Agreement between MRI and pathologic analyses for determination of tumor size and correlation with immunohistochemical factors of invasive breast carcinoma

Abstract

Background: There may be discordance between tumor size determined by magnetic resonance imaging (MRI) and that observed during pathologic analyses. Purpose: To evaluate MRI-pathology concordance of tumor size in patients with invasive breast carcinoma. Material and Methods: Data from 307 invasive breast carcinomas were analyzed retrospectively. Preoperative breast MRI was reviewed for size, lesion type, morphology, and dynamic contrast-enhanced tumor kinetics. MRI tumor size was compared with tumor size measurements from the pathologic analysis. Concordance was defined as a difference in diameter of <= 0.5 cm. MRI-pathology concordance was compared according to clinical and histopathologic features. Results: The mean tumor size on MRI was 2.48 +/- 1.41 cm. Tumor measurements determined by MRI were not significantly different from those recorded in the pathologic reports (2.56 +/- 1.61 cm, P = 0.199). MRI-pathology concordance was found in 229/307 (74.6%) cases; the size was overestimated in 36 (11.7%) tumors and underestimated in 42 (13.7%). On univariate analysis, MRI-pathology discordance was associated with larger tumor size (P < 0.001), estrogen receptor (ER) negativity (P = 0.006), and lymphovascular invasion (P = 0.003). Human epidermal growth factor receptor 2 positive molecular subtype showed worse correlation between the tumor size measured by MRI and pathology compared with luminal A and luminal B subtypes (P = 0.008 and 0.007). On multivariate analysis, tumor size and ER status significantly influenced MRI-pathology concordance (P < 0.05). Conclusion: ER negativity and larger tumor size were strongly associated with MRI-pathology discordance in invasive breast carcinomas. Awareness of these factors might improve surgical planning.