动态增强MR及扩散加权成像对鼻咽癌同步放化疗疗效预测研究

发布时间：2018-04-09 12:44

本文选题：鼻咽癌　切入点：扩散加权成像　出处：《北京协和医学院》2014年博士论文

【摘要】：第一部分：扩散加权成像对鼻咽癌同步放化疗疗效的预测研究 [目的]：探讨3.0T MR扩散加权成像(diffusion-weighted imaging, DWI)及表观扩散系数(apparent diffusion coefficient, ADC)值预测鼻咽癌同步放化疗疗效的价值。 [材料与方法]：2009年10月至2012年12月于我院鼻咽镜活检病理证实并行同步放化疗的鼻咽癌患者共65例。所有患者于治疗前行DWI检查(b=0,800s/mm2),测量肿瘤平均、最高及最低ADC值。并于治疗中(剂量达50Gy)、治疗末(常规放化疗结束时)再次行MR检查,分别计算治疗中、末的肿瘤消退率。依据治疗末肿瘤消退率分为完全缓解组(complete response, CR)及非CR组,后者包括部分缓解组(partial response, PR)及病变稳定组(stable disease, SD)。将治疗前肿瘤ADC值分别与治疗中、末肿瘤消退率进行Spearman目关性分析,采用独立样本T检验比较CR组与非CR组治疗前肿瘤ADC值的差异,并采用受试者工作特征曲线(receiver operating characteristic curve, ROC)分析治疗前ADC值预测治疗末CR的阈值及诊断效能。 [结果]：治疗前肿瘤平均ADC值、最高ADC值与治疗中、末肿瘤消退率均呈负相关(r=-0.463～-0.552, P0.001)。CR组及非CR组的治疗前平均ADC值分别为0.98±0.12×10-3mm2/s VS1.14±0.13×10-3mm2/s、最高ADC值分别为1.16±0.21±0.13×10-3mm2/s VS.1.42±0.23×10"3mm2/s,组间差异有统计学意义(P均0.001)。ROC显示以肿瘤治疗前平均ADC值1.09×10"3mm2/s为阈值,预测同步放化疗后CR组的敏感性、特异性及准确性分别为82.50%(33/40)、76.00(19/25)%及80.00%(52/65),曲线下面积0.816(P0.001)。 [结论]：DWI及ADC值能够较好的预测鼻咽癌治疗后消退情况,有望为鼻咽癌个体化治疗提供依据。第二部分：动态增强MRI对鼻咽癌同步放化疗疗效的预测研究 [目的]：探讨动态增强MRI (dynamic contrast enhanced magnetic resonance imaging, DCE-MRI)预测鼻咽癌同步放化疗疗效的价值。 [材料与方法]：2009年10月至2012年12月于我院鼻咽镜活检病理证实并行同步放化疗的鼻咽癌患者共61例。所有患者于治疗前行DCE-MRI检查,测量并计算肿瘤最大截面的平均、高强化区及低强化区时间信号强度曲线类型(time-signal intensity curve, TIC)及半定量参数,包括达峰时间(time to peak, TTP)、正性增强积分(positive enhancement integral, PEI)、最大上升斜率(maximum slope of increase,MSI)、最大下降斜率(maximum slope of decrease, MSD)。于治疗中(放疗50Gy)及治疗末(常规治疗结束)再次行MR检查。依据治疗末肿瘤消退率分为完全缓解组(complete response, CR组)及非CR组,后者包括部分缓解组(partial response,PR)及病变稳定组(stable disease, SD)。采用Spearman相关性检验分析治疗前肿瘤DCE-MRI参数与治疗中、末肿瘤消退率的相关性。采用非参数检验Mann-WhitneyU法比较CR及非CR组间治疗前各参数的差异。并选取与治疗末消退率相关系数最高的参数,采用受试者工作特征曲线(receiver operating characteristic curve, ROC)分析其预测治疗敏感的阈值及效能。 [结果]：与治疗中和/或治疗末消退率具有统计学意义相关性的参数包括：治疗前肿瘤平均、高强化区及低强化区的曲线类型、TTP、高强化区的PEI(r:-0.307～-0.724,P：0.001-0.016)、肿瘤平均及低强化区的MSI、肿瘤平均、高强化区及低强化区的MSD(r:0.262～0.329, P:0.001～0.037)。CR及非CR组间治疗前DCE-MRI参数差异有统计学意义的参数包括：肿瘤的平均、高强化区及低强化区的曲线类型、TTP、肿瘤低强化区的MSI、肿瘤平均及高强化区的的MSD (P:0.001～0.038)。