發(fā)布時間：2018-05-28 13:31

  本文選題：乳腺腫瘤 + 磁共振成像�。� 參考：《北京協(xié)和醫(yī)學院》2014年博士論文

【摘要】：第一部分動態(tài)增強MRI預(yù)測乳腺癌新輔助化療療效 目的探討動態(tài)對比增強MRI(DCE-MRI)早期預(yù)測乳腺癌新輔助化療(NAC)療效的可行性。 資料與方法收集2011年5月～2014年1月計劃在我院行NAC的乳腺癌患者共112例,年齡23～67歲,中位年齡48歲,于化療前和NAC2周期后行DCE-MRI掃描,分別采用ROIw,ol。法和ROIhs法測量腫瘤三個方位最大徑、半定量參數(shù)最大上升斜率(MSI)、第二期強化程度(SI2%)、峰值強化程度(SIpeak%)、正性增強積分(PEI)、最大下降斜率(MSD)、達峰時間(TTP)以及R0Iwhole法測量定量參數(shù)容量轉(zhuǎn)移常數(shù)(Ktrans)、速率常數(shù)(Kep)、細胞外血管外間隙容積比(Ve)。按照MillerPayne分級系統(tǒng),將腫瘤反應(yīng)性分為組織學顯著反應(yīng)(MHR)和組織學非顯著反應(yīng)(NMHR)。分別比較MHR組及NMHR組化療前與NAC2周期后的量化參數(shù),并比較MHR組和NMHR組NAC2周期后各參數(shù)的變化值以及基線參數(shù)值,正態(tài)分布者行兩獨立(或配對)樣本t檢驗,非正態(tài)分布者行兩獨立樣本(或配對資料)的非參數(shù)檢驗,繪制受試者工作曲線(ROC),找到早期預(yù)測乳腺癌NAC療效的最佳參數(shù)及其診斷閾值,并比較兩種ROI選取方法的預(yù)測效能。 結(jié)果最終入組行半定量分析者72例,NMHR組53例,MHR組19例；行定量分析者62例,NMHR組45例,MHR組17例。無論MHR組還是NMHR組,NAC前與化療2周期后的腫瘤大小均有統(tǒng)計學差異(P0.001)；MHR組NAC前與化療2周期后的半定量參數(shù)及定量參數(shù)均有統(tǒng)計學差異(P0.05)；NMHR組除MSD(ROIwhole)、MSD(ROIhs)、TTP(ROIhs)、KepVe外(P值分別為0.615,0.205,0.085,0.118,0.236),其余DCE-MRI參數(shù)在NAC前與化療2周期后的差異均有統(tǒng)計學意義(P0.05)；NAC2周期后腫瘤各徑線消退率、半定量參數(shù)變化值及定量參數(shù)變化值在MHR組和NMHR組間的差異均有統(tǒng)計學意義(P0.05),即MHR組參數(shù)下降較NMHR組明顯；而化療前的各參數(shù)在兩組間的差異無統(tǒng)計學意義(P0.05)。ROC分析顯示以腫瘤三徑線幾何均值的變化值(△G)預(yù)測NAC療效的曲線下面積、敏感度、特異度和診斷閾值分別為0.908、84.2%、92.5%和-43.7%,以K。。的變化值預(yù)測的曲線下面積、敏感度、特異度和診斷閾值分別為0.890、76.5%、93.3%和-54.8%。比較ROIwhole法和R0Ihs法測得的半定量參數(shù)MSI、SI2%、SIpeak%、PEI和MSD變化值預(yù)測NAC療效的ROC曲線下面積,P值分別為0.769、0.588、0.490、0.807和0.793。 結(jié)論DCE-MRI量化參數(shù)可于化療2周期預(yù)測乳腺癌NAC的療效,而化療前的參數(shù)尚不能預(yù)測化療的最終療效,其中△G、△Kep是較好的預(yù)測指標,兩種ROI選取方法預(yù)測乳腺癌NAC療效的效能相當。 第二部分動態(tài)增強MRI參數(shù)與乳腺癌分子亞型的關(guān)系研究 目的探討動態(tài)對比增強MRI (DCE-MRI)量化參數(shù)與乳腺癌不同分子亞型及預(yù)后因子的關(guān)系。 資料與方法以第一部分中收集的2011年5月-2014年1月計劃在我院行新輔助化療(NAC)的112例乳腺癌患者為研究對象,行DCE-MRI檢查,分別采用ROIwhole法和ROIhs法測量半定量參數(shù)最大上升斜率(MSI)、第二期強化程度(SI2%)、峰值強化程度(SIpeak%)、正性增強積分(PEI)、最大下降斜率(MSD)、達峰時間(TTP)以及ROIwhole法測量定量參數(shù)容量轉(zhuǎn)移常數(shù)(Ktrans)、速率常數(shù)(Kep)、細胞外血管外間隙容積比(Ve)。采用SP法檢測免疫組化指標ER、PR、HER-2及Ki-67,對HER-2(++)者行免疫熒光原位雜交法(FISH)檢測,以此將乳腺癌分為Luminal A、Luminal B、HER-2+和三陰性乳腺癌(TNBC)四個亞型。采用兩獨立樣本t檢驗或Mann-WhitneyU檢驗比較ER、PR和HER-2不同表達狀態(tài)時各參數(shù)的差別,行單因素方差分析比較不同分子亞型間的半定量參數(shù),行Mann-Whitney U檢驗對不同分子亞型間定量參數(shù)兩兩比較,P≤0.05為差異有統(tǒng)計學意義。 