【作者】 郝绍瑞；

【导师】 李兰娟；

【作者基本信息】 浙江大学 ， 内科学， 2011， 博士

【摘要】 据世界卫生组织报道,全球约20亿人曾感染过HBV,其中3.5亿人为慢性HBV感染者,每年约有100万人死于HBV感染所致的肝衰竭、肝硬化和HCC。我国现有的慢性HBV感染者约9300万人,其中慢性乙型肝炎患者约2000万例。每年各种原因引起的终末期肝病800万,死于肝病约50万。肝衰竭过程中发生很多内源性物质的积聚,比如胆红素,氨,谷氨酰胺,乳酸,芳香族氨基酸,自由脂肪酸,苯酚,硫醇,苯二氮类以及促炎症细胞因子。由于肝细胞大量急剧坏死,肝衰竭患者病情危重,预后极差,内科综合治疗患者病死率高达60～80%。急性肝衰竭的标准治疗方法是原位肝移植,但由于供肝的短缺,很多肝衰竭病人在肝移植等待过程中死亡。人工肝支持系统(ALSS)是一套独特有效的肝衰竭救治方案,极大地降低了肝衰竭患者病死率。急性、亚急性重型肝炎治愈好转率由20.0%提高至80.0%,慢性重型肝炎由5.6%提高至48.8%。然而,临床上具有相同生化指标的病人却有着不确定的预后,有些病人会治愈,有些病人则死亡。至今仍没有令人满意的预后预测指标。同时,人工肝治疗前后血浆代谢物水平发生显著变化,至今还没有在代谢组学水平研究人工肝治疗效果的研究。本研究使用超高效液相色谱-质谱技术研究不同预后以及人工肝治疗前后的慢加急(亚急)性乙肝肝衰竭患者的血浆代谢物差异。本研究选择进行了2次以上非生物型人工肝支持系统治疗的慢加急(亚急)性乙肝肝衰竭患者50例(共进行296例次人工肝支持治疗),取患者血浆进行超高效液相色谱联合四级杆飞行时间质谱的代谢组学分析,数据经Masslynx和Simca-P进行归一化、主成份分析(PCA)和正交最小二乘分析(OPLS)计算。使用Simca-P+12.0构建了可以预测慢加急(亚急)性乙肝肝衰竭预后的代谢组学正交最小二乘模型(OPLS),该模型基于第一次人工肝治疗前所采集的血浆样本,解释能力和预测能力优秀(R2Y=0.943.Q2=0.913),ROC曲线下面积0.968(95% CI [0.951，0.985]),明显优于MELD(the Model for End-stage Liver Diseases)评分的预测能力(0.737,95% CI [0.578,0.896])。鉴定出了三类十三种生物标记物：溶血卵磷脂(lysophosphatidylcholine, LPC),初级脂肪酰胺(primary fatty acid amides, FAA)和结合型胆汁酸(conjugated bile acids)。LPC和结合型胆汁酸是病人好转的保护因素,而脂肪酰胺是病人好转的危险因素。本模型的预测截断值是0.196,在这一点模型具有对好转组和未好转组的最佳区分,模型的敏感度95%,特异度87%。这个基于代谢轮廓分析的代谢组学模型不仅提供了对慢加急(亚急)性乙肝肝衰竭预后的优良区分和预测,而且可提供潜在肝移植需求的早期和精确预警。分析了非生物型人工肝支持系统治疗慢加急(亚急)性乙肝肝衰竭前后血浆代谢组学的变化。PCA图分析发现治疗后代谢轨迹基本一致,OPLS模型可以很好的区分治疗前后,模型经CV-ANOVA检验极显著(P=8.7×10-10)。根据VIP得分由高到低,鉴定出8种生物标记物。PCA图中,预后不同的病人分别集中在X轴两侧分组明显,而人工肝治疗前后的变化主要表现在纵轴上(第二主成份)由此推断除了人工肝治疗,病人本身的状态也是预后的决定性因素。在治疗前后差异最大的物质中,溶血卵磷脂在治疗后升高,甘氨鹅脱氧胆酸在治疗后降低。这种治疗效果在连续3次人工肝治疗中有相同的趋势。本研究阐明了非生物型人工肝支持系统在小分子层面对患者内环境的改善作用,同时分析了治疗前后的生物标志物在多次人工肝治疗过程中的动态变化过程,并且分析了这些标志物在不同预后患者的血浆中的动态变化的异同。更多还原

