3.0T MRI联合DWI和DCE在直肠癌中的临床应用价值

硕士学位论文论文题目：3.0T MRI联合DWI及DCE在直肠癌中的临床应用价值The Clinical Application Value of 3.0MRI Combining with DWI and DCE in Rectal Cancer研究生姓名:朱芸指导教师:张俊祥副教授学科专业:影像医学与核医学研究方向:影像诊断论文工作时间:2013年9月至2015年3月学位论文独创性声明本人郑重声明：所提交的学位论文是本人在导师指导下进行的研究工作和取得的研究成果。

本论文中除引文外，所有实验、数据和有关材料均是真实的。

本论文中除引文和致谢的内容外，不包含其他人或其它机构已经发表或撰写过的研究成果。

其他同志对本研究所做的贡献均已在论文中作了声明并表示了谢意。

学位论文作者签名：日期：学位论文使用授权声明研究生在校攻读学位期间论文工作的知识产权单位属蚌埠医学院。

学校有权保存本学位论文的电子和纸质文档，可以借阅或上网公布本学位论文的部分或全部内容，可以采用影印、复印等手段保存、汇编本学位论文。

学校可以向国家有关机关或机构送交论文的电子和纸质文档，允许论文被查阅和借阅。

（保密论文在解密后遵守此规定）保密论文注释：本学位论文属于保密论文，密级：_____保密期限为_____年。

学位论文作者签名：导师签名：日期：日期：目录中文摘要 (1)英文摘要 (4)正文前言 (8)第一部分材料与方法 (10)结果 (14)讨论 (21)结论 (25)第二部分材料与方法 (26)结果 (29)讨论 (40)结论 (45)第三部分材料与方法 (46)结果 (47)讨论 (51)结论 (53)参考文献 (54)致谢 (59)附录A中英文术语和缩略语对照表 (60)附录B攻读硕士期间发表文章情况 (61)附录C综述 (62)3.0T MRI联合DWI及DCE在直肠癌中的临床应用价值中文摘要研究目的探讨3.0T磁共振联合扩散加权成像及动态增强扫描技术在直肠癌的早期诊断、准确的术前分期、手术方式的选择中的临床应用价值及其与病理的分期、分型、分化程度之间的相关性。

研究方法1、一般资料及扫描方法：检查采用Philips Achieva3.0T MRI，8通道相控阵表面线圈，对60例直肠癌患者行常规、高分辨磁共振、扩散加权成像及常规增强扫描，其中随机选取30例患者做动态增强轴位扫描。

检查前1天常规口服聚乙二醇进行肠道准备，检查前禁食8小时，检查前10分钟给予盐酸山莨菪碱10mg肌肉注射。

2、扩散加权成像(diffusion weighted imaging,DWI)：b值取0和800s/mm2，采用单次激发自旋平面回波成像(spin echo-echo planar imaging,SE-EPI)技术,TR/TE=2750ms/最小值，FOV400mm×350mm，矩阵180×256，NSA=2，层厚6mm，层间距1mm。

测量直肠癌的表观扩散系数(apparent diffusion coefficient，ADC)值，比较影响直肠癌的预后因素(如病理类型、分化程度、肿瘤有无突破肠壁生长、淋巴结有无转移、直肠系膜筋膜有无受累、肿瘤的大体分型、肿瘤发生的部位、年龄及性别等)与肿瘤平均ADC值之间的相关性。

3、对60例直肠癌患者分别采用直肠常规+高分辨MRI及常规+高分辨MRI+DWI序列这两种方法对肿瘤进行术前分期，以术后病理结果为金标准，比较这两种方法分期的准确性、敏感性、特异性、阳性预测值及阴性预测值，并比较这两种方法与术后病理结果的一致性。

评估直肠系膜筋膜的受累情况，以术后病理结果为金标准，比较评估的准确性及其与术后病理结果的一致性。

在矢状位上准确测量肿块下缘距肛缘的距离，与手术后大体标本的结果进行比较。

4、动态增强扫描(dynamic contrast enhanced magnetic resonance imaging，DCE-MRI):采用e-THRIVE_dyn_BH SENSE轴位扫描，对比剂选用钆喷酸葡胺注射液(Gd-DTPA)，用高压注射器经肘静脉注入，注射流率3.0ml/s，注射剂量0.1mmol/kg。

