分享:
分享到微信朋友圈
X
临床研究
酰胺质子转移成像与体素内不相干运动成像评估子宫内膜腺癌组织学分级的价值
孔雅晴 曲倩倩 明蕾 王哲 邓凯

Cite this article as: Kong YQ, Qu QQ, Ming L, et al. Value of amide proton transfer imaging and intravoxel incoherent motion imaging in estimating histologic grades of endometrial adenocarcinoma[J]. Chin J Magn Reson Imaging, 2022, 13(9): 46-52.本文引用格式:孔雅晴, 曲倩倩, 明蕾, 等. 酰胺质子转移成像与体素内不相干运动成像评估子宫内膜腺癌组织学分级的价值[J]. 磁共振成像, 2022, 13(9): 46-52. DOI:10.12015/issn.1674-8034.2022.09.009.


[摘要] 目的 探讨酰胺质子转移(amide proton transfer, APT)成像和体素内不相干运动(intravoxel incoherent motion, IVIM)成像在评估子宫内膜腺癌(endometrial adenocarcinoma, EA)组织学分级中的价值。材料与方法 39名病理证实EA的患者接受了盆腔磁共振检查(其中Ⅰ级16例,Ⅱ级14例,Ⅲ级9例),扫描序列包括APT、IVIM。测量并比较各级别之间APT值和IVIM各衍生参数值(f,D,D*)的差异,然后采用受试者工作特征(receiver operating characteristic, ROC)曲线和DeLong检验评估差异具有统计学意义的参数的鉴别效能。最后通过Spearman相关分析评估APT值、IVIM各衍生参数值和组织学分级两两之间的相关性。结果 Ⅱ级和Ⅲ级EA的APT值(3.17%±0.43%、3.45%±0.42%)显著高于Ⅰ级EA的APT值(2.66%±0.63%),差异具有统计学意义(P=0.012,P=0.001),Ⅱ级和Ⅲ级EA的f值(0.170%±0.049%、0.150%±0.011%)显著低于Ⅰ级EA的f值(0.220%±0.080%),差异具有统计学意义(P=0.027,P=0.010)。APT值、f值以及二者联合诊断Ⅰ级与Ⅱ级、Ⅰ级与Ⅲ级EA的曲线下面积(area under the curve, AUC)分别为0.78、0.70、0.83;0.86、0.81、0.92,差异无统计学意义。Spearman相关性分析结果显示APT值与EA组织学分级相关系数rs=0.578(95% CI:0.27,0.79;P<0.001),f值与EA组织学分级相关系数rs=-0.416(95% CI:-0.66,-0.11;P=0.008)。结论 APT值和f值与EA组织学分级相关,APT值和f值对EA病理级别的鉴别效能无显著差异,APT成像和IVIM成像有助于术前预测EA组织学分级。
[Abstract] Objective To investigate the effect of amide proton transfer (APT) imaging and intravoxel incoherent motion (IVIM) imaging in estimating histologic grades of endometrial adenocarcinoma (EA).Materials and Methods A total of 39 patients with pathologically confirmation of EA underwent pelvic magnetic resonance imaging (grade Ⅰ, n=16; grade Ⅱ, n=14; grade Ⅲ, n=9), including APT and IVIM sequences. APT values and IVIM-derived parameters (f, D, D*) of lesions were calculated respectively by two radiologists. We used a Shapiro-Wilk test to evaluate whether the APT values and IVIM-derived parameters were normally distributed. The intraclass correlation coefficient (ICC) was calculated to describe the correlations of the measurements of each parameter between two readers. The differences of APT values and IVIM-derived parameters were calculated and compared among three grades. Then the receiver operating characteristic (ROC) curves and DeLong test were used to evaluate the efficiency of the parameters with statistically differences. Spearman's correlation analysis was also used between APT values and histologic grades, between IVIM-derived parameters and histologic grades, and between APT values and IVIM-derived parameters.Results All parameters in our research were normally distributed. The ICC of APT values and IVIM-derived parameters between the two readers showed excellent interobserver agreements (APT: 0.955; f: 0.958; D: 0.964; D*: 0.825). The APT values for grade Ⅱ and grade Ⅲ (3.17%±0.43%, 3.45%±0.42%) were significantly higher than grade Ⅰ (2.66%±0.63%), with statistically significant differences (P=0.012, P=0.001). The f values for grade Ⅱ and grade Ⅲ (0.170%±0.049%, 0.150%±0.011%) were significantly lower than grade Ⅰ (0.220%±0.080%), with statistically significant differences (P=0.027, P=0.010). The ROC analysis showed the AUC (area under the curve) of the APT values, f values and the combined diagnosis of both were as follows: grade Ⅰ vs. grade Ⅱ: 0.78, 0.70, 0.83; grade Ⅰ vs. grade Ⅲ: 0.86, 0.81, 0.92. There were no significant differences in AUC comparison among APT values, f values and combined diagnosis of both. APT values showed a positive correlation with histologic grades (rs=0.578, P<0.001) while f values showed a negative correlation (rs=-0.416, P=0.008).Conclusions APT values and f values were correlated with the histologic grades of EA, and there was no significant difference of APT values and f values in differentiating the histologic grades of EA. APT and IVIM imaging may be helpful to predict histologic grades of EA.
[关键词] 酰胺质子转移成像;体素内不相干运动;子宫内膜腺癌;病理分级;功能磁共振成像
[Keywords] amide proton transfer imaging;intravoxel incoherent motion imaging;endometrial adenocarcinoma;histologic grades;functional magnetic resonance imaging

