Sun T T, Liu W H, Zhang Y Q, Li L H, Wang R, Ye Y Y
Southeast University, Nanjing 210009, China.
Zhonghua Yi Xue Za Zhi. 2017 Aug 1;97(29):2266-2270. doi: 10.3760/cma.j.issn.0376-2491.2017.29.007.
To explore the differential between the value of dynamic contrast-enhanced MRI quantitative pharmacokinetic parameters and relative pharmacokinetic quantitative parameters in breast lesions. Retrospective analysis of 255 patients(262 breast lesions) who was obtained by clinical palpation , ultrasound or full-field digital mammography , and then all lessions were pathologically confirmed in Zhongda Hospital, Southeast University from May 2012 to May 2016. A 3.0 T MRI scanner was used to obtain the quantitative MR pharmacokinetic parameters: volume transfer constant (K(trans)), exchange rate constant (k(ep))and extravascular extracellular volume fraction (V(e)). And measured the quantitative pharmacokinetic parameters of normal glands tissues which on the same side of the same level of the lesions; and then calculated the value of relative pharmacokinetic parameters: rK(rans)、rk(ep) and rV(e).To explore the diagnostic value of two pharmacokinetic parameters in differential diagnosis of benign and malignant breast lesions using receiver operating curves and model of logistic regression. (1)There were significant differences between benign lesions and malignant lesions in K(trans) and k(ep) (=15.489, 15.022, respectively, <0.05), there were no significant differences between benign lesions and malignant lesions in V(e)(=-2.346, >0.05). The areas under the ROC curve(AUC)of K(trans), k(ep) and V(e) between malignant and benign lesions were 0.933, 0.948 and 0.387, the sensitivity of K(trans), k(ep) and V(e) were 77.1%, 85.0%, 51.0% , and the specificity of K(trans), k(ep) and V(e) were 96.3%, 93.6%, 60.8% for the differential diagnosis of breast lesions if taken the maximum Youden's index as cut-off. (2)There were significant differences between benign lesions and malignant lesions in rK(trans), rk(ep) and rV(e) (=14.177, 11.726, 2.477, respectively, <0.05). The AUC of rK(trans), rk(ep) and rV(e) between malignant and benign lesions were 0.963, 0.903 and 0.575, the sensitivity of rK(trans), rk(ep) and rV(e) were 85.6%, 71.9%, 52.9% , and the specificity of rK(trans), rk(ep) and rV(e) were 94.5%, 92.7%, 60.6% for the differential diagnosis of breast lesions.(3)There was no significant difference in the area under the ROC curve between the predictive probability of quantitative pharmacokinetic parameters and the prediction probability of relative quantitative pharmacokinetic parameters(=0.867, =0.195). There was no significant difference between the quantitative parameter values (K(trans,) k(ep)) and the relative quantitative parameter values (rK(trans,) rk(ep)) in diagnosis of breast lesions, which were important parameters in differential diagnosis of benign and malignant breast lesions.
探讨动态对比增强磁共振成像(MRI)定量药代动力学参数与相对药代动力学定量参数在乳腺病变中的价值差异。回顾性分析2012年5月至2016年5月在东南大学附属中大医院经临床触诊、超声或全数字化乳腺摄影发现并经病理证实的255例患者(262个乳腺病变)。采用3.0T MRI扫描仪获取定量MR药代动力学参数:容积转运常数(K(trans))、交换率常数(k(ep))和血管外细胞外容积分数(V(e))。并测量病变同侧同水平正常腺体组织的定量药代动力学参数;然后计算相对药代动力学参数值:rK(trans)、rk(ep)和rV(e)。采用受试者工作特征曲线和逻辑回归模型探讨两种药代动力学参数在乳腺良恶性病变鉴别诊断中的诊断价值。(1)K(trans)和k(ep)在良性病变与恶性病变之间存在显著差异(分别为=15.489、15.022,<0.05),V(e)在良性病变与恶性病变之间无显著差异(=-2.346,>0.05)。恶性与良性病变之间K(trans)、k(ep)和V(e)的ROC曲线下面积(AUC)分别为0.933、0.948和0.387,以最大约登指数为截断值时,K(trans)、k(ep)和V(e)对乳腺病变鉴别诊断的敏感性分别为77.1%、85.0%、51.0%,特异性分别为96.3%、93.6%、60.8%。(2)rK(trans)、rk(ep)和rV(e)在良性病变与恶性病变之间存在显著差异(分别为=14.177、11.726、2.477,<0.05)。恶性与良性病变之间rK(trans)、rk(ep)和rV(e)的AUC分别为0.963、0.�03和0.575,rK(trans)、rk(ep)和rV(e)对乳腺病变鉴别诊断的敏感性分别为85.6%、71.9%、52.9%,特异性分别为94.5%、92.7%、60.6%。(3)定量药代动力学参数的预测概率与相对定量药代动力学参数的预测概率之间的ROC曲线下面积无显著差异(=0.867,=0.195)。定量参数值(K(trans, k(ep))与相对定量参数值(rK(trans, rk(ep))在乳腺病变诊断中无显著差异,它们是乳腺良恶性病变鉴别诊断的重要参数。