ROC分析显示,以肿瘤低强化区TTP70s为阈值预测肿瘤治疗敏感,其敏感性、特异性及准确性分别为83.33%(30/36),76.00%(19/25)及80.33%(49/61),曲线下面积0.817(P0.001)。 [结论]：DCE-MRI半定量参数能够较好的预测鼻咽癌治疗后消退情况与治疗敏感性,有望为鼻咽癌个体化治疗提供依据。第三部分：动态增强MRI半定量及定量模型对鼻咽癌同步放化疗疗效的预测研究 [目的]：探讨动态增强MRI (dynamic contrast enhanced magnetic resonance imaging, DCE-MRI)半定量及定量Tofts模型参数预测鼻咽癌同步放化疗疗效的价值。 [材料与方法]：2013年1月至2013年12月于我院鼻咽镜活检病理证实并行同步放化疗的鼻咽癌患者共36例。所有患者于治疗前行DCE-MRI,并采用Tofts模型及GenIQ软件计算肿瘤半定量及定量参数,包括最大上升斜率MaxSlop、最大对比增强比率(Contrast Enhancement Ratio, CER)、初始钆曲线下面积(initial area under the gadolinium curve, IAUGC)、容量转移常数(volume transfer constant, Ktrans)、速率常数(rate constant, Kep)及血管外细胞外间隙容积(the extravascular-extracellular volume fraction, Ve)。于治疗中(剂量达50Gy)及治疗末(常规放化疗结束时)再次行MRI检查,分别计算肿瘤治疗中、末肿瘤消退率。依据治疗末肿瘤消退率分为完全缓解组(complete response, CR组)及非CR组,后者包括部分缓解组(partial response, PR)及病变稳定组(stable disease, SD)。将治疗前肿瘤DCE-MRI参数与治疗中、末肿瘤消退率进行Spearman相关性分析,并比较CR组与非CR组疗前DCE-MRI的参数差异。采用受试者工作特征曲线(receiver operating characteristic curve, ROC)分析治疗前DCE-MRI参数对预测肿瘤治疗敏感性的阈值及效能。 [结果]：治疗前全部肿瘤的平均及热点IAUGC、Ktrans、Kep、平均MaxSlop,热点CER与治疗中、末肿瘤消退率均呈正相关(r：0.341-0.608,P：0.001-0.042)；治疗前肿瘤的平均CER仅与治疗末肿瘤消退率呈正相关(r=0.350,P=0.036)。治疗末CR与非CR组间治疗前肿瘤的热点CER (1.802±0.297VS.1.567±0.293)、 IAUGC (0.236±0.040VS.0.193±0.031)、平均Ktrans (0.178±0.033min-1VS.0.136±0.033min-1)、热点Ktrans (0.228±0.042min-1VS.0.177±0.031min-1)、平均Kep(1.246±0.313min-1VS.0.925±0.228min-1)的差异均有统计学意义(P=0.001～0.024)。ROC分析显示治疗前平均Kep值≥1.11min-1为阈值预测肿瘤治疗末CR,其敏感性、特异性及准确性分别为71.43%(15/21)、86.67%(13/15)及77.78%(28/36),曲线下面积为0.832(P=0.001)。 [结论]：DCE-MRI半定量及定量参数均能够较好的预测鼻咽癌治疗后消退情况与治疗敏感性,有望为鼻咽癌个体化治疗提供依据。
[Abstract]:The first part: the prediction of the effect of diffusion weighted imaging on the efficacy of concurrent chemoradiotherapy for nasopharyngeal carcinoma
[Objective] to explore the value of 3.0T MR diffusion weighted imaging (diffusion-weighted imaging, DWI) and apparent diffusion coefficient (apparent diffusion coefficient, ADC) in predicting the efficacy of concurrent chemoradiotherapy for nasopharyngeal carcinoma.
