MR增强参数预测乳腺癌新辅助化疗疗效的前瞻性研究

摘要

Objective:To investigate the predictive value of the dynamic contrast enhanced MR parameters on neoadrnjuvant chemotherapy (NAC) of breast cancer. Methods: Ninety patients with invasive breast cancer underwent 3.0T MRrnscans after NAC but before the surgery. Dynamic parameters of time signal intensity curve (TIC) were recorded for regionrnof interest (ROI) ,which was measured by the high value method and the freehand method. Dynamic parameters measuredrnafter NAC were compared between the major histological response (MHR) group and non major histological responsern(NMHR) group using t test. Receiver operating characteristic (ROC) curve was used to define the optimal cutoff for thernprediction of MHR for different dynamic parameters of MR. Results:For different MP groups of breast cancer after NAC,rnsignificant differences were found for the signal enhancement value at second phase (SI2% ) ,the maximum enhancement valrnue of TIC (Slpeak% ) , and the maximum steepest slope (Smax) by both ROI measurement methods (P<0. 001 ). Parametersrnof the MHR group were lower than that of NMHR group. According to the area under ROC curve,parameters measured byrnROI high value method showed a better capability than that of the freehand method to predict the MP classification afterrnNAC of breast cancer. The optimal cutoff values of SI2% ,SIpeak%,and Smax was 83% ,229.45% and 404. 995 for ROI highrnvalue method,respectively. Conclusion: For breast cancer after NAC, dynamic parameters provided by both ROI measurernment methods could be used to predict the effect of NAC. The ROI high value method showed a higher capability of predicrntion than the freehand method.%目的:探讨乳腺癌新辅助化疗(NAC)后MR多期动态增强参数用于预测NAC疗效的价值.方法:应用3.0T MR对90例经NAC治疗后的浸润性乳腺癌患者行乳腺扫描.采用病灶兴趣区(ROI)高值法和手绘ROI法绘制病灶时间-信号强度曲线(TIC),记录TIC各相关参数值.采用t检验比较组织学显著反应(MHR)组和组织学非显著反应(NMHR)组NAC后乳腺癌病灶各参数的差异.以术后病理MHR者为阳性,绘制受试者工作特征(ROC)曲线,探讨MR多期动态增强参数预测MHR的最佳临界值.结果:乳腺癌NAC后不同MP术后评级组,两种ROI测量方法测得的病灶第二期强化程度(SI2%)、峰值强化程度(SIpeak%)、最大线性斜率(Smax)参数均有显著差异,MHR组参数值低于NMHR组(P<0.001).由ROC曲线下面积可得ROI高值法对乳腺癌NAC后手术病理分级的预测能力略优于ROI手绘法.ROI高值法预测MHR的最佳临界值分别为SI2%=83%、SIpeak%=229.45%、Smax=404.995.结论:NAC后乳腺癌病灶采用ROI高值法和手绘法测得的MR多期动态增强参数,均可预测新辅助化疗疗效,ROI高值法优于ROI手绘法.