【作者】 王娟；

【导师】 王伟；

【作者基本信息】 北京中医药大学 ， 中西医结合基础， 2012， 博士

【摘要】 慢性心力衰竭中医证候生物学诊断模式及代谢组学研究研究目的探讨慢性心衰证候分布组合特点及影响因素,对慢性心衰患者进行多项生物学指标检测和代谢组学核磁波谱分析,将一般统计学方法与数据挖掘方法相结合,在宏观和微观层面挖掘与疾病及证候相关的代谢物,并与代谢路径结合,提取敏感性和特异性强的证候特征指标群,为慢性心衰证候客观化诊断及中药复方干预提供参考依据。研究方法1.临床横断面研究：全面采集630例慢性心衰患者人口学资料、现病史信息、中医四诊信息及诊断结果,对证候属性进行判别,描述慢性心衰证候分布特点和组合规律并对影响因素进行Logistic回归分析。2.慢性心衰证候生物学诊断模式研究：检测符合纳入标准的100例慢性心衰患者临床生化指标及超声心动图结果,运用数据挖掘方法对各证候生物学指标进行相关分析和特征提取,在此基础上用神经感知器方法构建各证候生物学诊断模式。3.慢性心衰患者血浆代谢组学研究：对比46例冠心病慢性心衰患者和15例健康人1H-NMR的血浆代谢物谱,寻找与疾病和证候相关的标志性代谢物群,将其与代谢通路结合,为慢性心衰及证候的生物学基础研究提供新的方法和思路。研究结果(一)临床横断面流行病学研究结果1.630例慢性心力衰竭患者以男性居多,构成比为58.25%。55岁以下患者占14.92%,55岁到70岁之间患者占35.24%,70岁以上患者占49.84%。纳入患者的原发病为冠心病、高心病和扩心病,其中冠心病患者占61.95%,高心病患者占29.22%,扩心病患者占8.83%。心功能I级患者占2.22%,心功能Ⅱ级患者人数占24.92%,心功能Ⅲ级患者占51.11%,心功能Ⅳ级患者占21.75%。合并症按出现频次高低依次为高血压,心律失常,糖尿病,脑血管意外,高血脂症等。2.630例慢性心衰患者出现8种证候,其中气虚证为基本证候,其次为血瘀证和水停证,再次为阴虚证、痰浊证、阳虚证、热证和气滞证。证候组合方式有单一证型、证组合、三证组合、四证组合、五证组合和六证组合型。气虚证为最常见的单证型,占45.16%；二证组合以气虚血瘀最多,占53.98%；三证组合以气虚血瘀水停最多,占35.22%；四证组合以气虚血瘀阳虚水停最多,占23.9]%；五证组合以气虚血瘀阳虚水停兼夹痰浊证最多,占58.70%；六证组合仅1例,为气虚血瘀阳虚水停兼夹阴虚痰浊证。其中以三证组合方式为最常见,其次是四证组合和两证组合。性别、年龄、原发病因、心功能和并发症等可影响慢性心衰证候分布。(二)慢性心衰证候生物学诊断模式研究1.经T检验、ROC曲线分析和Logistic回归筛选出的气虚证指标为BASO、RBCHGB等；经T检验、ROC曲线分析和Logistic回归筛选出的血瘀证指标为：LYMPH、LYMPHL、MONO等；经T检验、ROC曲线分析和Logistic回归筛选出的痰浊证指标为：WBC、NEUT、LYMPH等；经T检验、ROC曲线分析和Logistic回归筛选出的水停证指标为：WBC、BASO、NEUT1等；经T检验、ROC曲线分析和Logistic回归筛选出的阴虚证指标为：WBC、MONO、EO等；经T检验、ROC曲线分析和Logistic回归筛选出的阳虚证指标为：RBC、HGB、MCH等。2.运用神经网络MLP(多层感知器)方法对慢性心衰各证候的理化指标群进行建模处理,血瘀证测试样本和预测样本中的准确率分别为75.4%和82.4%；痰浊证测试样本和预测样本中的准确率分别为79.3%和83.3%；水停证测试样本和预测样本中的准确率分别为81.8%和81.0%；阴虚证测试样本和预测样本中的准确率分别为83.0%和82.4%；阳虚证测试样本和预测样本中的准确率分别为91.1%和90.5%。(三)慢性心衰患者血浆代谢组学研究1.应用主成分分析和正交偏最小二乘两种模式识别方法对慢性心衰患者和健康人血浆1H-NMR代谢组进行了判别分析,结果显示慢性心衰患者与健康对照组图中主成分积分值基本集中分布于椭圆形散点图(95%置信区内)的4个区域,无明显交叉和重叠,表明慢性心衰患者与健康人之间存在血浆成分代谢产物谱的显著差异。其中慢性心衰患者组乳酸、肌酸、脯氨酸等含量明显高于健康对照组,而组氨酸、甘氨酸、谷氨酸等含量低于健康对照组。2.慢性心衰血瘀证患者与非血瘀证患者CPMG谱图差异代谢物包括组氨酸、甘氨酸、缬氨酸等下降,乳酸、丙氨酸、丙酮酸等升高；LED图谱差异代谢物主要为高密度脂蛋白下降、低密度脂蛋白和极低密度脂蛋白的升高。3.慢性心衰痰浊证患者血浆CPMG图谱差异代谢物包括丙氨酸、葡萄糖、胆碱等下降,乳酸升高。血浆LED图谱差异代谢物包括高密度脂蛋白、低密度脂蛋白、极低密度脂蛋白和不饱和脂肪酸下降。4.慢性心衰水停证患者血浆CPMG图谱差异代谢物包括葡萄糖、胆碱、TMAO等下降,乳酸、丙氨酸、乙酸等升高。血浆LED图谱差异代谢物包括低密度脂蛋白、极低密度脂蛋白和不饱和脂肪酸下降。5.慢性心衰阴虚证患者血浆CPMG图谱差异代谢物包括葡萄糖、缬氨酸、脯氨酸等降低,乳酸、糖蛋白升高。血浆LED图谱差异代谢物为低密度脂蛋白、极低密度脂蛋白和糖蛋白的升高。6.慢性心衰阳虚证患者血浆CPMG图谱差异代谢物包括葡萄糖、脂蛋白降低,乳酸、丙氨酸、谷氨酸、丙酮酸升高。血浆LED图谱差异代谢物包括谷氨酸、糖蛋白升高和高密度脂蛋白、低密度脂蛋白和极低密度脂蛋白的降低。结论1.通过研究,认识到慢性心衰发病以老年人为主的特点及病因分布情况,结合指南精神,提醒我们全方位、多靶点介入以控制危险因素并积极治疗高危人群原发病,有助于预防发展为心力衰竭,或延缓慢性心力衰竭发病进程。慢性心衰证候分布以气虚、水停、血瘀、痰浊、阴虚和阳虚为常见证候,气虚血瘀水停为关键病机,探讨慢性心衰的中医证型与心功能分级、中医药如何防治心力衰竭,减轻并发症,降低患者再住院率和死亡率有重要意义。2.贫血倾向和肾功能损害是慢性心衰气虚证可能的生物学基础；红细胞体积改变,血液流变学改变,单核-巨噬细胞介导的炎症反应等为血瘀证可能的生物学基础；血脂紊乱,炎性反应、高尿酸水平等为痰浊证可能的生物学基础；炎性反应,高尿酸,补体系统抑制等可能为水停证的内在生物学基础；炎性反应,贫血倾向,血脂紊乱等为阴虚证可能的生物学基础；贫血倾向,肝功下降,免疫激活等为阳虚证可能的内在生物学机制。本研究在运用ROC曲线和逻辑回归分析后得到的理化指标相互印证和补充并阐述其生物学意义后,对慢性心衰主要证候相关理化指标群进行人工网络建模处理,在测试和预测样本中都得到较高准确率,因此在证候的生物学基础研究过程中应将多种数据发掘方法互参以揭示证候本质所在。3.运用代谢组学方法对慢性心衰患者和健康人血浆分析结果显示OPLSDA模式识别效果优于PCA,可以降低临床样本不均一性,提高模式识别的分类效果,在临床数据分析中值得推广。慢性心衰与能量代谢、氨基酸代谢,糖代谢和脂代谢紊乱关系密切,反映慢性心衰状态下能量代谢紊乱、神经内分泌失调、免疫功能减弱、血液微循环障碍及抗损伤修复减退的病理机制。此外,慢性心衰各证候具有不同的代谢模式,所建模型能够将各组区分,其中小分子代谢物较大分子代谢物差异更明显,表明代谢组学技术及方法可通过代谢模式实现中医证型区分,为更好地将代谢组学方法用于疾病／证候的判别分析做了有益的探索。更多还原

