【作者】 史琦；

【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】 目的1.建立冠心病UAP患者中医证候/证候要素相对应的临床四诊信息组合及关联网络,形成四诊信息可视化网络图,探讨四诊信息在复杂网络中分布模式的特点和意义；2.发掘冠心病UAP患者证候/证候要素与临床常规理化指标及生物学指标之间的关联关系,尝试建立两者之间的对应性数据平台,探讨其组合模式在冠心病UAP中医证候判断中的作用和意义；3.建立基于血浆代谢组学的冠心病UAP患者与健康对照组之间的区分模式；筛选证候相关的特征性代谢产物,尝试建立冠心病UAP患者证候在代谢层面的识别模式。方法1.前期文献调查研究采用数据库检索及Shannon熵互信息方法筛选同冠心病UAP患者证候/证候要素密切相关的中医临床四诊信息、临床常规理化指标及反映病理生理机制的生物学指标；2.临床流行病学调查研究结合文献整理和专家调研结果制定冠心病UAP临床四诊信息采集表,完整、准确采集1576例确诊为冠心病UAP患者的临床资料；3.理化指标检测在流行病学调查过程中,全面采集411例冠心病UAP患者的149项临床常规理化指标；采用酶联免疫吸附试验方法(ELISA)检测130冠心病UAP患者及30例正常对照组血浆TNF-α、MMP-9、ET、NO、APN、sICAM-1、Hey、Ps、TAFI及HO-1水平,检测过程由专业人员负责,严格实行质量控制；4.代谢组学实验研究对45例冠心病UAP患者及15例正常人的血浆样本进行氢核磁共振(1H-NMR)检测,对其血浆中的内源性小分子及大分子代谢产物进行全面鉴定；5.冠心病UAP中医证候识别模式的建立(1)采用互信息的方法构建107项四诊信息节点之间的关联关系,采用Pajek2.0软件绘制四诊信息复杂网络的可视化图,包括不同类别节点图、k-核心网络图及抽象化图,进行网络的度分析、聚类分析、网络核心及连通性分析；(2)采用CHAID决策树及ADTree决策树方法对冠心病UAP患者中医证候／证候要素与临床常规指标及生物学指标之间的关联关系进行挖掘。采用分层交叉验证方法进行验证,以敏感性、特异性及判断正确率均大于70%视为模型成立；(3)采用SIMCA-P12.0软件中OPLS/O2PLS-DA方法进行数据分析,结合载荷图、VIP图及t检验/非参数检验结果,筛选出与冠心病UAP患者血瘀证、气虚证最可能相关的特征性代谢产物,进一步采用Statistica6.0软件进行聚类分析,验证特征性代谢产物对证候的识别效果。结果1.四诊信息复杂网络对证候的识别模式网络属性指标计算结果显示：冠心病UAP患者、冠心病合并高血压及冠心病合并糖尿病患者四诊信息网络中的107个节点度值在0-6之间；三个网络中分别有3-5个各自连通的网络,连通分量编号为1的网络最复杂。网络抽象化图显示：UAP患者的网络骨架由胸痛、耳鸣、胸闷、咳嗽、气短和五心烦热组成；合并高血压患者的网络骨架由心悸、恶心呕吐、头晕、胸痛、胸闷和畏寒肢冷组成；合并糖尿病患者的抽象化网络由胸痛、胸闷、五心烦热、咳嗽、倦怠乏力和气短组成。k-核心网络结果显示：冠心病UAP患者及合并高血压患者的网络中均有5个节点的k值为4,形成了4-核心网络；合并糖尿病患者的网络中有17个节点的k值为3,构成了3-核心网络。四诊信息网络节点分类图显示：三个网络的中心为提示气虚、阳虚和气滞的节点组合。网络中心周围依次排列了可不同程度的提示气虚、阳虚、阴虚、血瘀、气滞、脾虚、热蕴及痰浊证候的四诊信息组合。2.临床常规理化指标对证候的识别模式应用SPSS17.0CHAID算法,共形成13个证候/证候要素的识别模型。血瘀证决策树模型包括TSH,左室径MPV, DBIL, PTA, QT间期,QRS时限,ALB8个属性指标；气虚证决策树模型包括X TAL, RDW-CV, K, TSH, MONO, hs-CRP, LDL, A峰8个属性指标；阳虚证决策树模型包括D-Ⅱ聚体PDW, FT4, LP(a), Cl, PT6个属性指标；寒凝证决策树模型包括CRP, RDW-CV, AST, PT, HDL5个属性指标；痰浊证决策树模型包括D-Ⅱ聚体,Tc, MCHC, CK尿镜检RBC5个属性指标；气滞证决策树模型包括P-LCR, INR, PTA, E峰,Na, TP, MCHC7个属性指标；痰瘀互阻证决策树模型包括is-CRP, TBIL, GGT, PLT, FBG, P-R司期6个属性指标；气滞血瘀证决策树模型包括左室后壁厚度,PLT, NE%, BSA%4个属性指标；瘀热互结证决策树模型包括hs-CRP, LP(a), MONO, FIB, RDW-CV, MCV6个属性指标；气阴两虚证决策树模型包括MCH, P, MCV, EC, QRS时限5个属性指标；心阳气虚证决策树模型包括D-Ⅱ聚体,MPV,E峰,P,PT5个属性指标；阴阳两虚证决策树模型包括FT3,ALT, MONO%3个属性指标；痰热互结证决策树模型包括Cl,缩短分数,RDW-CV,血常规RBC,D-Ⅱ聚体,CK-MB, PTA, BUN8个属性指标。3.生物学指标对证候的识别模式血瘀证ADTree模型包含Ps、MMP-9、NO、sICAM-1、TAFI、Hey、HO-17个属性指标；气虚证ADTree模型包含TNF-α、NO、TAFI、sICAM-1、Hey、APN、Ps、HO-18个属性指标；阴虚证ADTree模型包含MMP-9、APN、sICAM-1、ET、HO-1共5个属性指标；阳虚证ADTree模型包含TAFI、Ps、sICAM-1、HO-1、MMP-9共5个属性指标；寒凝证ADTree模型包括sICAM-1、NO、APN、MMP-9、Ps、ET、HO-17个属性指标；痰浊证ADTree模型包括TNF-α、Ps、NO、HO-1、sICAM-1、TAFI6个属性指标；气滞证ADTree模型包括HO-1、ET、TAFI、Hey、sICAM-1、NO6个属性指标；热蕴证ADTree模型包含TAFI、TNF-α、MMP-9、Hey、Ps、HO-1、APN共7个属性指标。4.特征性代谢产物对证候的识别模式共检测到冠心病UAP患者和正常对照组血浆内源性代谢成分39种,其中小分子物质34种,大分子物质5种。CPMG及LED代谢产物的OPLS/O2PLS-DA积分矩阵图显示：冠心病UAP患者与正常对照组,冠心病血瘀证与非血瘀证,气虚证与非气虚证的分布区沿第一主成分t(1)轴方向完全分开,模型拟合度较好。特征性代谢物筛选结果显示：血瘀证的特征性代谢产物是缬氨酸和丙酮；气虚证的特征性代谢产物包括乙酰谷氨酸、赖氨酸、缬氨酸和肉碱。结论1.复杂网络的分析方法可用于冠心病UAP中医四诊信息分布模型的研究,具体包括挖掘识别证型的核心四诊信息或四诊信息群、分析疾病的基本证型、总结各证型之间的相关关系及关联程度等；2.