Comparison of Concordance of Tumor Size Measured using Ultrasonography, Magnetic Resonance Imaging, and Pathology in Breast Cancer

Abstract

Preoperative evaluations of the tumor size in breast cancer are essential for obtaining negative margins and performing oncoplastic surgery. Here, the tumor sizes measured using ultrasonography (USG), magnetic resonance imaging (MRI), and pathology were compared. Patients with a single mass observed using USG who underwent surgery between March 2016 and February 2019 were reviewed. Patients with multiple lesions, positive surgical margins, or had undergone neoadjuvant chemotherapy were excluded. The largest tumor diameter was measured during each examination. The intraclass correlation coefficient (ICC) was used for statistical analysis to compare concordance in tumor size. Nine hundred and ninety-six patients were analyzed and divided into subgroups according to their menopausal status, MRI features, pathological subtypes, and intrinsic subtypes. In all patients, MRI has higher concordance with the pathological findings than USG. In subgroup analysis, there was a significant difference between the ICC of USG and MRI for the premenopausal, non-mass enhancement (NME), and lobular types. According to the intrinsic subtype analysis, in luminal A and human epidermal growth factor receptor 2 (HER-2) breast cancer, MRI revealed higher ICC values than USG. In triple-negative breast cancer (TNBC), USG showed a higher ICC, but the difference between USG and MRI was small. MRI may be more helpful in predicting the tumor size for patients who are premenopausal, have NME, and have lobular cancer. On the other hand, USG is considered more helpful in luminal A and HER-2. Concordance differs depending on the clinicopathological characteristics and should be considered in the surgical plan.

Keywords

• breast neoplasms
• ultrasonography
• magnetic resonance imaging
• pathology