两种模型动态对比增强磁共振成像在宫颈癌病理分级中的应用价值  

Dynamic contrast-enhanced magnetic resonance imaging in pathological grading of cervical cancer:Extended Tofts Linear model versus Extended Tofts model

作　　者：章俞 赵振华[1] 韦明珠[1] 杨立铭[1] 黄亚男 阮雅文[2] 王丽 ZHANG Yu;ZHAO Zhenhua;WEI Mingzhu;YANG Liming;HUANG Yanan;RUAN Yawen;WANG Li(Department of Radiology,Affiliated Shaoxing Hospital of Zhejiang University School of Medicine,Shaoxing 312000,China)

机构地区：[1]绍兴市人民医院(浙江大学绍兴医院)放射科,312000 [2]绍兴市人民医院(浙江大学绍兴医院)妇产科,312000 [3]绍兴市人民医院(浙江大学绍兴医院)病理科,312000

出　　处：《浙江医学》2021年第2期148-152,I0004,共6页Zhejiang Medical Journal

基　　金：浙江省医药卫生科研基金项目(2017KY663);浙江省医药卫生科技计划项目(2015KYB402);绍兴市公益性技术应用研究项目(2018C30086、2018C30119);绍兴市人民医院院内青年基金一般项目(2018YB23)。

摘　　要：目的探讨Extended Tofts Linear模型和Extended Tofts模型动态对比增强磁共振成像(DCE-MRI)在宫颈癌病理分级中的应用价值。方法回顾性分析2016年1月至2019年10月绍兴市人民医院就诊且手术病理证实为子宫颈癌的31例患者,术前行盆腔多期DCE-MRI检查,选择Extended Tofts Linear和Extended Tofts血流动力学模型分别计算宫颈癌的定量灌注参数[包括容量转运常数(Ktrans)、速率常数(Kep)、血管外细胞外间隙容积分数(Ve)、血管间隙容积分数(Vp)],按术后病理分级分为低、中、高分化宫颈癌组,对比分析3组宫颈癌灌注参数的差异;筛选出有统计学意义的参数,绘制鉴别不同分化程度宫颈癌的ROC曲线,对比两种模型得到的定量参数在宫颈癌病理分级中的诊断效能。结果在高、低分化组宫颈癌间,Extended Tofts Linear模型的Ktrans值[高分化(0.518±0.180)、低分化(1.032±0.408)min-1]和Extended Tofts模型[高分化(0.525±0.283)、低分化(1.487±0.991)min-1]比较差异有统计学意义(P<0.05),高、低分化宫颈癌组间Kep、Ve、Vp比较差异均无统计学意义(均P>0.05)。而两个模型低分化与中分化、中分化与高分化组之间的Ktrans、Kep、Ve、Vp比较差异均无统计学意义(均P>0.05)。Extended Tofts Linear模型中Ktrans鉴别高、低分化宫颈癌的AUC大于Extended Tofts模型中Ktrans的AUC。联合诊断:利用logistic回归分析,求出Extended Tofts Linear和Extended Tofts两种模型Ktrans联合诊断概率值,结果显示两个模型联合诊断AUC均大于单个模型的AUC。结论Extended Tofts Linear模型和Extended Tofts模型DCE-MRI得到的Ktrans对高、低分化宫颈癌的病理分级具有一定价值,且前者的整体效能优于后者,同时两个模型联合用于宫颈癌病理分级的鉴别价值优于单个模型灌注参数,联合Ktrans鉴别诊断灵敏度、特异度最高。Objective To evaluate the application of dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI)with Extended Tofts Linear(ETL)model and extended tofts(ET)model in pathological grading of cervical cancer.Methods Thirty one patients with cervical cancer confirmed by pathology who underwent preoperative DCE-MRI from January 2016 to December 2019 in Shaoxing People's Hospital were enrolled in this retrospective study.We prospectively measured and analyzed quantitative perfusion parameters transfer constant(Ktrans),efflux rate constant(Kep),extravascular extracellular space volume ratio(Ve),blood plasma volume ratio(Vp),using signal-input two-compartment tracer kinetic models(ETL model and ET model).According to the postoperative pathological grade,the patients were divided into low,moderately and highly differentiated cervical cancer groups,and the differences of perfusion parameters among the three groups were compared and analyzed.The ROC curve was used to distinguish different degrees of differentiation of cervical cancer,and the diagnostic efficacy of the quantitative parameters obtained from the ETL and ET models in the pathological grading of cervical cancer was compared.Results There were significant differences in the Ktrans value in the ETL model(0.518±0.180 vs 1.032±0.408 min-1)and the Ktrans value in the ET model(0.525±0.283 vs 1.487±0.991 min-1)between well differentiated and poorly differentiated cervical cancer groups(P<0.05).There was no significant difference in Kep,Ve and VP between the two groups.There was no significant difference in quantitative perfusion value(Ktrans/Kep/Ve/Vp)between low differentiation group and moderate differentiation group,moderate differentiation group and high differentiation group(P>0.05).The area under the ROC curve of Ktrans obtained from ETL model(0.869±0.072)for differentiating well-differentiated cervical cancer from poorly differentiated cervical cancer was larger than that of Ktrans from ET model(0.843±0.077).Logistic regression analysis showed that th

关 键 词：宫颈癌 病理 动态对比增强磁共振成像 模型 

分 类 号：R737.33[医药卫生—肿瘤] R445.2[医药卫生—临床医学]