結(jié)果最終入組行半定量分析的101例患者,有4例未行FISH檢測,Luminal A型15例,Luminal B型56例,HER-2+型12例,TNBC14例。PR+者的MSD高于PR-者(P=0.002,0.016),HER-2+者的MSI高于HER-2-者(P=0.021,0.011),PR-者的SI2%(ROI,,hole)高于PR+者(P=0.031),HER-2-者的TTP(ROIhs)高于HER-2+者(P=0.029),其余半定量參數(shù)在不同受體表達狀態(tài)時的差異無統(tǒng)計學意義(P0.05)。除MSD(ROI,,hole)外(P=0.045),其余半定量參數(shù)在不同分子亞型間的差異無統(tǒng)計學意義,但進一步兩兩比較,各組間差異均無統(tǒng)計學意義。最終入組行定量分析的82例患者,有3例未行FISH檢測,Luminal A型13例,Luminal B型42例,HER-2+型11例,TNBC13例。不同受體表達狀態(tài)時定量參數(shù)的差異無統(tǒng)計學意義(P0.05)。Lumianl A型和TNBC、LumianlA型和HER-2+型的Ktrans值有統(tǒng)計學差異(P=0.026,0.047),Luminal B型和TNBC的Kep值有統(tǒng)計學差異(P=0.013),其余任何兩型間的定量參數(shù)值均無統(tǒng)計學差異。 結(jié)論不同PR表達狀態(tài)時的MSD不同,PR+者高于PR-者,HER-2過表達者的MSI較正常表達者顯著性增高；TNBC和HER-2+型的Ktrans值高于Luminal A型,TNBC的K。。值高于Luminal B型,即TNBC和HER-2+型的局部血流灌注較高,Luminal A和Luminal B型則相反,揭示了各分子亞型的不同生物學特性。 第三部分不同分子亞型對動態(tài)增強MRI預(yù)測乳腺癌NAC療效準確性的影響 目的探討乳腺癌不同分子亞型對動態(tài)對比增強MRI(DCE-MRI)量化參數(shù)預(yù)測乳腺癌新輔助化療(NAC)療效準確性的影響。 資料與方法收集2011年5月～2014年1月計劃在我院行NAC的乳腺癌患者共112例,分別于化療前和NAC2周期后行DCE-MR T掃描,分別采用R0Iwhole法和R0Ihs法測量腫瘤三個方位最大徑、半定量參數(shù)最大上升斜率(MSI)、第二期強化程度(SI：%)、峰值強化程度(SIpeak%)、正性增強積分(PEI)、最大下降斜率(MSD)、達峰時間(TTP)以及ROIwhole法測量定量參數(shù)容量轉(zhuǎn)移常數(shù)(Ktrans)、速率常數(shù)(Kep)、細胞外血管外間隙容積比(V。)。按照MillerPayne分級系統(tǒng),將腫瘤反應(yīng)性分為組織學顯著反應(yīng)(MHR)和組織學非顯著反應(yīng)(NMHR)。采用SP法檢測免疫組化指標ER.PR.HER-2及Ki-67,對HER-2(++)者行免疫熒光原位雜交法(FISH)檢測,以此將乳腺癌分為Luminal A、Luminal B、HER-2+和三陰性乳腺癌(TNBC)四個亞型。行二分類Logi stic回歸,從患者年齡、腫瘤形態(tài)、化療方案和分子亞型中篩選出影響病理與影像預(yù)測結(jié)果一致性的因素；計算DEC-MRI參數(shù)預(yù)測乳腺癌不同分子亞型NAC療效的敏感度、特異度、陽性預(yù)測值、陰性預(yù)測值和準確度。 結(jié)果最終入組69例,MHR組19例,NMIIR組50例；Luminal A型9例,Luminal B型39例,HER2+型8例,TNBC13例。經(jīng)過篩選,分子分型是影響I)CEMRl預(yù)測乳腺癌NAC療效準確性的因素,對SIpenk%(ROIwhole)、MSD(ROIwhone)及SI：%(ROIts)三個參數(shù)的影響較大,比值比分別為2.580、0.437和2.569。定量參數(shù)(Ktrans、Kep、Vec)及左右徑預(yù)測不同分子亞型乳腺癌NAC療效的準確性均較高；對于腫瘤前后徑和MSD(ROIwhole),Lumianl的準確度較高,分別為87.5%和89.6%,高于HER-2+型(75.0%和87.5%)和TNBC(76.9%和61.5%)；對于上下徑及半定量參數(shù)MSI(ROIwhole)、SI2%(ROIwhols)、SIpeak%(ROIwhole)、PEI(ROIwhole)、MSI(ROIhs)、SI2%(R0Ihs)、SIpeak%(ROIh、)、PEI(ROIhs)、MSD(ROIhs),Luminal型乳腺癌預(yù)測的準確性較差(分別為68.7%,77.1%,70.8%,62.5%,73.0%,79.2%,64.6%,64.6%,66.7%,72.9%),HER-2+(分別為100.0%,87.5%,87.5%,87.5%,87.5%,100.0%,100.0%,87.5%,87.5%,100.0%)和TNBC的準確性較高(分別為92.3%,100.0%,100.0%,100.0%,100.0%,100.0%,100.0%,100.0%,92.3%,84.6%)。 