Prediction of recurrence risk of estrogen receptor-positive and human epidermal growth factor receptor-2 negative breast cancer using a multi-parameter regression model based on diffusion kurtosis imaging / 中华放射学杂志

ABSTRACT

Objective:To explore the predictive value of a regression model based on diffusion kurtosis imaging (DKI) parameters for prediction of the recurrence risk in patients with estrogen receptor (ER)-positive, human epidermal growth factor receptor 2 (HER-2)-negative early invasive breast cancer.Methods:A retrospective cross-sectional study was designed. The clinicopathological (age, histological grade, Ki-67 level, etc.) and imaging data of 50 patients (50 lesions) with ER-positive, HER-2 negative early invasive breast cancer who underwent treatment at Wuxi People′s Hospital from January 2016 to December 2018 were retrospectively analyzed. All patients were female, aged 29 to 81 years, and underwent pre-operation conventional MRI and DKI examinations. The volume of breast fibroglandular tissue (FGT), background parenchymal enhancement (BPE), and internal enhancement features were recorded; the peak enhancement (PH), peak enhancement rate, time to peak, mean kurtosis (MK), and mean diffusivity (MD) were calculated. Based on the 21-gene recurrence risk scores, patients were divided into low recurrence risk group and medium-high recurrence risk group. Independent sample t test, Mann-Whitney U test, χ2 test were used to compare the differences of various indicators between the two groups. Two logistic models were constructed with age, PH, MD, and MK as independent variables (Pre1), and with Ki-67, age, PH, MD, and MK as independent variables (Pre2), respectively. The efficacy of the models in predicting low recurrence risk in patients was assessed using receiver operating characteristic curve and area under the curve (AUC). Results:There were 25 cases in the low recurrence risk group and 25 cases in the medium-high recurrence risk group. The differences in age, FGT, PH, MD, MK, and Ki-67 between the low recurrence risk group and the medium-high recurrence risk group were statistically significant (all P<0.05), while other indexes showed no statistically significant differences (all P>0.05). The AUC of Pre1 in predicting low recurrence risk of ER-positive, HER-2 negative early invasive breast cancer was 0.87, with a sensitivity of 0.76 and specificity of 0.88. The AUC of Pre2 for predicting the low recurrence risk of ER-positive, HER-2 negative early invasive breast cancer was 0.92, with a sensitivity of 0.84, and specificity of 0.92. Conclusions:A multi-parameter model based on DKI can effectively predict the recurrence risk of ER-positive and HER-2 negative breast cancer. The model with combination of Ki-67 can further improve the predictive efficacy, and help effectively identify patients at low recurrence risk.