To assess the association between multiparametric MRI features of primary tumors and axillary lymph node tumor burden in women with invasive breast cancer. In this retrospective study, women diagnosed with invasive breast cancer who underwent 3T multiparametric MRI, including diffusion-weighted imaging (DWI) from 2019-2020, were evaluated. Two radiologists reviewed T2-weighted images (T2WI) for peritumoral edema and intratumoral necrosis, and measured apparent diffusion coefficient (ADC) values by manually placing regions of interest within breast tumors. We also analyzed quantitative kinetic features of breast cancer using computer-aided diagnosis (CAD) and clinical-pathologic characteristics. Uni- and multivariable logistic regression analyses were conducted to identify predictors of a high axillary nodal burden (≥3 positive nodes). In total, 301 women (mean age, 54.13 years) were evaluated. Forty-three (14.3%) had a high axillary nodal burden by surgical pathology. Multivariate analysis revealed that factors significantly associated with high axillary nodal burden included peritumoral edema (OR:7.970; p<0.001), lower tumor ADCmax (≤ 1.098 × 10-3 mm2/s) (OR:6.978; p<0.001), larger tumor size (>2cm) (OR:2.986; p=0.046), lobular histology (OR:12.620; p<0.001), and the presence of lymphovascular invasion (OR:3.622; p=0.003). CAD-derived kinetic features did not show an association with axillary nodal burden. In subgroup analysis of 238 patients with early clinically node-negative breast cancer, both peritumoral edema (OR:7.831; p=0.002) and lower tumor ADCmax (≤ 1.098 × 10-3 mm2/s) (OR:8.002; p=0.002) remained significant predictors. Our results suggest that peritumoral edema as viewed in T2WI and the ADCmax value of breast cancer in DWI are valuable for predicting axillary nodal burden in women with invasive breast cancer. Multiparametric MRI features of a primary tumor are useful for predicting axillary nodal burden in patients with invasive breast cancer.
Read full abstract