冠心病UAP各证候／证候要素的主要生物学机制为：(1)血瘀证：以TSH降低为主的甲状腺功能异常,MPV异常的高凝状态,以PTA水平下降的出血倾向,QT间期延长及易发心律失常,以ALB降低的蛋白质代谢紊乱,DBIL增高的保护功能,以Ps降低的血小板活化功能异常,以MMP-9升高的细胞外基质代谢紊乱；代谢物角度提示脂质代谢、氨基酸代谢异常及血管内皮损伤；(2)气虚证：以K离子降低的电解质紊乱,以hs-CRP升高、RDW-CV、MONO及TNF-α水平异常的炎症反应,心室舒张早期充盈功能减退,TSH水平呈代偿性的增高,NO水平降低的内皮功能损伤；代谢物角度提示糖代谢、脂质代谢及氨基酸代谢异常；(3)阴虚证：MMP-9异常的细胞外基质代谢紊乱,APN水平降低的脂质代谢异常；(4)阳虚证：以D-Ⅱ聚体升高的高凝状态,以PDW异常及Ps降低的血小板活化功能异常及血栓形成倾向,TAFI异常的凝血纤溶功能紊乱,LP(a)升高的脂质代谢异常,以Cl离子为主的电解质紊乱；(5)寒凝证：CRP异常的炎症反应,以PT升高为主的外源性凝血功能异常,以HDL下降的脂质代谢异常,以sICAM-1水平异常的细胞间粘附状态的异常,以ET水平升高的内皮功能损伤；(6)痰浊证：D-Ⅱ聚体异常的凝血纤溶功能改变,QTc延长及易发心律失常,以MCHC上升的贫血倾向,以CK升高倾向的心肌损伤,以TNF-α明显升高的炎症状态,以Ps升高的血小板活化功能亢进；(7)气滞证：以P-LCR升高,1NR降低,PTA水平降低的出血凝血异常,左室舒张功能受损,以TP下降的蛋白质代谢紊乱及肝损害倾向,MCHC水平升高的贫血倾向,以HO-1明显减少的抗氧化应激能力下降,以ET水平异常的内皮功能紊乱；(8)热蕴证：以TAFI异常的凝血纤溶功能紊乱,以TNF-α、Hcy升高的炎症反应,以Ps明显升高的血小板活化功能亢进；(9)痰瘀互阻证：以hs-CRP异常的炎症反应,GGT水平升高的氧化应激反应,PLT升高的凝血状态,糖代谢异常；(10)气滞血瘀证：左室后壁厚度改变的心室重构,PLT升高的凝血和血栓倾向,NE%升高的炎症状态；(11)瘀热互结证：以hs-CRP升高,MONO水平异常,RDW升高的炎症反应,FIB水平升高的凝血倾向；(12)气阴两虚证：MCH异常的贫血倾向,P离子水平降低的电解质紊乱,MCV升高的可能炎症状态,以QRS时限延长的心室收缩功能下降；(13)心阳气虚证：D-Ⅱ聚体升高的凝血纤溶功能异常,PT、MPV降低的低血栓倾向,左室舒张功能受损；(14)阴阳两虚证：FT3水平明显降低的甲状腺功能异常及机体的保护性抑制状态,MONO%异常的可能炎症状态,ALT水平升高的CHD风险性增高；(15)痰热互结证：Cl离子降低的严重电解质紊乱,RDW升高的炎症反应,RBC计数增高的血液流变性质异常,CK-MB异常改变的心肌受损,PTA升高的凝血倾向。更多还原

【Abstract】 Objective1. To establish networks of symptoms from four diagnostic methods that TCM syndromes/syndrome elements corresponding from unstable angina pectoris (UAP) patients, form the visualization graphs of symptoms from four diagnostic methods complex networks, and to explore the characteristics and significance of the distribution modes of symptoms from four diagnostic methods in complex network.2. To find the relationships between the traditional Chinese medicine (TCM) syndromes/syndrome elements from UAP patients and the biological parameters, establish the corresponding data platform, explore the roles that biological parameters or their combination modes have played in TCM syndromes identification of UAP patients.3. To establish the distinguished modes between UAP patients and healthy control group based on plasma metabonomics, screen the characteristic metabolites related to the syndromes, establish TCM syndromes identification modes in the metabolic level.Methods1. Literature reviewBy the methods of database retrieval and Shannon entropy mutual information, we screened the symptoms from four diagnostic methods and biological parameters that are both closely related to TCM syndromes/syndrome elements and to the pathophysiological mechanisms from UAP patients.2. Clinical epidemiological investi’gation and researchCombinations of literatures review and experts investigation were used to develop the UAP symptoms from four diagnostic methods questionnaires. We collected the clinical data of1576UAP patients.3. Testing of physicochemical indexWe collected149clinical testing indexes from411UAP patients and tested the levels of plasma TNF-α, MMP-9, ET, NO, APN, sICAM-1, Hcy, Ps, TAFI and HO-1of130UAP patients and30healthy people by ELISA method. Professionals were responsible for the inspections and the quality controls were executed strictly.4. Metabolomics experimental studyWe detected the plasma samples of45UAP patients and15healthy people with hydrogen nuclear magnetic resonance (1H-NMR) detection, and identified the plasma endogenous metabolites from micro-molecules to macro-molecules.5. Establishment of identification modes of TCM syndromes(1) We identified useful relationships among107symptoms from four diagnostic methods by means of Distance-based Mutual Information Model (DMIM). Pajek software2.0was used to analyze the node degrees, the node core values, the connected component and the clustering coefficients of the symptoms from four diagnostic methods network. Then we drew the K-core network graphs, classified graphs according to different colors and different degrees and the abstract graphs.