Differences in tumour heterogeneity based on dynamic contrast-enhanced MRI between tumour and peritumoural stroma for predicting Ki-67 status of invasive ductal carcinoma. Issue 6 (June 2021)

• Record Type: Journal Article
• Authors: He, X.; Zhou, J.; Ye, S.; Cheng, Q.; Miao, H.; Xu, N.; Li, J.; Pan, Z.; Cheng, J.; Wang, M.
• Is Part Of: Clinical radiology. Volume 76:Issue 6(2021)
• Page Start: 470.e13
• Page End: 470.e22
• Publication Date: 2021-06
• Subjects: Medical radiology -- Periodicals; Radiotherapy -- Periodicals; Radiotherapy -- Periodicals; Radiology -- Periodicals; Societies, Medical -- Periodicals; Medical radiology; Radiotherapy; Electronic journals; Periodicals; 616.0757
• DOI: 10.1016/j.crad.2020.12.008
• Languages: English
• ISSNs: 0009-9260

Abstract

Abstract : AIM: To evaluate and compare the heterogeneity of intratumour and peritumour areas in the prediction of Ki-67 of invasive ductal carcinoma (IDC) and the predictive accuracy of different contrast frames based on dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI). MATERIALS AND METHODS: This study included 88 patients with histologically confirmed IDC with 57 patients with high Ki-67 status and 31 patients with low Ki-67 status. All patients underwent DCE-MRI before surgery. A grey-level co-occurrence matrix (GLCM) was performed on slice-matched images from six frames by drawing the region of the interest (ROI) on the inner and outer regions of the tumours. The correlations between texture characteristics and Ki-67 status of lesions were analysed, using the Mann–Whitney test and receiver operating characteristic curve analysis. RESULTS: In the high-Ki-67 group, the entropy was significantly higher than that of the low-Ki-67 group (p<0.001). The entropy obtained, based on the tumour boundary as a band-like area inside and outside at the first post-contrast series, revealed the highest receiver operating characteristic (AUC = 0.765). In the multivariate analysis, a higher entropy value (>4.305; p<0.001) remained independently associated with a high-Ki-67 status after adjustment for menopausal status, tumour size, histologic grade, oestrogen receptor (ER) status, and progesterone receptor (PR) status. The other parameters did not show significant differences between the high- and low-Ki-67 groups. CONCLUSION: Heterogeneity analysis based on DCE-MRI could discriminate between high- and low-Ki-67 status. Texture characteristics from the band-like region inside and outside the tumour boundary could predict the Ki-67 status and showed higher accuracy. Highlights: Heterogeneity analysis could discriminate high- and low-Ki-67 status based on DCE-MRI. The tumor showed better performance in predicting Ki-67 status. The contrast frames of DCE-MRI were independent of the prediction accuracy.