比较不同分化程度的直肠癌及淋巴结的时间-信号强度曲线(time-intensity curve，TIC)，观察总结TIC曲线特征及有无规律性。

测量病变强化达峰值时的信号强度(SIpeak)、达峰时间(time to peak,TTP)、起始信号强度(S0)及起始强化时间(T0)，并进行比较。

5、统计学方法:采用SPSS19.0软件包统计分析数据，计量资料以均数±标准差( x±s)表示，比较采用独立样本均数的t检验、配对样本均数的t检验和单因素方差分析等。

一致性采用Kappa检验，相关性用Pearson相关分析。

当P＜0.05时认为差异有统计学意义。

研究结果1、直肠癌的平均ADC值在分化程度(P=0.000)、病理T1-2与T3-4(P=0.015)、病理N0与N1-2(P=0.001)、直肠系膜筋膜受累与未受累(P=0.014）间差异均有统计学意义。

直肠癌的平均ADC值与肿瘤大体分型（P=0.971）、肿瘤的部位（P=0.960）、性别（P=0.592）之间差异无统计学意义。

直肠癌的平均ADC值与年龄（r=0.038）无明显相关性。

2、单独使用常规+高分辨MRI与常规+高分辨MRI+DWI序列这两种方法分别进行术前T分期的准确性为76.67%和88.33%，与术后病理结果的一致性比较Kappa值为0.628和0.814。

这两种方法对T1分期的敏感性、特异性、阳性预测值、阴性预测值分别为50.00%和75.00%、100.00%和100.00%、100.00%和100.00%、96.55%和98.25%，对T2分期的敏感性、特异性、阳性预测值、阴性预测值分别为75.00%和85.00%、90.00%和95.00%、78.95%和89.47%、87.80%和92.68%，对T3分期的敏感性、特异性、阳性预测值、阴性预测值分别为82.76%和93.10%、77.42%和87.10%、77.42%和97.10%、82.76%和93.10%，对T4分期的敏感性、特异性、阳性预测值、阴性预测值分别为71.43%和85.71%、94.34%和98.11%、62.50%和85.71%、96.15%和98.11%。

单独使用常规+高分辨MRI与常规+高分辨MRI+DWI序列这两种方法分别进行术前N分期的准确性为60.00%和71.67%，与术后病理结果的一致性比较Kappa值为0.331和0.517。

这两种方法对N0分期的敏感性、特异性、阳性预测值、阴性预测值分别为63.64%和72.73%、74.07%和81.48%、75.00%和82.76%、62.50%和70.97%，对N1分期的敏感性、特异性、阳性预测值、阴性预测值分别为57.14%和71.43%、74.36%和79.49%、54.55%和65.22%、76.32%和83.78%，对N2分期的敏感性、特异性、阳性预测值、阴性预测值分别为50.00%和66.67%、87.04%和92.59%、30.00%和50.00%、94.00%和96.15%。

高分辨MRI判断直肠系膜筋膜受累情况的准确性为91.67%，与术后病理结果的一致性比较Kappa值为0.798。

高分辨MRI测量肿瘤下缘距肛缘的距离，和手术后大体标本测量结果比较，差异无统计学意义（P=0.337）。

3、根据肿瘤强化峰值出现的早晚及对比剂退出的情况将时间-信号强度曲线大致分为Ⅰ、Ⅱ、Ⅲ3种类型。

Ⅰ型是无明确峰值出现，整个检查过程信号强度始终处于缓慢上升阶段；Ⅱ型是强化早期信号明显上升，随后上升幅度减慢，逐渐达峰值，达峰值后随之出现一平台期；Ⅲ型是强化早期信号明显上升,逐渐达峰值，其后信号逐渐下降。

直肠癌TIC曲线表现不一，没有一定的规律性，这3种类型都有。

不同分化程度的直肠癌的SIpeak-S0、TTP-T0值不同，随着肿瘤分化程度的下降，SIpeak-S0逐渐增加、TTP-T0逐渐延长，但是差异无统计学意义(P ＞0.05)。