[materials and methods]:2009 from October to December 2012 in our hospital epipharyngoscope biopsy of nasopharyngeal carcinoma patients with parallel chemoradiotherapy were 65 cases. All the patients in the treatment of DWI examination before (b=0800s/mm2), measuring tumor average, maximum and minimum value of ADC. And in the treatment (dose of 50Gy), at the end of treatment (conventional radiotherapy at the end of chemotherapy) again underwent MR, were calculated at the end of treatment, the rate of tumor regression. Based on the treatment of late cancer remission rate divided into remission group (complete, response, CR) and non CR group, the latter including partial remission group (partial response, PR) and stable disease group (stable disease, SD). Before treatment and treatment of tumor ADC values were, at the end of Spearman related tumor regression rate were analyzed, compared with independent sample T test group CR before treatment difference of tumor ADC values with non CR group, and the receiver operating characteristic curve (receiver operating Characteristic curve, ROC) was used to analyze the threshold and diagnostic efficiency of CR at the end of the treatment of ADC at the end of treatment.
[results]: before treatment, the mean tumor ADC values, and the highest ADC value at the end of treatment, the tumor regression rate were negatively correlated (r=-0.463, -0.552, P0.001).CR group and non CR group before treatment the average ADC = 0.98 + 0.12 + 0.13 VS1.14 * 10-3mm2/s * 10-3mm2/s, the highest ADC value was 1.16 + 0.21 + 0.13 + 0.23 * 10-3mm2/s VS.1.42 * 10 3mm2/s, there was significant difference between the groups (P 0.001).ROC display with an average ADC value of 1.09 * 10 3mm2/s tumor before treatment for the threshold, prediction of the concurrent chemoradiotherapy group CR sensitivity, specificity and accuracy were 82.50% (33/40), 76 (19/25)% and 80% (52/65), the area under the curve of 0.816 (P0.001).
[conclusion]:DWI and ADC values can be used to predict the regression of nasopharyngeal carcinoma after treatment, and it is expected to provide a basis for the individualized treatment of nasopharyngeal carcinoma.
The second part: the prediction of the effect of dynamic enhanced MRI for nasopharyngeal carcinoma with concurrent chemoradiotherapy
[Objective] to explore the value of dynamic enhanced MRI (dynamic contrast enhanced magnetic resonance imaging, DCE-MRI) to predict the curative effect of concurrent chemoradiotherapy for nasopharyngeal carcinoma.
[materials and methods]:2009 from October to December 2012 in our hospital epipharyngoscope biopsy of nasopharyngeal carcinoma patients with parallel chemoradiotherapy were 61 cases. All the patients in the treatment of DCE-MRI examination before, measure and calculate the maximum average section of tumor, high intensity region and low enhancement area time signal intensity curve type (time-signal intensity curve, TIC) and semi quantitative parameters, including the peak time (time, to, peak, TTP) positive enhancement integral (positive enhancement, integral, PEI), the maximum rise slope (maximum slope of increase, MSI), the largest decline (maximum slope of decrease slope, MSD). In the treatment (radiotherapy 50Gy) and the end of treatment (the end of the routine treatment) again examined by MR. According to the treatment at the end of the tumor regression rate divided into remission group (complete response, CR group) and non CR group, the latter including partial remission group (partial response, PR) and stable disease (Group Stable disease, SD). Using the Spearman correlation analysis before treatment and treatment of tumor DCE-MRI parameters, at the end of tumor regression rate correlation. The difference with non parametric test method of Mann-WhitneyU CR and non CR groups before treatment and select the parameters. The extinction rate of phase relationship with the highest number of parameters and the end of treatment, the subjects operating characteristic curve (receiver operating characteristic curve, ROC) to analyze the prediction of treatment and efficacy of threshold sensitive.