孔雅晴 1, 2   曲倩倩 1   明蕾 1   王哲 1   邓凯 1*  

1 山东第一医科大学第一附属医院(山东省千佛山医院)放射科,济南 250014

2 山东第一医科大学(山东省医学科学院),济南 271016

*邓凯,E-mail:289954749@qq.com

作者利益冲突声明:全体作者均声明无利益冲突。


基金项目: 山东省科技发展计划 2012YD18060
收稿日期:2022-05-19
接受日期:2022-08-30
中图分类号:R445.2  R737.33 
文献标识码:A
DOI: 10.12015/issn.1674-8034.2022.09.009
本文引用格式:孔雅晴, 曲倩倩, 明蕾, 等. 酰胺质子转移成像与体素内不相干运动成像评估子宫内膜腺癌组织学分级的价值[J]. 磁共振成像, 2022, 13(9): 46-52. DOI:10.12015/issn.1674-8034.2022.09.009

       子宫内膜癌(endometrial carcinoma, EC)是女性生殖系统常见的恶性肿瘤之一,以腺癌多见,严重威胁女性健康。根据2009年国际妇产科联盟的标准[1],将EC的病理分级分为三级:高分化癌G1,中分化癌G2,低分化癌G3。大约80%的EC为Ⅰ~Ⅱ级的子宫内膜腺癌(endometrial adenocarcinoma, EA),其余为Ⅲ级的EA和其他组织学类型的EC,如透明细胞癌或浆液性癌。EC的病理分级与患者预后直接相关[2],且肿瘤级别的高低直接影响临床治疗方式的选择,因此术前准确评估肿瘤的组织学分级对于临床治疗计划以及改善患者预后有着重要的意义[3]。MRI技术具有较高的软组织分辨率,已广泛应用于EC术前诊断。酰胺质子转移(amide proton transfer, APT)成像是一种新型的化学交换饱和转移(chemical exchange saturation transfer, CEST)MRI技术,该技术通过检测组织内源性蛋白质或多肽中的酰胺质子与水质子的交换速率来反映细胞内蛋白质浓度和pH值[4, 5],从而将MRI技术的应用范围扩展到分子水平。体素内不相干运动(intravoxel incoherent motion, IVIM)采用多b值,基于多种模型,减小了扩散加权成像(diffusion weighted imaging, DWI)单指数模型受到的肿瘤内部毛细血管微灌注的影响,能够定量评估毛细血管微灌注产生的扩散系数D*及灌注系数f,可以更加准确地反映水分子扩散程度,同时获得灌注相关信息,更加全面地反映肿瘤微观情况[6]。作为功能MRI技术,近年来APT成像和IVIM成像在EC的术前诊断中得到了初步应用[7, 8, 9, 10],但关于分析及比较二者在评估EA分级中的价值的应用鲜有报道。本研究目的是探讨APT成像和IVIM成像在评估EA术前组织学分级中的价值。

1 材料与方法

1.1 研究对象

       这项前瞻性研究得到了我院伦理委员会的批准,批准文号:[2021]伦审字(S1036)号,获得了每位患者的知情同意并已签署知情同意书。自2021年4月至2022年7月,根据临床症状及超声或CT检查结果,60名怀疑为EC的患者治疗前接受了磁共振检查(包括APT和IVIM序列)。在60名患者中,21名患者由于以下原因被排除在外:(1)病变范围太小无法测量(n=7);(2)术后病理结果为子宫内膜息肉(n=6);(3)术后病理结果为子宫内膜浆液性癌(n=1);(4)术后病理结果为子宫内膜非典型增生(n=3);(5)扫描序列不全或图像存在过多伪影(n=3);(6)未进行手术(n=1),最终39名病理证实为EA的患者被纳入研究,年龄29~68岁。图1显示了关于患者选择的简要流程图。在39名患者中,术后病理结果为Ⅰ级的有16例,Ⅱ级为14例,Ⅲ级为9例。

图1  患者纳入流程图。APT 为酰胺质子转移成像;IVIM为体素内不相干运动成像。
Fig. 1  Patient selection flowchart. APT: amide proton transfer; IVIM: intravoxel incoherent motion.