结论1、直肠癌的平均ADC值与肿瘤的病理类型、分化程度、肿瘤的T、N分期、直肠系膜筋膜浸润状态等因素有一定的相关性。

随着肿瘤分化程度的降低、浸润程度的增加、病变范围突破肠壁、有肠周淋巴结转移及直肠系膜筋膜受累时ADC 值明显降低。

直肠癌的平均ADC值与肿瘤的大体分型、肿瘤发生的部位、年龄及性别等因素无明显相关性。

因此肿瘤ADC值可以作为评价肿瘤侵袭性及预后的一个定量指标。

2、MRI常规+高分辨+DWI序列有利于直肠癌的诊断及提高术前TN分期的准确性，有利于明确肿瘤对肠壁及邻近组织器官的侵犯程度，对肿瘤T分期的准确性较高，对N分期的准确性一般，在直肠癌的诊断及术前分期中发挥着重要作用。

此外，高分辨MRI还能评估直肠系膜筋膜的受累情况，能较为准确地测量肿瘤下缘距肛缘的距离。

3、MRI动态增强扫描有利于直肠癌的诊断，不仅能反映病变的形态特征，还能够反映对比剂进入、流出肿瘤的动态过程，提供关于肿瘤局部的血液循环情况的信息。

[关键词]：直肠癌；磁共振成像；扩散加权成像；动态增强扫描；肿瘤分期The Clinical Application Value of3.0MRI Combining withDWI and DCE in Rectal CancerABSTRACTObjectiveTo discuss the clinic value of3.0T MRI combined DWI and DCE-MRI in the selection of early diagnosis,accurate pre-operational staging and surgical method of rectal cancer,as well as its correlation with staging,typing and differentiating of pathology.Materials and Methods1.General information and scanning method:Philips Achieva 3.0T MRI, SENSE-XL-TORSO8-channel phased-array surface coils.60cases of rectal cancer patients underwent conventional sequences,high-resolution MRI,diffusion-weighted imaging and conventional enhanced scan,which selected30cases plus for dynamic contrast enhanced.Take polyethylene glycol for bowel preparation on one day before the examination;fasting8hours before the examination;intramuscularly the patients with10mg anisodamine10minutes before the examination.2.DWI:Use spin echo-echo planar imaging(SE-EPI)sequence,b takes0and 800s/mm2,TR/TE=2750ms/min,FOV400mm×350mm,matrix180×256,NSA=2,slice thickness6mm,slice gap1mm.Measure the apparent diffusion coefficient(ADC) value of rectal cancer;compare the correlation between the prognostic factors(e.g. pathological pattern,differentiated degree,whether the tumor grows by breaking through the intestinal wall,whether lymph gland is metastasis,whether mesorectal fascia is involved,gross types of tumor,location of tumor,age and gender)of rectal cancer and the average ADC value of tumor.3.60cases of patients with rectal cancer were use conventional sequences and HR MRI with and without DWI,and use these two methods for preoperative staging of tumors,the result of postoperative pathology is regarded as a gold pare the accuracy,sensibility,specificity,positive predictive value and negative predictive value of these two staging method,as well as the consistency between these methods and postoperative pathology.Evaluate the affected situation of mesorectal fascia,the result of postoperative pathology is regarded as the gold pare the accuracy of evaluation and consistency of postoperative pathology result.Measure thedistance between the inferior border and anal verge of lump from the sagittal view, and compare it to the result of pathologic sample after operation.4.DCE-MRI:e-THRIVE_dyn_BH SENSE axial scanning,select Gd-DTPA as contrast agent,and inject it with high pressure injector from elbow vein.The injection flow rate:3.0ml/s,injection dose:0.1mmol/pare the rectal cancers of different differentiated degree and the time-intensity curve(TIC)of lymph node,observe and summarize the TIC curve characters and its regularity.Measuring the signal indensity of peak enhancement(SIpeak),the time to peak(TTP),the starting signal strength(S0) and the starting time of strengthening(T0),and take them into comparison.5.Statistical Methods:SPSS19.0software package is adopted to statistically analyze data,the measurement data is shown in the form of( x±s),the comparison adopts t test of independent sample mean,t test of paired sample mean and one-way analysis of variance,etc.Consistency adopts Kappa inspection,while correlation is analyzed by Pearson.Test levelα=0.05,with P<0.05was considered to be statistically significant. Results1.The averaged ADC value of rectal cancer is of statistical significance with differentiated degree(P=0.000),pathology T1-2and T3-4(P=0.015),pathology N0 and N1-2(P=0.001),affected mesorectal fascia and unaffected one(P=0.014).The averaged ADC value of rectal cancer has no statistical significance with gross type of tumor(P=0.971),tumor position(P=0.960),gender(P=0.592).There is no significant correlation between the average ADC value and age(r=0.038)of rectal cancer.2.The methods Convention+high resolution MRI as well as convention+high resolution MRI+DWI sequence