The Accuracy of Breast MR Imaging for Measuring the Size of a Breast Cancer: Analysis of the Histopathologic Factors

Abstract

BackgroundThe objective of the study was to compare the accuracy of different magnetic resonance (MR) sequences for measuring tumor size and to evaluate whether the imaging and histopathologic features affect the accuracy of the tumor size estimation on the MR sequence. Patients and MethodsEight hundred women were included. The maximum diameter of the tumor was measured on T2-weighted (T2W) sequences, early-subtracted dynamic contrast-enhanced (DCE) T1-weighted (T1W) sequences, and maximal intensity projection (MIP) reconstructions. Agreement between the MR imaging and pathology-determined size were analyzed. Using the best MR sequence to measure the tumor size, the relationship between the accuracy and the imaging and histopathologic features were evaluated. ResultsTumor measurement showed a good agreement with the pathology-determined size, and with the best results using MIP (k = 0.805) compared with the early-subtracted DCE T1W sequence (k = 0.802) and the T2W sequence (k = 0.779). On MIP, the tumors of patients with minimal or mild background parenchymal enhancement, a mass, invasive ductal carcinoma (IDC), pathology-determined size < 2 cm, positive estrogen receptor, negative HER2, luminal A type, nuclear and histologic grade 1, negative nodal status, negative lymphovascular invasion, and negative extensive intraductal component were significantly more accurately estimated. The independent factors associated with the accuracy of tumor measurement were a mass, IDC, and the pathology-determined size < 2 cm. ConclusionOur study showed that tumor size is most accurately measured on MIP, especially for IDC appearing as a mass on MR imaging and with the pathology-determined size < 2 cm.