(2) By Chi-Square Automatic Interaction Detection (CHAID) decision tree and ADTree, we established identification modes to explain the relationships between TCM syndromes/syndrome elements of UAP patients and the biological parameters. Folds cross validations were used in this research to minimize the bias produced by random sampling of the training and test data samples. The modes were successfully formed if the sensitivity, specificity and the accuracy were all higher than70%.(3) We analyzed the metabolomics data by OPLS/O2PLS-DA method in SIMCA-P12.0software. Characteristic metabolites related to blood stasis syndrome and qi deficiency syndrome were screened based on the combination of load matrix, VIP graphs and/tests/non parametric test results. Statistica6.0software was used for further clustering analysis to validate the identification effects of the characteristic metabolites.Results1. Symptoms from four diagnostic methods complex networks to identify the syndromes Calculation results of the network attribute indexes showed that:in the networks of UAP, coronary heart disease (CHD) combined with hypertension (HT) and CHD combined with diabetes mellitus (DM) patients, the degree values of107nodes were from0to6. There were3to5connective networks respectively in the three complex networks. and the networks with connected component No.1were the most complex ones. The abstract graphs displayed that: the network of UAP patients were made up of chest pain, tinnitus, chest distress, cough, short breath and burning sensation of five centres. The network of CHD combined with HT included cardiopalmus, nausea and vomiting, dizziness, chest pain, chest distress, cold abdomen and waist. Nodes of CHD combined with DM network were chest pain, chest distress, burning sensation of five centres, cough, hypodynamia and short breath. Results of k-core networks:in the networks of UAP and CHD combined with HT, there were5nodes with the k core value4, which made4-core networks. In the networks of CHD combined with DM, there were17nodes with the k core value3, which formed a3-core network. Classification figures according to different colors showed that:the central nodes suggested the qi deficiency, yang deficiency and qi stagnation syndrome in the three networks. Around the central arranged combinations of symptoms from four diagnostic methods reflected qi deficiency, yin deficiency, blood stasis, phlegm, yang deficiency, heat deposition, qi stagnation and spleen deficiency syndromes.2. Clinical routine test indexes to identify the syndromesIdentification modes of13syndromes/syndrome elements of UAP patients were formed by CHAID method. The identification mode of blood stasis contained TSH, Left ventricular diameter, MPV, DBIL, PTA, Q-T interval, QRS and ALB. The identification mode of qi deficiency contained X TAL, RDW-CV, K, TSH, MONO, hs-CRP, LDL and A peak. The identification mode of yang deficiency contained D-Ⅱ dimer, PDW, FT4, LP(a), CI and PT. The identification mode of cold coagulation contained CRP, RDW-CV, AST, PT and HDL. The identification mode of phlegm contained D-Ⅱ dimer, QTc, MCHC, CK and urine RBC. The identification mode of qi stagnation contained P-LCR, INR, PTA, E peak, Na, TP and MCHC. The identification mode of phlegm-blood stasis contained hs-CRP, TBIL, GGT, PLT, FBG and P-R interval. The identification mode of qi stagnation-blood stasis contained