MRI and intraoperative pathology to predict nipple-areola complex (NAC) involvement in patients undergoing NAC-sparing mastectomy

Abstract Background Nipple–areola sparing mastectomy (NSM) with immediate implant reconstruction is an option for patients with non-locally advanced breast cancer. The prediction of occult tumour involvement of the nipple–areola complex (NAC) may help select candidates to NSM. Patients and methods We prospectively recorded clinical and pathological data, magnetic resonance imaging (MRI) results and intraoperative pathological assessments of the subareolar (SD) and proximal nipple ducts (ND) of 112 consecutive breast cancer patients scheduled for NSM. All parameters were correlated with final pathological NAC assessment by univariate and multivariate analysis. Results Thirty-one patients (27.7%) had tumour involvement of the NAC. At univariate analysis, age ( p  = 0.001), post-menopausal status (0.003), tumour central location ( p  = 0.03), tumour–NAC distance measured by MRI ( p  = 0.000) and intraoperative pathologic assessment (SD + ND) ( p  = 0.000) were significantly correlated with NAC involvement. At multivariate analysis, only MRI tumour–NAC distance ( p  = 0.008) and menopausal status ( p  = 0.039) among all preoperative variables retained statistical significance. The sensitivity and specificity of MRI tumour–NAC distance were 32.2% and 88.6% and those of intraoperative pathologic assessment were 46.7% and 100%, respectively. Sensitivity, specificity and accuracy of the double assessment (MRI plus intraoperative pathology) were 50.0%, 96.2% and 84.1%, respectively. Conclusion Intraoperative pathologic assessment and tumour–NAC distance measured by MRI are the most important predictors of occult NAC involvement in breast cancer patients. A negative pathological assessment and a tumour–NAC distance ⩾ 5 mm allow optimal discrimination between NAC positive and NAC negative cases and may serve as a guide for the optimal planning of oncological and reconstructive surgery.