Agarwal Khushbu, Sharma Uma, Mathur Sandeep, Seenu Vurthaluru, Parshad Rajinder, Jagannathan Naranamangalam R
Department of NMR & MRI Facility, All India Institute of Medical Sciences, New Delhi, Delhi 110029, India.
Department of Pathology, All India Institute of Medical Sciences, New Delhi, Delhi 110029, India.
Magn Reson Imaging. 2018 Jun;49:116-122. doi: 10.1016/j.mri.2018.02.004. Epub 2018 Feb 14.
To evaluate the utility of fat fraction (FF) for the differentiation of different breast tissues and in various breast tumor subtypes using in vivo proton (H) magnetic resonance spectroscopy (MRS).
H MRS was performed on 68 malignant, 35 benign, and 30 healthy volunteers at 1.5 T. Malignant breast tissues of patients were characterized into different subtypes based on the differences in the expression of hormone receptors and the FF was calculated. Further, the sensitivity and specificity of FF to differentiate malignant from benign and from normal breast tissues of healthy volunteers was determined using receiver operator curve (ROC) analysis.
A significantly lower FF of malignant (median 0.12; range 0.01-0.70) compared to benign lesions (median 0.28; range 0.02-0.71) and normal breast tissue of healthy volunteers (median 0.39; range 0.06-0.76) was observed. No significant difference in FF was seen between benign lesions and normal breast tissues of healthy volunteers. Sensitivity and specificity of 75% and 68.6%, respectively was obtained to differentiate malignant from benign lesions. For the differentiation of malignant from healthy breast tissues, 76% sensitivity and 74.5% specificity was achieved. Higher FF was seen in patients with ER-/PR- status as compared to ER+/PR+ patients. Similarly, FF of HER2neu+ tumors were significantly higher than in HER2neu- breast tumors.
The results showed the potential of in vivo H MRS in providing insight into the changes in the fat content of different types of breast tissues and in various breast tumor subtypes.
利用体内质子(H)磁共振波谱(MRS)评估脂肪分数(FF)在鉴别不同乳腺组织及各种乳腺肿瘤亚型中的效用。
对68例恶性、35例良性和30例健康志愿者在1.5T条件下进行H MRS检查。根据激素受体表达差异将患者的恶性乳腺组织分为不同亚型,并计算FF。此外,使用受试者工作特征曲线(ROC)分析确定FF区分恶性与良性以及健康志愿者正常乳腺组织的敏感性和特异性。
观察到恶性病变(中位数0.12;范围0.01 - 0.70)的FF显著低于良性病变(中位数0.28;范围0.02 - 0.71)和健康志愿者的正常乳腺组织(中位数0.39;范围0.06 - 0.76)。良性病变与健康志愿者的正常乳腺组织之间的FF无显著差异。区分恶性与良性病变的敏感性和特异性分别为75%和68.6%。区分恶性与健康乳腺组织时,敏感性为76%,特异性为74.5%。与ER+/PR+患者相比,ER-/PR-状态患者的FF更高。同样,HER2neu+肿瘤的FF显著高于HER2neu-乳腺肿瘤。
结果表明体内H MRS在洞察不同类型乳腺组织及各种乳腺肿瘤亚型脂肪含量变化方面具有潜力。