1.2 仪器与方法

       采用3.0 T磁共振成像仪(Ingenia CX, Philips Healthcare, the Netherlands),体部18通道相控阵线圈,采取仰卧位及脚先进的方式减少小肠蠕动伪影,并适度充盈膀胱。所有患者术前均行MRI检查,除常规MRI序列(如T2WI-SPAIR、T1WI-TSE等)外,所有患者还接受了APT、IVIM序列检查,其中IVIM b值为0、50、100、150、200、400、600、800、1000、1200 s/mm2。根据常规MRI序列确认子宫位置及形态后,以矢状位T2WI-SPAIR及冠状位T2WI-TSE为定位图像,在子宫短轴位行APT及IVIM序列扫描。之后注射对比剂行动态增强磁共振成像(dynamic contrast enhanced magnetic resonance imaging, DCE-MRI)序列扫描。各序列扫描参数如表1所示。

表1  各扫描序列参数
Tab. 1  Details of MRI imaging parameters

1.3 图像处理

       将APT图像上传至工作站(IntelliSpace Portal. V9.0.4.31010)测量APT值。IVIM图像通过MRIcroGL及MITK软件进行后处理并测量IVIM衍生参数f、D、D*值。由两名对妇科影像有15年以上诊断经验的放射科副主任医师在对患者临床、病理信息不知情的情况下分别独立阅片并绘制感兴趣区(region of interest, ROI)。首先,通过T2及DCE序列图像大致确定病变范围。其次,每位医师参考T2及DCE序列图像并通过将APT与T2序列融合的方式,在APT及IVIM原始图像上绘制多边形ROI(选择与APT序列相同层面的病变在IVIM序列图像上绘制ROI),选取病灶直径大于1 cm的层面进行勾画(大于3个层面),在尽量避开囊变、坏死及肿瘤边缘等区域的基础上包括尽可能多的肿瘤实性成分,保证每位医师选取的ROI面积尽可能相近。最后,分别记录两位医师感兴趣区的APT值及IVIM衍生参数值并取两位医师测量的平均值用于最终分析,相关图片如图2, 3, 4所示。

图2  女,59 岁,Ⅰ级子宫内膜腺癌患者。2A:短轴位高b 值体素内不相干运动(IVIM)成像原始图像,宫腔内见高信号肿块;2B:IVIM衍生参数扩散系数(D)值的伪彩图;2C:IVIM衍生参数伪扩散系数(D*)值的伪彩图;2D:IVIM衍生参数灌注分数(f)值的伪彩图;2E:与T2 成像融合的酰胺质子转移成像(APT)的伪彩图;2F:病理图像(HE,×200)。两位医师测得的病变平均APT值、f值、D值、D*值分别为2.90%、0.390%、0.50×10-3 mm2/s、0.036 mm2/s。
Fig. 2  A 59-year-old woman with grade Ⅰ endometrial adenocarcinoma (EA). 2A: Short axial high b-value intravoxel incoherent motion (IVIM) original map with a high signal mass in the uterine cavity; 2B: Pseudo colored map of IVIM-derived parameter diffusion coefficient (D) value; 2C: Pseudo colored map of IVIM-derived parameter false diffusion coefficient (D*) value; 2D: Pseudo colored map of IVIM-derived parameter perfusion fraction (f) value; 2E: Pseudo colored map of amide proton transfer (APT) fused with T2 imaging; 2F: Pathological image (HE, ×200).AveragedAPT value, f value, D value and D* value of EAlesion obtained by two readers were 2.90%, 0.390%, 0.50×10-3 mm2/s, 0.036 mm2/s respectively.
图3  女,62 岁,Ⅱ级子宫内膜腺癌患者。3A:短轴位高b 值体素内不相干运动(IVIM)成像原始图像,宫腔内见高信号肿块;3B:IVIM衍生参数扩散系数(D)值的伪彩图;3C:IVIM衍生参数伪扩散系数(D*)值的伪彩图;3D:IVIM衍生参数灌注分数(f)值的伪彩图;3E:与T2 压脂成像融合的酰胺质子转移成像(APT)的伪彩图;3F:病理图像(HE,×200)。两位医师测得的病变平均APT值、f值、D值、D*值分别为3.20%、0.170%、0.60×10-3 mm2/s、0.018 mm2/s。
Fig. 3  A 62-year-old woman with grade Ⅱ endometrial adenocarcinoma (EA). 3A: Short axial high b-value intravoxel incoherent motion (IVIM) original map with a high signal mass in the uterine cavity; 3B: Pseudo colored map of IVIM-derived parameter diffusion coefficient (D) value; 3C: Pseudo colored map of IVIM-derived parameter false diffusion coefficient (D*) value; 3D: Pseudo colored map of IVIM-derived parameter perfusion fraction (f) value; 3E: Pseudo colored map of APT fused with fat suppressed T2 imaging; 3F: Pathological image (HE, ×200).AveragedAPT value, f value, D value and D* value of EA lesion obtained by two readers were 3.20%, 0.170%, 0.60×10-3mm2/s, 0.018 mm2/s respectively.
图4  女,64岁,Ⅲ级子宫内膜腺癌患者。4A:短轴位高b值体素内不相干运动(IVIM)成像原始图像,宫腔内见高信号肿块;4B:IVIM衍生参数扩散系数(D)值的伪彩图;4C:IVIM衍生参数伪扩散系数(D*)值的伪彩图;4D:IVIM衍生参数灌注分数(f)值的伪彩图;4E:与T2 压脂成像融合的酰胺质子转移成像(APT)的伪彩图;4F:病理图像(HE,×200)。两位医师测得的病变平均APT值、f值、D值、D*值分别为3.50%、0.150%、0.70×10-3 mm2/s、0.020 mm2/s。
Figure 4  A 64-year-old woman with grade Ⅲ endometrial adenocarcinoma (EA). 4A: Short axial high b-value intravoxel incoherent motion (IVIM) original map with a high signal mass in the uterine cavity; 4B: Pseudo colored map of IVIM-derived parameter diffusion coefficient (D) value; 4C: Pseudo colored map of IVIM-derived parameter false diffusion coefficient (D*) value; 4D: Pseudo colored map of IVIM-derived parameter perfusion fraction (f) value; 4E: Pseudo colored map of amide proton transfer (APT) fused with fat suppressed T2 imaging; 4F: Pathological image (HE, ×200). Averaged APT value, f value, D value and D* value of EA lesion obtained by two readers were 3.50%, 0.150%, 0.70×10-3mm2/s, 0.020 mm2/s respectively.