【Abstract】 Object ivesThe purpose of the study of patients with chronic heart failure was to investigate the complicated TCM syndromes, in which the biological indexes used to build the diagnostic pattern for each syndrome based on data mining methods were adopted. We further analyzed the metabolites closely related to chronic heart failure and different syndromes by1H-NMR techniques, so as to provide new ideas in syndrome diagnosis, prevention and treatment of heart failure with TCM formulas.Methods1.Clinical cross-sectional study:630patients were enrolled and we discriminated syndromes under the criteria based on the demographic data, illness history as well as TCM four diagnostic information collected. Complicated TCM syndromes and related factors were described.2.Biological diagnostic pattern study:100patients with CHF were included and their physical and chemical indexes were analyzed by T-test, ROC curve and logistic regression to build the diagnosis pattern by means of artificial networks.3.The metabolic biomarker study:Plasma samples of46patients with chronic heart failure caused by coronary heart diseases and15healthy people were experimented by1H-NMR to find the metabolites related to chronic heart failure and different syndromes, PCA and OPLSDA methods were used and the results were compared.Results1.The male patients accounted for58.25%in630patients with CHF. Patients above70years occupied49.84%.Coronary heart diseases is the leading cause of chronic heart failure. Patients with NYHA cardiac function of grade III took up51.11%.The most common complications are hypertension, arrhythmia, diabetes diseases, cerebral vascular accident, hyperlipemia etc. There were8kinds of syndromes, qi deficiency was the most, followed by blood stasis and water retention, phlegm turbid, yin deficiency, yang deficiency, qi stagnation, heat. There were single syndrome to six syndrome combinations. Qi deficiency was the most common syndrome in single type, accounting for45.16%. Qi deficiency combined with blood stasis was the top two syndrome combination, accounting for53.98%. Qi deficiency and blood stasis combined with water retention was the top three syndrome combination, accounting for35.22%. Qi deficiency and blood stasis and water retention combined with Yang Deficiency syndrome was the top four syndrome combination, accounting for29.31%. Qi deficiency and blood stasis and water retention combined with yang deficiency as well as phlegm turbid was top five syndrome combination, accounting for58.70%. Three syndromes combination was the most common type. Sex, age, cause of illness, heart function and complications were related to the development of CHF.2. Characteristic indexes of qi deficiency includes:BASO、RBC、HGB etc; Characteristic indexes of blood stasis includes:LYMPH、LYMPHL、MONO etc; Characteristic indexes of phlegm turbid includes:WBC、NEUT、LYMPH etc; Characteristic indexes of water retention includes:WBC、BASO、NEUT1etc; Characteristic indexes of yin deficiency includes:WBC、 MONO、EO etc; Characteristic indexes of yang deficiency includes:RBC、HGB、MCH etc. The accuracy of