are respectively adopted,and the accuracy of preoperative T staging is76.67%and88.33%.Compared with the consistency of postoperative result,the Kappa value is0.628and0.814.The sensibility,specificity, positive predictive value,negative predictive value of these two methods on T1 staging is50.00%and75.00%,100.00%and100.00%,100.00%and100.00%, 96.55%and98.25%;on T2staging is75.00%and85.00%,90.00%and95.00%, 78.95%and89.47%,87.80%and92.68%;on T3staging is82.76%and93.10%, 77.42%and87.10%,77.42%and97.10%,82.76%and93.10%;and on T4staging is 71.43%and85.71%,94.34%and98.11%,62.50%and85.71%,96.15%and98.11%. The accuracy of these two methods on pre-operation N-staging is60.00%and71.16%respectively,and compared to the consistency of preoperative pathology result,the Kappa value is0.331and0.517respectively.The sensitivity,specificity,positive predictive value,negative predictive value of these two methods on N0staging is 57.14%and71.43%,74.36%and79.49%,54.55%and65.22%,76.32%and83.78% respectively;on N2staging is50.00%and66.67%，87.04%and92.59%,30.00%and 50.00%,94.00%and96.15%.The accuracy of high resolution MRI on measuring the affected situation of mesorectal fascia is91.67%,and compared to the consistency of post-operational pathology result,the Kappa value pared the distance between the inferior border of tumor to anal verge measured under the high resolution MRI,with the measuring result of gross sample after operation,the difference is not statistically significant(P=0.037).3.According to the appearance time of tumor intensity peak as well as the quitting time of contrast agent,the time-single intensity curve can be approximately divided into three types:type I,type II and type III.In type I,there is no clear peak,and the whole process is at a slowly rising state;in type II,the signal rises significantly at the early stage of enhancement.The rising range slows down to a peak level.After the peak,there is a platform stage;in type III,the signal rises significantly at the early stage of enhancement.After reaching at the peak,the signal declines gradually.In these three types,the TIC curve of rectal cancer presents differently without certain order.Under differentiated degree,the SIpeak-S0,TTP-T0value of rectal cancer is different.With the declining of tumor differentiated degree,the SIpeak-S0rises gradually while TTP-T0also lengthens gradually,but the difference is not statistically significant.(P＞0.05)Conclusion1.The average ADC value of rectal cancer is of certain correlation with the pathological type,differentiated degree of tumor,T,N staging of tumor,and infiltration state of mesorectal fascia.Under the condition where the differentiated degree of tumor decreases,infiltration degree increases,the lesion range breaks through the intestinal wall,lymphatic metastasis and the mesorectal fascia is affected, the ADC value declines significantly.The average ADC value of rectal cancer is not statistically significant with the gross type of tumor,tumor position,age and gender. Therefore,the ADC value of tumor can be regarded as a quantitative index to evaluate tumor invasion and prognosis.2.MRI convention+high resolution+DWI sequence is conductive to the diagnosis of rectal cancer and improving the accuracy of TN staging.It is also conductive to ensuring the invasion degree of tumor to intestinal wall and other organs nearby.This method has high accuracy on T staging of tumor but normal accuracy on N staging.Thus,it plays an important role in the diagnosis of tumor and preoperative staging.In addition,high resolution MRI can either evaluate the affected condition of mesorectal fascia,or accurately measure the distance between the inferior borders of tumor to anal verge.3.MRI dynamic enhancement scan is conductive to the diagnosis of rectal cancer, which can not only reflect the morphological characteristics of lesion,but presents the dynamic process of contrast agent when it inflows into or outflows the tumor.Further, it also provides the information of blood circulation condition of the local tumor. [Key words]:Rectal cancer;Magnetic resonance imaging;Diffusion weighted imaging;Dynamic contrast-enhanced magnetic resonance imaging;Tumor stage前言一般把发生在齿状线至直肠乙状结肠交界处之间的肿瘤认为是直肠癌，在消化道肿瘤中[1-3]占有很大的比例。