Abstract P4-02-04: Magnetic resonance imaging to predict nipple involvement in breast cancer patients

Abstract

Abstract Introduction. The implementation of Nipple-Sparing Mastectomy (NSM) as a treatment option for selected cases of breast cancer has risen great interest among breast surgeons. The preservation of the nipple-areola complex (NAC) can lead to extremely favorable psychological effects in breast cancer patients treated with this type of procedure. However, to ensure the oncologic safety of this technique it is of utmost importance to evaluate the likelihood of NAC involvement pre-operatively. In this study we evaluate the contribution of Breast Magnetic Resonance Imaging (MRI) in predicting the involvement in the NAC in breast cancer patients. Materials and Methods: We studied 170 mastectomy specimens from 165 breast cancer patients (five patients had bilateral disease) affected by Ductal Carcinoma in situ (DCIS)(n=19) or Invasive Ductal Carcinoma (n=151), stages I, II or IIIA. Every patient was pre-operatively studied using a 1.5 Tesla, 4-channel in vivo dedicated surface breast coil MRI. The parameters we investigated were: type of index lesion enhancement pattern (nodular or non-nodular), size of the index lesion, enhancement between the index lesion and the NAC, enhancement of the nipple, thickening of the areola, nipple retraction and size of the nipple in comparison with the contra-lateral nipple. The retro-areolar area and papilla were evaluated in histological sections of 4μm to identify DCIS and Invasive Ductal or Lobular carcinomas. One radiologist, blinded to the result of the histological evaluation of the papillae, performed the evaluation of the MRIs. Results. In univariate analysis, type of lesion enhancement in MRI, enhancement between index lesion and the papilla, distance between the index lesion and the papilla, enhancement of the papilla and nipple retraction had a statistically significant correlation with neoplastic involvement of the NAC (p<0.05). Using multivariate analysis, among the previously mentioned parameters, enhancement between the index lesion and the NAC, and nipple retraction remained as statistically significant predictors of nipple involvement in breast cancer patients (p < 0,001 e 0,010, respectively). The Negative Predictive Value of the combination of these two variables was 89.5%. According to this model that used the combination of those two variables, the probability of neoplastic involvement of the NAC was 73.9% in the presence of enhancement between the index lesion and the papilla combined with nipple retraction; 46% in the presence of enhancement between the index lesion and the papilla without nipple retraction in the MRI; 26.9% if there is only nipple retraction in the MRI; and 9.9% in the absence of these two characteristics. The sensibility of this model composed by those 2 variables to identify neoplastic involvement of the NAC was 29.7% (CI95%: 15.9% - 47%), specificity was 97.7% (CI95%: 93,5% - 99,5%), positive predictive value was 78,6% (CI95%: 49,2% - 95,3%) and negative predictive value was 83,3% (CI95%: 76,5% - 88,8%). Conclusion. We can conclude that the probability of the NAC being cancer-free is around 90% when there is no enhancement between the index lesion and the nipple, and there is no retraction of the nipple in the pre-operative MRI study of the breast. Citation Format: Piato JR, Chala LF, Alves-Jales RD, Dória MT, Mota BS, Messias AP, Goncalves R, Mano MS, Soares JM, de Barros N, Filassi JR, Baracat EC. Magnetic resonance imaging to predict nipple involvement in breast cancer patients. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P4-02-04.