diagnosis pattern used by MLP was higher than80%on average for each syndrome in both training samples and validation samples.3.PC A and OPLSDA methods were used to analyze the distinguished metabolites between the patients and healthy people, including Lac, Cr, Tyr etc, and the OPLSDA methods had better classification. The characteristic metabolites of the blood stasis syndrome were as follows: Lac, Ala His etc increased while Gly,Val, Glu etc reduced; The characteristic metabolites of the phlegm turbid syndrome were as follows:Lac increased while Ala, Glu, Cho, Pro, Gly, HLDL, VLDL\LDL, UFA reduced; The characteristic metabolites of the water retention syndrome were as follows:Lac, Ala, N-Ac, O-Ac increased while Glu, Cho, TMAO, Pro, VLDL\LDL, UFA reduced; The characteristic metabolites of the yin deficiency syndrome were as follows:Lac, N-Ac, O-Ac, VLDL\LDL increased while Glu, Val, Pro, Ala, Carnitine reduced; The characteristic metabolites of the yang deficiency syndrome were as follows:Lac, Ala, N-Ac, Glutamate increased while Glu, HLDL, VLDL\LDL reduced.Conelusion1.More attention should be paid to the old in prevention and treatment of CHF. Getting rid of the risk factors, controlling the complications and avoiding bad habits play an important role in reducing the mortality and re-hospitalization rate of chronic heart failure patients. Qi deficiency, water retention, blood stasis, phlegm turbid, yin deficiency and yang deficiency are the basic syndromes in this disease. Qi deficiency and blood stasis combined with water retention is the key pathogenesis, which can help clinical doctors get a better understanding of the syndrome differentiation in the treatment of CHF.2.Qi deficiency syndrome related to anemia and decreased renal function; Blood stasis syndrome related to blood cell volume changes, inflammatory response etc; Phlegm turbid syndrome related to lipids disorder, inflammatory response, high UA level etc; Water retention syndrome related to inflammatory response, high UA level, immune activation etc. Yin deficiency syndrome related to inflammatory response, anemia, lipids disorder etc. Yang deficiency syndrome related to anemia, reduced liver function, immune activation etc. As far as the diagnosis model for each syndrome being concerned. We should make full use of all kinds of data mining methods to do the TCM syndrome research.3.The OPLSDA methods have great potential to distinguish the metabolic biomarkers, better than PCA in this study, which should be widely applied in clinical studies. We figured out the metabolic patterns of patients with CHF, related to the glucose, protein, lipid metabolism as well as energy supplication, which reflect the patho-physiological mechanisms of the diseases, such as nerve-endocrine disorders, immune dysfunction, microcirculation disorder and so on. Different metabolites were found in each TCM syndrome of CHF, which could be obviously seen from the small molecular metabolites, provided new ideas of syndrome diagnosis as well as the prevention and treatment of chronic heart failure by Traditional Chinese Medicine.更多还原