LVPWT, PLT, NE%and BSA%. The identification mode of blood-heat stasis contained hs-CRP, LP(a), MONO, FIB, RDW-CV and MCV. The identification mode of qi-yin deficiency contained MCH, P, MCV, EC and QRS. The identification mode of qi-yang deficiency contained D-Ⅱ dimer, MPV, E peak, P and PT. The identification mode of yin-yang deficiency contained FT3, ALT and MONO%. The identification mode of phlegm-heat stasis contained CI, FS, RDW-CV, RBC, D-Ⅱ dimer, CK-MB, PTA and BUN.3. Biological parameters to identify the syndromesNodes of blood stasis ADTree mode were Ps. MMP-9, NO, sICAM-1, TAFI, Hcy and HO-1. Nodes of qi deficiency ADTree mode were TNF-α, NO. TAFI, sICAM-1, Hcy, APN, Ps and HO-1. Nodes of yin deficiency ADTree mode were MMP-9, APN, sICAM-1, ET and HO-1. Nodes of yang deficiency ADTree mode were TAFI, Ps, sICAM-1, HO-1and MMP-9. Nodes of cold coagulation ADTree mode were sICAM-1, NO, APN, MMP-9, Ps, ET and HO-1. Nodes of phlegm ADTree mode were TNF-α, Ps, NO, HO-1, sICAM-1and TAFI. Nodes of qi stagnation ADTree mode were HO-1, ET, TAFI, Hcy, sICAM-1and NO. Nodes of heat deposition ADTree mode were TAFI, TNF-α, MMP-9, Hcy, Ps, HO-1and APN. 4. Characteristic metabolites to identify the syndromes39endogenous metabolites had been detected, of which34were micro-molecules and5were macro-molecules. OPLS/O2PLS-DA integral matrix graphs of CPMG and LED showed that: distribution region of UAP patients and healthy people, blood stasis and non-blood stasis patients, qi deficiency and non-qi deficiency patients could be completely separated along the t(1) axis direction. The separated modes had a high fitting degree. Characteristic metabolites screening results showed that:characteristic metabolites of blood stasis were valine and acetone. Characteristic metabolites of qi deficiency were acetyl glutamic acid, Lysine, valine and carnitine.Conclusion1. Complex network analysis has contributed a lot to the study on distribution modes of symptoms from four diagnostic methods. With it, we can find the symptoms from four diagnostic methods or their groups that can identify TCM syndromes, analyze the core syndromes of disease, summarize the relationships between various syndromes and the correlation degrees.2. Main biological mechanism of TCM syndrome/syndrome elements for UAP patients(1) Main biological mechanisms of blood stasis syndrome contained thyroid function abnormalities with lower TSH; abnormal MPV and coagulation state; bleeding tendency with lower PTA level; prolongation of QT interval and being prone to cardiac arrhythmia; protein metabolic disorder with lower ALB; protection function with increased DBIL; platelet activation function abnormalities with the change of Ps; disorder in extracellular matrix metabolism with higher MMP-9; disorder in lipid metabolism, amino acid metabolism and vascular endothelium injury;(2) Main biological mechanisms of qi deficiency syndrome contained electrolyte disturbances with lower K ion; inflammatory response with higher hs-CRP and abnormal level of RDW-CV, MONO and TNF-α; ventricular early diastolic dysfunction; compensatory increase in TSH; impaired endothelial function with reducing NO; disorder in glucose, lipid and amino acid metabolism;(3) Main biological mechanisms of yin deficiency syndrome contained disorder in extracellular matrix metabolism with abnormal MMP-9; disorder in lipid metabolism with reducing