1.4 统计学方法

       使用 SPSS 22.0及MedCalc软件进行统计学分析,P<0.05表明差异具有统计学意义。首先通过Shapiro-Wilk检验分析APT值及IVIM各衍生参数值是否符合正态分布,符合正态分布的参数值用平均值±标准差的形式表示。使用组内相关系数(intra-class correlation coefficient, ICC)评估两位医师测量各参数结果的一致性,评估标准:≥0.75为一致性很好,0.60~0.74为一致性良好,0.40~0.59为一致性中等,<0.40为一致性差。方差齐性检验之后,采用单因素方差分析比较各级EA之间APT值及IVIM各衍生参数值(f、D、D*)的差异。之后采用受试者工作特征(receiver operating characteristic, ROC)曲线评估差异具有统计学意义的参数对EA级别的鉴别效能,并采用Delong检验比较各曲线下面积(area under the curve, AUC)的差异性。同时,对于有诊断意义的参数,我们还采用了logistic回归进行联合诊断。最后采用Spearman相关性分析,评估APT值、IVIM衍生参数值与组织学分级之间以及APT值和IVIM各衍生参数值之间的相关性。

2 结果

2.1 两位医师测量结果一致性检验

       本研究中所有参数值均符合正态分布。两位医师测量的各参数值的组内相关系数分别为:APT 0.955(95% CI:0.913~0.976);f 0.958(95% CI:0.922~0.978);D 0.964(95% CI:0.932~0.981);D* 0.825(95% CI:0.692~0.904),两位医师对APT值及IVIM各衍生参数值的测量结果均表现出很好的一致性。

2.2 三级之间APT值和IVIM衍生参数值的比较

       三级之间APT值和IVIM衍生参数值的比较见表2。Ⅰ级EA的APT值和IVIM衍生参数(f、D、D*)值分别为2.66%±0.63%、0.220%±0.080%、(0.80±0.19)×10-3 mm2/s、(0.023±0.010) mm2/s;Ⅱ级EA各参数值分别为3.17%±0.43%、0.170%±0.049%、(0.77±0.15)×10-3 mm2/s、(0.020±0.007) mm2/s;Ⅲ级EA各参数值分别为3.45%±0.42%、0.150%±0.011%、(0.69±0.10)×10-3 mm2/s、(0.017±0.007) mm2/s。

2.2.1 Ⅰ级 vs. Ⅱ级

       Ⅰ级病变的APT值显著低于Ⅱ级病变(P=0.012),Ⅰ级病变的f值显著高于Ⅱ级病变(P=0.027)。Ⅰ级与Ⅱ级病变的D值、D*值差异无统计学意义。

2.2.2 Ⅰ级 vs. Ⅲ级

       Ⅰ级病变的APT值显著低于Ⅲ级病变(P=0.001),Ⅰ级病变的f值显著高于Ⅲ级病变(P=0.010)。Ⅰ级与Ⅲ级病变的D值、D*值差异无统计学意义。

2.2.3 Ⅱ级 vs. Ⅲ级

       Ⅱ级和Ⅲ级病变之间APT值、f值、D值和D*值差异均无统计学意义。

表2  三级EA之间APT值和IVIM衍生参数值的比较
Tab. 2  Comparisons of APT values and IVIM-derived parameters among three grades

2.3 APT值、f值及二者联合诊断鉴别EA级别的效能评估及比较

2.3.1 Ⅰ级 vs. Ⅱ级

       APT值、f值及二者联合诊断鉴别Ⅰ级和Ⅱ级EA的AUC、敏感度、特异度分别为:0.78(95% CI:0.59~0.91)、92.8%、68.7%;0.70(95% CI:0.50~0.85)、71.4%、68.7%;0.83(95% CI:0.64~0.94)、92.8%、81.2%。APT值与f值的AUC之间(P=0.520,Z=0.63)、APT值与联合诊断的AUC之间(P=0.540,Z=0.59)以及f值与联合诊断的AUC之间(P=0.120,Z=1.54)的差异无统计学意义(表3图5, 6)。