結(jié)論定量參數(shù)和在延遲MRI圖像(高分辨率)上測得的腫瘤左右徑預(yù)測NAC療效的準確性較高,不受分子亞型影響；通過大多數(shù)半定量參數(shù)預(yù)測TNBC和HER-2+型NAC療效的準確性較高,Luminal型則較低。
[Abstract]:Part I dynamic contrast-enhanced MRI predicts the efficacy of neoadjuvant chemotherapy for breast cancer.
Objective to explore the feasibility of dynamic contrast-enhanced MRI (DCE-MRI) in predicting the efficacy of neoadjuvant chemotherapy (NAC) in breast cancer.
Materials and methods 112 cases of breast cancer were collected from May 2011 to January 2014 in our hospital. The age of 112 cases of breast cancer was 23~67 years old. The median age was 48 years old. DCE-MRI scan was performed before and after the chemotherapy and NAC2 cycle. The maximum diameter of three azimuth, the maximum slope of semi quantitative parameters (MSI), and second phases of strengthening of the tumor were measured by ROIw, ol. and ROIhs. Degree (SI2%), peak Intensification (SIpeak%), positive enhancement integral (PEI), maximum descent slope (MSD), peak time (TTP) and quantitative parameter volume transfer constant (Ktrans), rate constant (Kep) and extracellular space volume ratio (Ve) of extracellular blood vessel (Ve) measured by R0Iwhole method. According to MillerPayne grading system, the tumor reactivity was divided into a significant histological response (MHR) and histology non significant response (NMHR). Compare the quantitative parameters of the MHR and NMHR groups before and after the NAC2 cycle, and compare the changes of the parameters and the baseline parameters of the MHR and NMHR groups after NAC2 cycle, the normal distribution of the two independent (or paired) sample t test, and the non normal distribution of the two independent samples (or paired data). Nonparametric test, the subject work curve (ROC) was drawn to find the best parameters for early prediction of breast cancer NAC and its diagnostic threshold, and to compare the predictive effectiveness of the two ROI selection methods.
