【作者】 胡以达；

【导师】 王学峰；

【作者基本信息】 重庆医科大学 ， 神经病学， 2012， 博士

【摘要】 目的本研究拟采用前瞻性设计，通过多模态功能磁共振技术，从脑功能与脑结构两个方面寻找能够早期预测、预报抗癫痫药物疗效的影像学标记，阐释其可能的机制。方法本研究使用GE3.0T磁共振采集新诊断未服药癫痫患者服用抗癫痫药物前的磁共振静息状态功能数据，弥散张量成像数据以及高分辨率三维结构相数据。完成后随访观察并评价新诊断癫痫患者用药效果，根据癫痫控制情况将新诊断患者分为癫痫治疗无效的未控制组与癫痫治疗有效的控制组。从新诊断癫痫患者脑部低频振幅活动入手，结合临床随访观察结果，利用血氧水平依赖信号，通过与正常对照、耐药性癫痫患者以及癫痫控制良好即将停药患者进行对比，以探寻对药物敏感性不同患者脑区活动的差异，继而构建感兴趣区作为功能连接的节点，辐射全脑探寻功能网络的改变，同时利用弥散张量成像技术对患者全脑白质纤维微结构改变进行评价。结果共计53名被试者纳入本研究最后数据分析，新诊断未服药癫痫患者21人，其中随访观察证实为癫痫治疗无效的未控制组11人，癫痫治疗有效的控制组10人，与新诊断癫痫患者年龄、性别、利手相匹配正常的对照13人，符合耐药性癫痫诊断标准患者8人，癫痫控制良好即将停药患者11人。结果发现：①新诊断癫痫患者未控制组、患者控制组以及正常对照组三组在双侧楔叶、双侧舌回，右侧枕叶中部以及右侧额下回的低频活动存在显著性差异；相对于正常对照和癫痫控制组，癫痫未控制组在枕叶中、下部尤其是楔叶、舌回等视觉皮层区域的低频活动明显增高，而在额中回、额下回以及左侧顶下小叶处的低频活动受到抑制；癫痫未控制患者低频活动异常增高的视觉皮层区域也在耐药性癫痫患者组中被观察到有低频活动的异常增高，这一区域的异常信号在新诊断癫痫患者癫痫控制组和癫痫控制良好即将停药患者组中没有被观察到；计算枕叶视觉皮层区功率谱后发现在0.01-0.08HZ区域的低频振幅活动，存在癫痫未控制患者＞癫痫控制患者＞正常对照被试的趋势；计算每一位癫痫未控制患者，癫痫控制患者和正常被试Brodmann17区的平均低频振幅率值，进行三组的单因素方差分析，结果显示三组在Brodmann17区平均低频振幅值存在显著差异（F=6.279，P=0.005）；进行两两分析发现：癫痫未控制组与正常被试Brodmann17区的平均低频振幅率值有显著差异（t=3.660，P=0.001），癫痫未控制组与癫痫控制组也有显著差异（t=2.389，P=0.026），控制组Brodmann17区的平均低频振幅率值略高于正常被试，但没有统计差异（t=0.781，P=0.443）；构建受试者工作曲线后发现，癫痫未控制组-癫痫控制组曲线下面积为0.846，p=0.004，95%CI [0.690-1.002]，证明Brodmann17区平均低频振幅值能将癫痫未控制患者与正常被试区分开，并且当平均低频振幅率取值1.146时，可获得81.8%的灵敏度与76.9%的特异性；癫痫未控制组-癫痫控制组曲线下面积为0.782，p=0.029，95%CI[0.585-0.979]，证明Brodmann17区平均低频振幅值能将癫痫未控制患者与癫痫控制患者区分开，并且当平均低频振幅率取值1.201时，可获得72.7%的灵敏度与70.0%的特异性。②全脑体素与Brodmann17区的功能连接发现，癫痫未控制组中与Brodmann17区正性功能连接异常增高区域是：左侧中央后回和双侧顶上小叶；负性功能连接异常增高的区域为：双侧后扣带回、双侧楔前叶下部、内侧前额叶、额上回和额中回。癫痫控制组患者中与Brodmann17区仅在左枕颞交界区有正性功能连接减弱。研究还发现，癫痫未控制组中与Brodmann17区存在异常功能连接的脑区个数与体积均显著多于癫痫控制组。③对反映脑白质微结构完整性的各向异性分数进行的基于白质范围的体素测量研究发现，癫痫未控制组患者双侧额叶下白质，双侧基底节周围白质和左侧额顶交界皮层下白质的各向异性分数异常降低。癫痫控制组患者仅扣带回中部上方白质存在各向异性分数下降，而左侧胼胝体前部各向异性分数则增高。结论本研究发现静息状态下枕叶视觉皮层区域尤其是Brodmann17区有异常增高低频电活动的新诊断癫痫患者对抗癫痫药物敏感性较低。药物控制这类患者癫痫发作更加困难的原因可能与其大脑Brodmann17区与顶上小叶之间构成了异常联系的功能网络有关。并且这类患者在额叶区域以及额顶交界区域的白质结构的损害可能与耐药性癫痫异常神经网络有关，而损伤数目以及范围的差异可能与患者对药物敏感性相关。更多还原