图5  APT 值、IVIM各衍生参数、APT 值和f 值联合鉴别I 级和Ⅱ级子宫内膜腺癌的ROC 曲线,AUC 值分别为0.78(P<0.05)、0.70(P<0.05)、0.55(P>0.05)、0.59(P>0.05)、0.83(P<0.05)。
图6  APT 值、IVIM各衍生参数、APT 值和f 值联合鉴别I 级和Ⅲ级子宫内膜腺癌ROC曲线,AUC 值分别为0.86(P<0.05)、0.81(P<0.05)、0.68(P>0.05)、0.68(P>0.05)、0.92(P<0.05)。APT为酰胺质子转移成像;IVIM为体素内不相干运动;f 为灌注分数;D为扩散系数;D*为伪扩散系数。
Fig. 5  The ROC curves of APT values, IVIM-derived parameters and combined diagnosis of APT values and f values to differentiate grade I and II endometrial adenocarcinoma,and AUC values were 0.78 (P<0.05), 0.70 (P< 0.05), 0.55 (P>0.05), 0.59 (P>0.05), 0.83 (P<0.05).
Fig. 6  The ROC curves of APT values, IVIM-derived parameters and combined diagnosis of APT values and f values to differentiate grade I and Ⅲ endometrial adenocarcinoma, and AUC values were 0.86 (P<0.05), 0.81 (P<0.05), 0.68 (P>0.05), 0.68 (P>0.05), 0.92 (P<0.05). APT: amide proton transfer; IVIM: intravoxel incoherent motion; f: perfusion fraction; D: diffusion coefficient; D*: false diffusion coefficient.
表3  各参数对子宫内膜腺癌级别的鉴别效能
Tab. 3  Efficiency of parameters in differentiating the grade of endometrial adenocarcinoma

2.3.2 Ⅰ级 vs Ⅲ级

       APT值、f值及二者联合诊断鉴别Ⅰ级和Ⅲ级EA的AUC、敏感度、特异度分别为:0.86(95% CI:0.66~0.96)、88.8%、87.5%;0.81(95% CI:0.59~0.93)、100.0%、68.7%;0.92(95% CI:0.74~0.99)、100.0%、81.2%。APT值与f值的AUC之间(P=0.670,Z=0.41)、APT值与联合诊断的AUC之间(P=0.480,Z=0.70)以及f值与联合诊断的AUC之间(P=0.080,Z=1.73)的差异无统计学意义(表3图5~6)。

2.4 Spearman相关性分析

       APT值与EA组织学分级呈正相关,相关系数rs=0.578(95% CI:0.27~0.79,P<0.001),f值与EA组织学分级呈负相关,相关系数rs=-0.416(95% CI:-0.66~-0.11;P=0.008)。D、D*值与组织学分级之间以及APT值与IVIM各衍生参数值之间没有显著相关性(P>0.05)。

3 讨论

       本研究探讨了APT、IVIM两种功能MRI技术在评估EA术前病理分级中的价值,结果显示APT值、f值与EA组织学分级相关,有助于术前预测EA分级,APT值和f值对EA级别的鉴别效能无显著差异。各级别EA之间D值和D*值差异无统计学意义。

3.1 APT成像在EA病理分级中的价值

       APT成像技术通过检测组织内源性蛋白质或多肽中的酰胺质子来反映细胞内蛋白质含量及酸碱度的变化,其原理在于通过探测水质子信号变化间接获得细胞内蛋白质信息,APT值的高低反映了细胞内的蛋白质浓度及pH值,且有研究表明APT值与之成正相关。本次研究结果显示Ⅱ级和Ⅲ级EA APT值要高于Ⅰ级EA,这与之前的研究结果一致[11],且与其他部位恶性肿瘤APT成像的研究结果相符[12]。针对本研究结果,结合APT成像原理及相关研究,可能有以下几个原因。首先是细胞密度的差异,与低级别EA相比,高级别肿瘤细胞更为活跃,细胞增殖较快,细胞密度较低级别大,从而产生更多的细胞内蛋白和多肽,酰胺质子浓度升高,与水质子交换速率增快,APT值升高。另外,最近有研究表明[13],Ⅱ型EC(通常为Ⅲ级子宫内膜样腺癌、浆液性癌和透明细胞癌以及癌肉瘤)血管内皮生长因子(vascular endothelial growth factor, VEGF)水平通常较高。VEGF本质是一种糖蛋白,在子宫内膜表达并在血管生成、血管功能及通透性等方面发挥关键作用[14],随着肿瘤级别的增高,VEGF含量升高(蛋白含量增加),血管生成增多[15],这可能也是导致本研究的结果的原因之一。其次可能是核异型性的差异。与低级别EA相比,高级别EA分化较差,核异型性更为显著,核异型性可诱导大分子物质和疏水性细胞膜之间的相互作用,从而促进蛋白质和多肽的释放[16]。最后可能与组织坏死程度有关。由于高级别肿瘤生长较快,肿瘤组织容易发生缺氧缺血性坏死,导致局部酸性环境的形成[17],pH值减低,降低了酰胺质子与水质子的交换速率。但组织坏死过程中会向周围释放蛋白及多肽[18],导致内源性蛋白含量增加,促进了酰胺质子和水质子的交换速率。可能这一过程对APT值的影响要高于pH值降低对APT值的影响,使得高级别EA的APT值较高。本研究结果中APT成像鉴别Ⅰ级和Ⅱ级、Ⅰ级和Ⅲ级EA的AUC分别为0.83、0.84,提示具有较高的鉴别效能。术前明确肿瘤分级直接影响临床治疗方式的选择及患者预后情况,可以术前行APT检查辅助临床诊断,但因其空间分辨率较差,对肌层浸润及淋巴转移评估不具优势,而DCE可根据子宫不同结构间强化的差别评估EC肌层浸润[19],因此可考虑与DCE成像联合进一步明确诊断。