Results in the final group, there were 72 semi quantitative analysts, 53 cases in group NMHR and 19 in group MHR, 62 cases of quantitative analysis, 45 cases in group NMHR, 17 cases in group MHR. The tumor size before NAC before and after 2 cycles of chemotherapy was statistically significant (P0.001); the semi quantitative parameters and quantitative parameters of MHR group before and after 2 cycles of chemotherapy were statistically significant. Difference (P0.05); group NMHR except MSD (ROIwhole), MSD (ROIhs), TTP (ROIhs), KepVe (P value was 0.615,0.205,0.085,0.118,0.236), and the difference between the remaining DCE-MRI parameters before and after 2 cycles of chemotherapy. The difference between the NMHR group and the NMHR group was statistically significant (P0.05), that is, the parameters of the MHR group were significantly lower than those in the NMHR group, but the difference between the parameters before the chemotherapy was not statistically significant (P0.05).ROC analysis showed that the variation value of the geometric mean of the tumor (delta G) predicted the area under the curve of the NAC efficacy, sensitivity, specificity and diagnostic threshold, respectively. For 0.908,84.2%, 92.5% and -43.7%, the area under the curve, sensitivity, specificity and diagnostic threshold were 0.890,76.5%, 93.3% and -54.8%. compared with the ROIwhole method and the R0Ihs method to determine the semi quantitative parameter MSI, and the SI2%, SIpeak%, PEI and MSD variations predicted the area under the curve. .807 and 0.793.
Conclusion DCE-MRI quantized parameters can be used to predict the curative effect of breast cancer NAC in 2 cycle of chemotherapy, and the parameters before chemotherapy can not predict the final curative effect of chemotherapy, of which Delta G and delta Kep are good predictors, and the effectiveness of the two ROI selection methods to predict the effect of breast cancer NAC is quite effective.
The second part is the relationship between dynamic enhanced MRI parameters and molecular subtypes of breast cancer.
Objective to explore the relationship between dynamic contrast enhanced MRI (DCE-MRI) quantification parameters and different molecular subtypes and prognostic factors of breast cancer.
Materials and methods: in the first part, 112 cases of breast cancer who were planned in January -2014 May 2011 in January -2014 were studied. The DCE-MRI examination was performed. The maximum slope of semi quantitative parameters (MSI) was measured by ROIwhole method and ROIhs method respectively. The second phase intensification degree (SI2%), and the peak intensification degree (SIpeak%) were taken. Sex enhancement score (PEI), maximum descent slope (MSD), peak time (TTP) and quantitative parameter volume transfer constant (Ktrans), rate constant (Kep), extracellular space volume ratio (Ve) of extracellular blood vessel (Ve) were measured by ROIwhole method. Immunochemical index ER, PR, HER-2 and Ki-67 were detected by SP method, and the immunofluorescence in situ hybridization (+ +) was detected by SP method. Breast cancer was divided into four subtypes of Luminal A, Luminal B, HER-2+ and three negative breast cancer (TNBC). Two independent samples t test or Mann-WhitneyU test were used to compare the difference of the parameters when ER, PR and HER-2 were expressed in different states. Single factor variance analysis was used to compare the semi quantitative parameters of different molecular subtypes. There was a significant difference between the quantitative parameters of sub subtypes 22 and P < 0.05.
Results there were 101 cases of semi quantitative analysis. There were 4 cases without FISH detection, 15 cases of Luminal A, 56 cases of Luminal B, 12 cases of HER-2+ type, and MSD higher than PR- in TNBC14 case.PR+. R-2+ (P=0.029), the difference of the other semi quantitative parameters in the expression of different receptors was not statistically significant (P0.05). Except MSD (ROI, hole), the difference between the other semi quantitative parameters was not statistically significant in the subtypes of different molecules, but there was no statistical significance in the further 22 ratio, and the final group was 8. In 2 cases, there were 3 cases without FISH detection, 13 cases of Luminal A, 42 cases of Luminal B, 11 cases of HER-2+ type, and TNBC13 cases. There was no statistical difference between the quantitative parameters of the expression of different receptors (P0.05).Lumianl A and TNBC. Difference (P=0.013), no significant difference was found in the quantitative parameters between the other two types.