【Abstract】 ObjectiveIn this prospective observational cohort study, we used multimode functionalmagnetic resonance imaging to search the biomarkers which could predictresponse of newly diagnosed epileptic patients to antiepileptic drugs beforedrug taking.MethodsIn this study, the MRI data of newly diagnosed epileptic patients, healthycontrols, drug-resistant epileptic patients, and drug withdrawal patients wereacquired by using G.E.3.0-Tesla scanner. Functional images data werecollected by using an echo-planar imaging sequence. Three dimensionalT1-weighted images data were collected magnetization-prepared rapidgradient echo imaging. Fractional anisotropy (FA) values were derived fromdiffusion tensor imaging (DTI). Different brain regions with low-frequencyfluctuations of spontaneous brain activity in newly diagnosed epilepticpatients with different response to antiepileptic drugs were examined by themean fractional low-frequency fluctuations (mfALFF) using resting-state functional MRI Then the different brain regions were chose for the region ofinterests (ROIs) to built functional connectivity maps to evaluate thedifferences of functional networks in different patients with differentresponse to drugs. Additionally, DTI technology was used to detect thedifferences of white fiber microstructures among different subjects.Results21newly diagnosed epileptic patients,13healthy controls,8drug-resistantepileptic patients and11drug withdrawal patients were recruited into thisstudy. The results showed that:1.mfALFF analysis: there were widespread mfALFF differences amongseizure uncontrolled (SUC) group, seizure controlled (SC) group, andhealthy controls throughout the bilateral cuneus lobe, bilateral lingual gyrus,right middle occipital gyrus, and right inferior frontal gyrus. The SUCpatients showed increased activity mainly in middle occipital gyrus,especially in the cuneus and lingual gyrus, while there was decreasedactivity in the middle frontal gyrus, the inferior frontal gyrus and the inferiorparietal lobule compared with the SC group and healthy controls. Wedemonstrated that the Brodmann17area showed the most significant groupdifferences in mfALFF, and the amplitude low-frequency fluctuations(ALFF) values in the region exhibited a general pattern of ALFF healthycontrols <ALFF patients <ALFF SUC patients. Moreover, by using themfALFF for the Brodmann17area as an index, SUC patients can be differentiated from the SC patients and the healthy controls with highspecificity and sensitivity.2.Functional connectivity analysis: the “default mode” was existed in allhealthy controls and patients. This mode mainly included the posteriorcingulated cortex, medial prefrontal cortex, inferior parietal lobule andinferior temporal gyrus. Compared with the healthy controls and the SCpatients, the SUC patients showed significant increased positive functionalconnectivity with the Brodmann17area in the left posterior central gyrusand the bilateral superior parietal lobule, while increased negative functionalconnectivity with the Brodmann17area in the bilateral posterior cingulatedcortex, bilateral inferior precuneus, medial prefrontal cortex, superior frontalgyrus, and middle frontal gyrus. Compared with the healthy controls, therewas only occipitotemporal junction region that showed decreased positiveconnectivity with Brodmann17area in SC patients. We also found that therewere more brain regions with abnormal functional connectivity in SUCpatients than in SC patients.3. DTI analysis: Compared with the healthy controls, the SUC patientsshowed significantly decreased fractional anisotropy values in the whitefibers under bilateral frontal cortex and the junction cortex of frontal andparietal. There was decreased fractional anisotropy value in the white fibersabove middle cingulated cortex and increased fractional anisotropy value inthe corpus callosum in the SC patients. ConclusionWe find that the abnormal high level of spontaneous brain activity in theoccipital visual cortex region, in particular, the Bromann17area,significantly correlates with the poor response to antiepileptic drugs innewly diagnosed epileptic patients. The abnormal connectivity amongBrodmann17area and superior parietal lobules and the damage of whitefiber microstructures under the frontal and the junction of frontal-parietalcortex may be involved in the formation of antiepileptic drug-resistantnetworks.更多还原