APN:(4) Main biological mechanisms of yang deficiency syndrome contained coagulation state with higher D-Ⅱ dimmer; platelet activation and thrombosis tendency with abnormal PDW and reducing Ps; coagulation and fibrinolysis disorders with abnormal TAFI; disorder in lipid metabolism with higher LP(a); electrolyte disorders with Cl ion;(5) Main biological mechanisms of cold coagulation syndrome contained inflammatory responses with abnormal CRP; exogenous coagulation abnormalities with elevated PT; disorder in lipid metabolism with lower HDL; abnormal intercellular adhesion state with abnormal sICAM-1; endothelial function injury with elevated ET;(6) Main biological mechanisms of phlegm syndrome contained coagulation and fibrinolytic function change with abnormal D-II dimmer; prolongation of QTc interval and being prone to cardiac arrhythmia; anaemic tendency with rising of MCHC; myocardial injury with increased CK; inflammatory response with significant rising of TNF-a; hyper-function in platelet activation with elevated Ps;(7) Main biological mechanisms of qi stagnation syndrome contained bleeding and coagulation abnormalities with increased P-LCR, decreased INR and PTA; left ventricular diastolic dysfunction; disorder in protein metabolism and liver damage tendency with decrease of TP; anaemic tendency with elevated levels of MCHC; disability in antioxidant with significantly reduced HO-1; disorder in endothelial function with abnormal ET;(8) Main biological mechanisms of heat deposition syndrome contained coagulation and fibrinolysis disorders with abnormal TAF1; inflammatory response with rising of TNF-a and Hcy; hyper-function in platelet activation with elevated Ps;(9) Main biological mechanisms of phlegm-blood stasis syndrome contained inflammatory response with abnormal hs-CRP; oxidative stress reaction with rising of GGT; coagulation condition with higher PLT; disorder in glucose metabolism;(10) Main biological mechanisms of qi stagnation-blood stasis syndrome contained ventricular remodeling with change of left ventricular posterior wall thickness; coagulation and thrombosis tendency with elevated PLT; inflammatory state with higher NE%;(11) Main biological mechanisms of blood-heat stasis syndrome contained inflammatory state with rising of hs-CRP and RDW and abnormal level of MONO; coagulation tendency with higher FIB;(12) Main biological mechanisms of qi-yin deficiency syndrome contained anaemic tendency with abnormality of MCH; electrolyte disturbances with reduced P ion; possible inflammatory state with rising of MCV; ventricular systolic dysfunction with prolongation of QRS;(13) Main biological mechanisms of qi-yang deficiency syndrome contained coagulation and fibrinolytic function change with higher D-H dimmer; thrombotic tendency with decrease in PT and MPV; left ventricular diastolic function;(14) Main biological mechanisms of yin-yang deficiency syndrome contained abnormal thyroid function and inhibition state with significantly reduced FT3; possible inflammatory state with MONO%anomalies; rising risk of CHD with higher ALT;(15) Main biological mechanisms of phlegm-heat stasis syndrome contained severe electrolyte disturbances with reduced Clion; inflammatory reaction with higher RDW; abnormal blood rheological function with increased RBC; myocardial damage with CK-MB abnormalities; clotting tendency with increased PTA.更多还原