Conclusion the MSD of different PR expressions is different, the PR+ is higher than the PR-, the MSI of HER-2 overexpression is significantly higher than that of the normal expression. The Ktrans value of TNBC and HER-2+ is higher than that of Luminal A. Different biological characteristics of subtypes.
The third part of the impact of different molecular subtypes on the accuracy of dynamic enhanced MRI in predicting the efficacy of NAC in breast cancer
Objective to investigate the effect of different molecular subtypes of breast cancer on the accuracy of dynamic contrast enhanced MRI (DCE-MRI) in predicting the efficacy of neoadjuvant chemotherapy (NAC) in breast cancer.
Data and methods 112 cases of breast cancer were collected from May 2011 to January 2014 in our hospital. DCE-MR T scan was performed before and after the chemotherapy of NAC. The maximum diameter of three azimuth, the maximum slope of semi quantitative parameter (MSI), the second intensification degree (SI:%) and the peak strengthening process were measured by R0Iwhole and R0Ihs respectively. Degree (SIpeak%), positive enhancement integral (PEI), maximum descent slope (MSD), peak time (TTP) and quantitative parameter volume transfer constant (Ktrans) measured by ROIwhole, rate constant (Kep), extracellular space volume ratio (V.) in extracellular space (V.). According to the MillerPayne grading system, the tumor reactivity was divided into histological significant reaction (MHR) and histology nonsignificant. Reaction (NMHR). The immunofluorescence index ER.PR.HER-2 and Ki-67 were detected by SP, and the HER-2 (+ +) was detected by immunofluorescence in situ hybridization (FISH), and the breast cancer was divided into four subtypes: Luminal A, Luminal B, HER-2+ and three negative breast cancer (TNBC). The factors that affect the consistency of the pathological and image prediction results were screened, and the sensitivity, specificity, positive predictive value, negative predictive value and accuracy of the DEC-MRI parameters were calculated to predict the efficacy of different molecular subtypes of NAC in breast cancer.
The results were 69 cases, 19 cases in group MHR, 50 cases in group NMIIR, 9 cases of Luminal A, 39 cases of Luminal B, 8 cases of HER2+, TNBC13 case. The molecular typing is the factor influencing the accuracy of I) CEMRl to predict the NAC effect of breast cancer. The accuracy of the 37 and 2.569. quantitative parameters (Ktrans, Kep, Vec) and the left and right diameters for the prediction of NAC for different molecular subtypes of breast cancer were higher. The accuracy of Lumianl was higher for the anterior and posterior diameter of the tumor and MSD (ROIwhole), 87.5% and 89.6%, respectively, higher than the HER-2+ (75% and 87.5%) and TNBC (76.9% and 61.5%), and MSI (ROI) for the upper and lower diameter and the semi quantitative parameter (ROI). Whole), SI2% (ROIwhols), SIpeak% (ROIwhole), PEI (ROIwhole), MSI (ROIhs), SI2% (R0Ihs), SIpeak% (R0Ihs), poor prediction accuracy (68.7%, 77.1%, 70.8%, 62.5%, 73%, 79.2%, 64.6%, 64.6%, 66.7%, 72.9% respectively). %, 100%) and TNBC were of high accuracy (92.3%, 100%, 100%, 100%, 100%, 100%, 100%, 100%, 92.3%, 84.6%) respectively.
Conclusion the accuracy of the quantitative parameters and the predicted NAC effect in the delayed MRI image (high resolution) is higher, not affected by the molecular subtype, and the accuracy of the TNBC and HER-2+ type NAC is higher and the Luminal type is lower by most semi quantitative parameters.
【學位授予單位】：北京協(xié)和醫(yī)學院
【學位級別】：博士
【學位授予年份】：2014
【分類號】：R737.9;R445.2

【參考文獻】

相關(guān)期刊論文 前2條

1 胡文娟;張蓓;潘玉林;謝蘊;劉維燕;;磁共振灌注成像對局部進展期乳腺癌新輔助化療療效的評價[J];放射學實踐;2011年01期

2 張仁知;周純武;李靜;歐陽漢;王少明;;MR增強參數(shù)預(yù)測乳腺癌新輔助化療療效的前瞻性研究[J];放射學實踐;2013年06期

，

本文編號：1946902