酰胺质子转移加权成像联合DKI预测乳腺癌新辅助化疗早期病理完全反应  

Early prediction of pathologic complete response to neoadjuvant chemotherapy in breast cancer by amide proton transfer weighted imaging combined with diffusion kurtosis imaging

作　　者：陈科威 潘君龙[1,3] 余成新[1,3] 罗雨晴 徐亚卡[3] 杨婷 CHEN Ke-wei;PAN Jun-long;YU Chen-xin(The First College of Clinical Medical Science,Three Gorges University,Hubei 443000,China)

机构地区：[1]三峡大学第一临床医学院放射科,湖北宜昌443000 [2]宁波大学附属妇女儿童医院放射科,浙江宁波315000 [3]宜昌市中心人民医院放射科,湖北宜昌443000 [4]复旦大学附属金山医院放射科,上海201508

出　　处：《放射学实践》2024年第8期1059-1066,共8页Radiologic Practice

基　　金：中华国际医学交流基金会SKY影像科研基金(Z-2014-07-2101)。

摘　　要：目的:探讨酰胺质子转移(APT)加权成像联合扩散峰度成像(DKI)对乳腺癌新辅助化疗(NAC)早期病理完全反应的预测效能。方法:本研究中纳入2022年7月-2023年8月在本院确诊为乳腺癌后接受新辅助化疗且在治疗前行DKI和APT加权成像检查的72例患者。搜集并记录每例患者的临床和病理信息。测量病灶的平均峰度系数(MK)、平均扩散系数(MD)和非对称磁化转移率(MTR_(asym))。根据Miller Payne分级系统评估患者对NAC的反应,定义Miller Payne 5级为病理完全反应(pCR),Miller Payne 1~4级为非pCR(non-pCR)。采用独立样本t检验及卡方检验比较pCR组与non-pCR组之间临床病理资料(包括一般临床资料及免疫组化结果)及影像学定量参数间的差异,随后对P值小于0.05的变量采用二元多因素Logistic回归分析构建联合模型,并使用受试者工作特性曲线(ROC)分析各参数单独或联合的诊断效能。结果:pCR组肿瘤最大径(MRI图像测量)明显小于non-pCR组(P=0.006)MD值高于non-pCR组(P=0.008),MTR_(asym)(3.5 ppm)值低于non-pCR组(P=0.006);联合模型的AUC为0.799,与各项单一参数AUC的差异均有统计学意义(Z=2.059~2.168,P=0.0302~0.0395)。结论:DKI和APT衍生参数及肿瘤最大径对早期预测乳腺癌NAC病理完全反应具有重要价值,且三者联合的预测效能优于单一因素。Objective:To investigate the predictive efficacy of amide proton transfer(APT)-weighted imaging combined with diffusion kurtosis imaging(DKI)for early pathological complete response(pCR)to neoadjuvant chemotherapy in breast cancer.Methods:Seventy-two patients who received neoadjuvant chemotherapy after diagnosis of breast cancer at our hospital from July 2022 to August 2023 and underwent DKI and APT-weighted imaging before treatment were included in this stu-dy.Clinical and pathological information was collected and recorded for each patient.Mean kurtosis coefficient(MK),mean diffusion coefficient(MD)and asymmetric magnetisation transfer rate(MTR_(asym))of the lesions were measured.Patient response to NAC was assessed according to the Mil-ler Payne grading system,defining Miller Payne grade 5 as pCR and Miller Payne grades 1~4 as non-pCR(non-pCR).Independent samples t-test and Chi-square test were used to compare the differences in clinicopathological data(including general clinical information and immunohistochemical results)and quantitative imaging parameters between the pCR and non-pCR groups,and a joint model was constructed by binary logistic multifactorial analysis of factors with a P-value less than 0.05,and the parameters were analyzed using the receiver operating characteristic(ROC)to determine the diagnostic efficacy of the parameters,individually or in combination diagnostic efficacy.Results:The maximum diameter of the tumor(measured in MRI images)in the pCR group was significantly smaller than that in the non-pCR group(P=0.006),the MD value was higher than that in the non-pCR group(P=0.008),and the value of MTR_(asym)(3.5ppm)was lower than that in the non-pCR group(P=0.006);the AUC of the combined model was 0.799,which was significantly different from that of single parameter(Z=2.059~2.168,P=0.0302~0.0395).Conclusion:DKI-and APT-derived parameters and maximum tumour diameter are valuable for early prediction of pathological complete response to NAC in breast cancer,and the predictive efficacy of the c

关 键 词：乳腺癌 新辅助化疗 病理完全反应 酰胺质子转移加权成像 扩散峰度成像 

分 类 号：R445.2[医药卫生—影像医学与核医学] R737.9[医药卫生—诊断学]