Dongfeng He1, Daqing Ma, Erhu Jin. 1. Department of Radiology, the Affiliated Beijing Friendship Hospital of Capital Medical University, Beijing, P.R. China.
Abstract
PURPOSE: To identify the dynamic contrast-enhanced magnetic resonance (DCE-MR) imaging features that may predict the outcome of patients with breast cancer. METHODS: DCE-MR images from 87 patients newly diagnosed with primary breast cancer were reviewed. The kinetic parameters (including cold spot, hot spot, and heterogeneity parameters) were derived from the DCE-MRI data. These parameters were used to thoroughly reflect the tumor status. The association of dynamic MR features (including kinetic and morphological features) with established prognostic indicators was evaluated. RESULTS: Malignant tumors with poor histomorphological indicators showed higher values of hot spot parameters (maximal initial Slope [maxSlope(i)] and maximal Washout [maxWashout]), higher values of a heterogeneity parameter- initial slope ratio (Slope(i) ratio) and lower values of a cold spot parameter (minimal initial slope [minSlope(iM)]) than those with favorable prognostic indicators. The heterogeneous internal enhancement pattern and rim-like enhancement pattern were more frequently observed in patients with poor prognostic indicators. Moreover, binary logistic regression analysis showed that kinetic parameters Slope(i) ratio (p=0.021), minSlope(i) (p=0.024), internal homogeneity (p=0.001), and maxSlope(iM) (p<0.001) were independently and significantly associated with histological grade, lymph node status, tumor size, and Ki-67, respectively. CONCLUSION: Our results suggest that all hot spot, cold spot, and heterogeneity parameters may be useful to noninvasively identify highly aggressive breast carcinomas. More importantly, cold spot and heterogeneity parameters may serve as crucial indicators to predict the outcome of breast cancer.
PURPOSE: To identify the dynamic contrast-enhanced magnetic resonance (DCE-MR) imaging features that may predict the outcome of patients with breast cancer. METHODS:DCE-MR images from 87 patients newly diagnosed with primary breast cancer were reviewed. The kinetic parameters (including cold spot, hot spot, and heterogeneity parameters) were derived from the DCE-MRI data. These parameters were used to thoroughly reflect the tumor status. The association of dynamic MR features (including kinetic and morphological features) with established prognostic indicators was evaluated. RESULTS:Malignant tumors with poor histomorphological indicators showed higher values of hot spot parameters (maximal initial Slope [maxSlope(i)] and maximal Washout [maxWashout]), higher values of a heterogeneity parameter- initial slope ratio (Slope(i) ratio) and lower values of a cold spot parameter (minimal initial slope [minSlope(iM)]) than those with favorable prognostic indicators. The heterogeneous internal enhancement pattern and rim-like enhancement pattern were more frequently observed in patients with poor prognostic indicators. Moreover, binary logistic regression analysis showed that kinetic parameters Slope(i) ratio (p=0.021), minSlope(i) (p=0.024), internal homogeneity (p=0.001), and maxSlope(iM) (p<0.001) were independently and significantly associated with histological grade, lymph node status, tumor size, and Ki-67, respectively. CONCLUSION: Our results suggest that all hot spot, cold spot, and heterogeneity parameters may be useful to noninvasively identify highly aggressive breast carcinomas. More importantly, cold spot and heterogeneity parameters may serve as crucial indicators to predict the outcome of breast cancer.