3.2 IVIM相关参数f值在EA病理分级中的价值

       IVIM是扩散加权成像的进阶模型,相比于传统DWI单指数计算模型,多b值、多种模型的IVIM成像能够更全面地反映肿瘤的微观情况。双指数模型是IVIM最为经典的模型,分别以D-mono及D-Bi(D值)、D*-mono及D*-Bi(D*值)、f-mono及f-Bi(f值)为其量化参数。本研究结果显示Ⅱ级和Ⅲ级EA的f值显著低于Ⅰ级EA (P=0.033,P=0.016),f是灌注系数,反映了肿瘤组织间毛细血管的通透性和血管的生成情况[20],此外还与核质比和组织间隙渗透压有关[21]。肿瘤的生长和侵袭有赖于肿瘤内部新生血管数量的增加,如上所述,随着肿瘤级别增高,VEGF及其受体表达增加,促进新生血管生成以及增加血管通透性[15],这似乎与本研究结果相矛盾,但也有研究表明新生血管生成过程中可能导致血管破裂,使得高级别肿瘤组织容易发生缺氧缺血性坏死[17,22],这可能导致肿瘤灌注减低。另外高级别肿瘤分化较差,核质比增高也可能导致了f值的降低。在之前的研究中[9,23, 24],f值与肿瘤级别的关系一直颇具争议,Li等[21]的研究结果显示各级别EC之间f值无显著差异,Shen等[23]认为高级别胶质瘤f值较高,而Bai等[24]认为低级别胶质瘤f值较高,因此仍需大量相关研究证实。在本研究中,f值在鉴别Ⅰ级和Ⅲ级EA时的鉴别效能(AUC=0.64)要低于APT值的鉴别效能(AUC=0.84),但差异没有统计学意义。

3.3 IVIM相关参数D值、D*值在EA病理分级中的价值

       本研究中各级别EA之间D值、D*值均无显著差异,这与之前的研究结果一致[9,25, 26]。理论上来说,D值反映的是水分子的扩散程度,主要与细胞密度有关[27]。如上所述,高级别肿瘤的细胞密度更高,核异型性较高,核仁增大,水分子扩散程度较低级别肿瘤更加受限,导致D值降低。在本次研究结果中,D值随肿瘤级别增高而减低,但差异无统计学意义,说明虽然水分子扩散程度是肿瘤特征的重要方面,但仍不能直接、全面地反映其生物学特性。且与低级别肿瘤相比,高级别肿瘤更易发生缺血坏死[17],有研究表明[28],坏死会导致ADC值的增高。与ADC值类似,D值反映的也是水的扩散程度,因此,肿瘤坏死可能也会导致高级别肿瘤D值升高,这可能是导致本结果中各级别之间D值差异无统计学意义的主要原因之一,当然也可能是样本量较少的缘故,尚需进一步扩大样本量。

       灌注参数D*值可以反映肿瘤内部灌注情况,对水分子的运动速度较敏感[29],虽然随肿瘤级别增高血供逐渐丰富,但同时会伴有微血管壁发育不良、扭曲变形、肿瘤细胞密度增大导致血管受压程度增加等情况,使微毛细血管血流量的变化不足以反映病理分级的改变,导致D*变化不明显[30]。另外也可能是因为标准差较大,数据的不稳定性和对信噪比水平的依赖性,使得D*值较难准确测量[31]

3.4 本研究的局限性

       我们的研究存在一些局限性。首先是样本量较少,且Ⅲ级病例数量较少,各级别病例分布欠均匀,今后需扩充样本量进一步验证。其次是在ROI绘制过程中避开了囊变坏死等区域,这可能不利于全面评估肿瘤特征,尚需更加深入的研究。

       综上所述,APT值和f值与EA组织学分级相关,鉴别效能较好,APT值和f值对EA病理级别的鉴别效能无显著差异,APT成像和IVIM成像有助于术前预测EA组织学分级。