[results]: at the end of treatment and treatment and / or extinction rate has statistical significance correlation parameters including: before treatment, the average tumor area and low, high strength enhancement curve type, TTP District, PEI District high strength (r:-0.307 ~ -0.724, P:0.001-0.016), and low average tumor enhancement area MSI, the average tumor. High strength and low area enhancement area MSD (r:0.262 ~ 0.329, P:0.001 ~ 0.037) including parameter differences before DCE-MRI treatment parameters.CR and non CR groups had statistical significance: the average tumor, high strength and low area enhancement curve type, area TTP, tumor enhancement area of low MSI, and the average tumor high enhancement area MSD (P:0.001 ~ 0.038).ROC analysis showed that with low tumor enhancement area TTP70s threshold prediction of tumor treatment sensitivity, the sensitivity, specificity and accuracy were 83.33% (30/36), 76% (19/25) and 80.33% (49/61), the area under the curve of 0.817 (P0.001).
[conclusion]:DCE-MRI semi quantitative parameters can be used to predict the retrogression and treatment sensitivity of nasopharyngeal carcinoma, and it is expected to provide a basis for the individualized treatment of nasopharyngeal carcinoma.
The third part: a dynamic enhanced MRI semi quantitative and quantitative model for the prediction of the efficacy of concurrent chemoradiotherapy for nasopharyngeal carcinoma
[Objective] to explore the value of MRI dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) semi quantitative and quantitative Tofts model parameters in predicting the efficacy of concurrent chemoradiotherapy for nasopharyngeal carcinoma.
[materials and methods]:2013 from January to December 2013 in our hospital epipharyngoscope biopsy of nasopharyngeal carcinoma patients with parallel chemoradiotherapy were 36 cases. All patients underwent DCE-MRI in the treatment of tumors were calculated, semi quantitative and quantitative parameters of the Tofts model and GenIQ software, including the maximum rise slope of MaxSlop, the maximum contrast enhancement ratio (Contrast Enhancement Ratio, CER), the area under the curve (initial area initial GD under the gadolinium curve IAUGC (volume), volume transfer constant transfer constant, Ktrans), the rate constants (rate constant, Kep) and extravascular extracellular space volume (the extravascular-extracellular volume fraction, Ve). In the treatment (dose of 50Gy) and at the end of treatment (conventional chemotherapy end) again examined by MRI were calculated in the treatment of cancer, tumor regression rate. Based on the end of treatment at the end of the tumor regression rate is slow The solution group (complete response, CR group) and non CR group, the latter including partial remission group (partial response, PR) and stable disease group (stable disease, SD). Before treatment and treatment of tumor DCE-MRI parameters, the late Spearman associated tumor regression rate were analyzed, and the difference ratio parameters compared with the CR group and non CR group before treatment DCE-MRI. The receiver operating characteristic curve (receiver operating characteristic curve, ROC) of DCE-MRI before the treatment parameters to predict the threshold and efficiency of treatment of tumor sensitivity.
[results]: the average and hot IAUGC, before treatment, all of the tumor Ktrans, Kep, average MaxSlop, CER and hot in the treatment of late cancer remission rate were positively correlated (r:0.341-0.608, P:0.001-0.042); the average CER before treatment and treatment of tumors only at the end of tumor regression rate was positively correlated (r=0.350, P=0.036) CER for the treatment of hot. CR and non CR groups before cancer treatment at the end of (1.802 + 0.297VS.1.567 + 0.293), IAUGC (0.236 + 0.040VS.0.193 + 0.031), average Ktrans (0.178 + 0.033min-1VS.0.136 + 0.033min-1), hot Ktrans (0.228 + 0.042min-1VS.0.177 + 0.031min-1), average Kep (1.246 + 0.313min-1VS.0.925 + 0.228min-1) had significant difference (P=0.001 ~ 0.024).ROC analysis showed that before treatment the average Kep value is greater than or equal to 1.11min-1 for the prediction of tumor treatment at the end of the CR threshold, the sensitivity, specificity and accuracy were 71.43% (15/21), 86.67% (13/15) and 77.78% (28/36), The area under the curve is 0.832 (P=0.001).
[conclusion]:DCE-MRI semi quantitative and quantitative parameters can better predict the regression and therapeutic sensitivity of nasopharyngeal carcinoma after treatment, which is expected to provide a basis for individualized treatment of nasopharyngeal carcinoma.

【学位授予单位】：北京协和医学院
【学位级别】：博士
【学位授予年份】：2014
【分类号】：R739.63;R445.2

【参考文献】