Povezanost morfoloških i kinetičkih osobitosti duktalnoga invazivnoga karcinoma dojke dobivenih magnetnom rezonancijom s patohistološkim prognostičkim pokazateljima Correlation of morphological and kinetic features at magnetic resonance imaging and pathohistological prognostic factors in invasive ductal breast cancer

摘要

Objective. The aim of this study was to correlate magnetic resnonace imaging (MRI) featuresudof invasive ductal carcinomas (IDC) with pathohistological prognostic factors, as well as withudmolecular subtypes of breast cancer. Such an association, if present, could have significantudtranslational implications for early identification of aggressive types of breast cancer, whereudrapid identification can expedite tumor treatment.udMethods. 114 women with IDC who underwent breast MRI within one month prior toudsurgery were included in this retrospective study. MRI features were analyzed and thenudinterpreted with a Göttingen score (GS) that included morphological (shape, margins andudpattern of enhancement) and kinetic characteristics (initial signal increase and post-initialudbehaviour of the time-signal intensity curve). Histological specimens were analyzed for tumorudsize, axillary lymph node status, histological grade, estrogene receptors (ER), progesteroneudreceptors (PR), HER2, and Ki-67.udResults. By multivariate analysis, a smooth margin was a significant, independent predictorudof a larger tumor size (p=0.041), lymph node invasion (p=0,013), and lower expression of ERud(p=0.022). High GS was significant, independent predictor of a higher histological gradeud(p=0.022) while round or oval shape of lesion was independent predictor of a higher PRudexpression (p=0,027).udConclusion. A smooth margin of breast cancer on breast MRI was able to predict positiveudaxillary lymph nodes, larger tumor size and lower expression of ER. Except for a higherudhistological grade, GS was not able to predict other unfavourable prognostic factors, mainlyuddue to the fact that smooth margins were assigned less points than spiculated margins.