[1]
Pecorelli S. Revised FIGO staging for carcinoma of the vulva, cervix, and endometrium[J]. Int J Gynaecol Obstet, 2009, 105(2): 103-104. DOI: 10.1016/j.ijgo.2009.02.012.
[2]
Lu KH, Broaddus RR. Endometrial Cancer[J]. N Engl J Med, 2020, 383(21): 2053-2064. DOI: 10.1056/NEJMra1514010.
[3]
王海尧, 薛长菊. 术前盆腹腔MRI扫描对子宫内膜癌诊治的指导价值[J]. 中国医学装备, 2019, 16(9): 79-82. DOI: 10.3969/J.ISSN.1672-8270.2019.09.021.
Wang HY, Xue CJ. Guidance value of preoperative MRI scan on abdominopelvic cavity in the diagnosis and treatment of EC[J]. China Med Equip, 2019, 16(9): 79-82. DOI: 10.3969/J.ISSN.1672-8270.2019.09.021.
[4]
Zhou JY, Payen JF, Wilson DA, et al. Using the amide proton signals of intracellular proteins and peptides to detect pH effects in MRI[J]. Nat Med, 2003, 9(8): 1085-1090. DOI: 10.1038/nm907.
[5]
Canese R. Editorial for "comparative analysis of amide proton transfer MRI and diffusion-weighted imaging in assessing p53 and ki-67 expression of rectal adenocarcinoma"[J]. J Magn Reson Imaging, 2020, 52(5): 1497-1498. DOI: 10.1002/jmri.27265.
[6]
Lecler A, Duron L, Zmuda M, et al. Intravoxel incoherent motion (IVIM) 3 T MRI for orbital lesion characterization[J]. Eur Radiol, 2021, 31(1): 14-23. DOI: 10.1007/s00330-020-07103-1.
[7]
Ochiai R, Mukuda N, Yunaga H, et al. Amide proton transfer imaging in differentiation of type Ⅱ and type I endometrial carcinoma: a pilot study[J]. Jpn J Radiol, 2022, 40(2): 184-191. DOI: 10.1007/s11604-021-01197-3.
[8]
Meng N, Fang T, Feng PY, et al. Amide proton transfer-weighted imaging and multiple models diffusion-weighted imaging facilitates preoperative risk stratification of early-stage endometrial carcinoma[J]. J Magn Reson Imaging, 2021, 54(4): 1200-1211. DOI: 10.1002/jmri.27684.
[9]
Li Y, Lin CY, Qi YF, et al. Three-dimensional turbo-spin-echo amide proton transfer-weighted and intravoxel incoherent motion MR imaging for type I endometrial carcinoma: correlation with Ki-67 proliferation status[J]. Magn Reson Imaging, 2021, 78: 18-24. DOI: 10.1016/j.mri.2021.02.006.
[10]
Li Y, Lin CY, Qi YF, et al. Non-invasive differentiation of endometrial adenocarcinoma from benign lesions in the uterus by utilization of amide proton transfer-weighted MRI[J]. Mol Imaging Biol, 2021, 23(3): 446-455. DOI: 10.1007/s11307-020-01565-x.
[11]
Meng N, Wang XJ, Sun J, et al. Evaluation of amide proton transfer-weighted imaging for endometrial carcinoma histological features: a comparative study with diffusion kurtosis imaging[J]. Eur Radiol, 2021, 31(11): 8388-8398. DOI: 10.1007/s00330-021-07966-y.
[12]
Nishie A, Takayama Y, Asayama Y, et al. Amide proton transfer imaging can predict tumor grade in rectal cancer[J]. Magn Reson Imaging, 2018, 51: 96-103. DOI: 10.1016/j.mri.2018.04.017.
[13]
Dobrzycka B, Mackowiak-Matejczyk B, Kinalski M, et al. Pretreatment serum levels of bFGF and VEGF and its clinical significance in endometrial carcinoma[J]. Gynecol Oncol, 2013, 128(3): 454-460. DOI: 10.1016/j.ygyno.2012.11.035.
[14]
Smith SK. Regulation of angiogenesis in the endometrium[J]. Trends Endocrinol Metab, 2001, 12(4): 147-151. DOI: 10.1016/s1043-2760(01)00379-4.
[15]
Kamat AA, Merritt WM, Coffey D, et al. Clinical and biological significance of vascular endothelial growth factor in endometrial cancer[J]. Clin Cancer Res, 2007, 13(24): 7487-7495. DOI: 10.1158/1078-0432.CCR-07-1017.
[16]
Tang Y, Dundamadappa SK, Thangasamy S, et al. Correlation of apparent diffusion coefficient with Ki-67 proliferation index in grading meningioma[J]. AJR Am J Roentgenol, 2014, 202(6): 1303-1308. DOI: 10.2214/AJR.13.11637.
[17]
Suzuki A, Maeda T, Baba Y, et al. Acidic extracellular pH promotes epithelial mesenchymal transition in Lewis lung carcinoma model[J]. Cancer Cell Int, 2014, 14(1): 129. DOI: 10.1186/s12935-014-0129-1.
