weighted sequences and dynamic contrast-enhanced imaging obtained after administration of gadolinium contrast.
Must Read: Mammography Shifts Breast Cancer Stage and Mortality
Two radiologists independently reviewed the imaging examinations, focusing on the nipple–areola complex. Evaluation incorporated several radiologic criteria, including nipple retraction or invasion, abnormal NAC enhancement, tumour-to-nipple enhancement, peri-areolar skin thickening, asymmetry or abnormal morphology of the NAC and tumour-to-nipple distance. Ancillary characteristics such as tumour size, enhancement pattern and breast density were also documented. Each case received a binary classification indicating the presence or absence of suspected NAC involvement.
Diagnostic Performance and Imaging Predictors
Histopathological examination confirmed NAC infiltration in 10 of 91 women in the CEM cohort and in 20 of 104 women in the MRI cohort. The difference between groups did not reach statistical significance. Sensitivity for detecting NAC involvement measured 60% for CEM and 50% for MRI. Specificity was nearly identical for both modalities at approximately 96%. Positive and negative predictive values were also comparable between the two imaging techniques.
Radiologists demonstrated high consistency when interpreting imaging findings. Agreement between readers reached more than 90% for both modalities, indicating strong reproducibility in identifying suspicious NAC features.
Multivariable analysis identified distinct predictors of NAC involvement for each modality. In the CEM model, peri-areolar skin thickening emerged as the strongest independent predictor. In the MRI model, abnormal enhancement of the nipple–areola complex and tumour-to-nipple distance were significantly associated with NAC infiltration. Increasing tumour-to-nipple distance reduced the likelihood of involvement, indicating a protective effect as the separation between tumour and nipple increased.
Several imaging features occurred more frequently in NAC-positive cases across both modalities. Abnormal NAC enhancement, peri-areolar thickening, tumour-to-nipple enhancement and nipple retraction all appeared among the findings associated with infiltration. These features correspond to radiologic indicators used in clinical practice to evaluate potential extension of tumour toward the nipple.
Clinical Context and Interpretation
The distribution of tumour histology showed that ductal carcinoma in situ frequently appeared in cases with NAC infiltration. In situ components were present in most NAC-positive cases in both imaging groups, although statistical analysis did not identify this factor as an independent predictor of involvement.
The overall frequency of NAC involvement measured 11.0% in the CEM cohort and 19.2% in the MRI cohort. These values align with previously reported ranges for NAC infiltration in breast cancer. Differences between the groups reflect variations in patient characteristics, including age distribution, because MRI was primarily performed in younger patients while CEM was used more frequently in older individuals or those unable to undergo MRI.
Sensitivity values for MRI in this dataset were lower than those reported in several previous analyses. Variations in tumour characteristics likely contributed to this difference. Some earlier reports included a higher proportion of invasive tumours with direct nipple invasion, whereas the present cohort contained more cases with in situ components that may produce subtler enhancement patterns on imaging.
Despite these differences in patient populations, the comparison of diagnostic performance revealed no statistically significant disparity between CEM and MRI. Both modalities achieved high specificity, suggesting that suspicious NAC findings strongly correspond with histopathological infiltration. This level of diagnostic confidence supports the role of imaging in guiding surgical decisions about nipple preservation.
CEM and MRI demonstrate comparable diagnostic performance in the preoperative evaluation of nipple–areola complex involvement in breast cancer. Sensitivity and specificity values are similar across modalities, and both techniques show strong reader agreement when applying established imaging criteria. Specific radiologic features, including peri-areolar thickening, NAC enhancement and tumour-to-nipple distance, contribute to identifying potential infiltration and can support surgical planning. The high specificity observed for both modalities reinforces the reliability of imaging findings when suspicion arises. CEM also offers advantages in accessibility, acquisition speed and patient tolerability, making it a viable option for individuals who cannot undergo MRI or when MRI availability is limited. Continued evaluation in prospective populations and the development of standardised imaging criteria may further clarify the role of each modality in assessing NAC involvement during breast cancer staging.
Source: European Radiology
Image Credit: iStock
References:
Lorenzon M, Minichetti P, Casotto L et al. (2026) Contrast-enhanced mammography versus breast MRI in the preoperative evaluation of the nipple-areola complex: data from a real-world setting. Eur Radiol: In Press.