[18]
Meng N, Wang X, Sun J, et al. Application of the amide proton transfer-weighted imaging and diffusion kurtosis imaging in the study of cervical cancer[J]. Eur Radiol, 2020, 30(10): 5758-5767. DOI: 10.1007/s00330-020-06884-9.
[19]
Fujii S, Kido A, Baba T, et al. Subendometrial enhancement and peritumoral enhancement for assessing endometrial cancer on dynamic contrast enhanced MR imaging[J]. Eur J Radiol, 2015, 84(4): 581-589. DOI: 10.1016/j.ejrad.2015.01.004.
[20]
Iima M, Reynaud O, Tsurugizawa T, et al. Characterization of glioma microcirculation and tissue features using intravoxel incoherent motion magnetic resonance imaging in a rat brain model[J]. Invest Radiol, 2014, 49(7): 485-490. DOI: 10.1097/RLI.0000000000000040.
[21]
Li XS, Wang P, Li DC, et al. Intravoxel incoherent motion MR imaging of early cervical carcinoma: correlation between imaging parameters and tumor-stroma ratio[J]. Eur Radiol, 2018, 28(5): 1875-1883. DOI: 10.1007/s00330-017-5183-3.
[22]
Williams E, Martin S, Moss R, et al. Co-expression of VEGF and CA9 in ovarian high-grade serous carcinoma and relationship to survival[J]. Virchows Arch, 2012, 461(1): 33-39. DOI: 10.1007/s00428-012-1252-9.
[23]
Shen NX, Zhao LY, Jiang JJ, et al. Intravoxel incoherent motion diffusion-weighted imaging analysis of diffusion and microperfusion in grading gliomas and comparison with arterial spin labeling for evaluation of tumor perfusion[J]. J Magn Reson Imaging, 2016, 44(3): 620-632. DOI: 10.1002/jmri.25191.
[24]
Bai Y, Lin YS, Tian J, et al. Grading of gliomas by using monoexponential, biexponential, and stretched exponential diffusion-weighted MR imaging and diffusion kurtosis MR imaging[J]. Radiology, 2016, 278(2): 496-504. DOI: 10.1148/radiol.2015142173.
[25]
Zhang Q, Yu XD, Lin M, et al. Multi-b-value diffusion weighted imaging for preoperative evaluation of risk stratification in early-stage endometrial cancer[J]. Eur J Radiol, 2019, 119: 108637. DOI: 10.1016/j.ejrad.2019.08.006.
[26]
Zhang JJ, Suo ST, Liu GQ, et al. Comparison of monoexponential, biexponential, stretched-exponential, and kurtosis models of diffusion-weighted imaging in differentiation of renal solid masses[J]. Korean J Radiol, 2019, 20(5): 791-800. DOI: 10.3348/kjr.2018.0474.
[27]
Yamada I, Sakamoto J, Kobayashi D, et al. Diffusion kurtosis imaging of endometrial carcinoma: correlation with histopathological findings[J]. Magn Reson Imaging, 2019, 57: 337-346. DOI: 10.1016/j.mri.2018.12.009.
[28]
Whittaker CS, Coady A, Culver L, et al. Diffusion-weighted MR imaging of female pelvic tumors: a pictorial review[J]. Radiographics, 2009, 29(3): 759-774. DOI: 10.1148/rg.293085130.
[29]
Federau C, Hagmann P, Maeder P, et al. Dependence of brain intravoxel incoherent motion perfusion parameters on the cardiac cycle[J/OL]. PLoS One, 2013, 8(8) [2022-05-19]. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0072856. DOI: 10.1371/journal.pone.0072856.
[30]
郝光宇, 陈蒙, 汪纯洁, 等. 体素内不相干运动及扩散峰度成像在宫颈鳞癌临床分期及病理分级中的评估价值[J]. 实用放射学杂志, 2022, 38(1): 99-103. DOI: 10.3969/j.issn.1002-1671.2022.01.024.
Hao GY, Chen M, Wang CJ, et al. The value of intravoxel incoherent motion and diffusion kurtosis imaging in clinical stage and pathological grade of cervical squamous cell carcinoma[J]. J Pract Radiol, 2022, 38(1): 99-103. DOI: 10.3969/j.issn.1002-1671.2022.01.024.
[31]
Marzi S, Piludu F, Vidiri A. Assessment of diffusion parameters by intravoxel incoherent motion MRI in head and neck squamous cell carcinoma[J]. NMR Biomed, 2013, 26(12): 1806-1814. DOI: 10.1002/nbm.3020.

上一篇 淋巴结包膜外侵犯的影像表现与前列腺癌病理分级的相关性研究
下一篇 MR集成序列在慢性冈上肌肌腱炎中的应用研究
  
诚聘英才 | 广告合作 | 免责声明 | 版权声明
联系电话:010-67113815
京ICP备19028836号-2