【Abstract】 According to the report of WHO,2 billion people have been infected by HBV globally, of whom 350 million is chronic HBV infection. about I million people die of HBV infection induced liver failure, liver cihhrosis and HCC each year. In China, chronic HBV infection by now is about 93 million, of whom about 20 million is chronic HBV hepatitis patients. End stage liver diseases induced by various causes each year is about 8 million, and 500 thousand of whom died of liver diseases. During liver failure, there are cumulations of a lot of endogenous substances, such as bilirubin, ammonium, glutamine, lactate, aromatic amino acids, free fatty acid, phonel, mercaptans, benzodiazepines and proinflammatory cytokines. Liver failure is a severe sympotom with very poor prognosis for the reason of massive necrosis of hepatocytes, the motality of intensive medicine care is as high as 60%-80%. The standard treatment of acute liver failure is orthotopic liver transplantation, however, because of limited donor livers, lots of patients died while waiting on the transplantation list. Artificial liver support system (ALSS) is an unique and effective liver failure cure plan, and profoundly decreases the motality of liver failure patients. With ALSS, recovery rate of acute, subacute severe hepatitis improved from 20% to 80%, chronic severe hepatitis from 5.6% to 48.4%. Nevertheless, clinically patients with similar biochemical indices had very different outcome, some recovery while others die. To date, there is still no satisfing predictive index. ALSS cause tremendous changes at the metabolites level, however, there is no study on the effect of ALSS at the metabolomic level yet. This study applies ultraperformance liquid chromotography- mass spectromatry technique to study the difference of plasma metabolite between different outcomes of acute on chronic HBV induced liver failure patients and between before and after the ALSS treatment.50 AOCLF patients were recruited each of whom accepted at least 2 sessions of ALSS (a total of 296 sessions), and the serum was analysed by UPLC Q-TOF MS. Normalization, PCA and OPLS analysis were performed by Masslynx and Simca P software.An metabolomic OPLS model was constucted using Simca-P+12.0 for prognosis of ACOLF. This model was based on the samples collected right before the first ALSS treatment rather than later ones because of comparable model efficacy and earlier time point. The concordance statistics of our model was 0.968 (95% CI [0.951.0.985]) which is superior to that of the MELD score (0.737,95% CI [0.578,0.896]). Three categories of markers were identified:lysophosphatidylcholine, primary fatty acid amides and conjugated bile acids. Lysophosphatidylcholine and conjugated bile acids were protect factors of living and primary fatty acid amides were risk factors. The cut-off point of predictive value from our model was greater than or equal to 0.196, which was the best discriminant of recovery and non-recovery group, with sensitivity 95% and specificity 87%. Our metabolomic model based on plasma UPLC-MS profile analysis provided not only excellent discrimination and prediction of the prognosis of HBV induced acute-on-chronic liver failure but also early and precise warning of liver transplantation.The serum metabolomic change due to ALSS treatment of ACOLF is analysed. PCA plot indicated almost identical metabolic track after ALSS. OPLS model descriminated the treatment effect perfectly and p value of the model by CV-ANOVA test was 8.7×10-10.8 biomarkers were identified according to the descending order of VIP. In PCA plot, prognosis was the main grouping factor, while the change according to ALSS was mainly showed on the y axies (the 2nd PC), indicating the status of patients is another deciding factor of prognosis besides ALSS. Among the most different substances, LPC level elevated through the treatment, GCDCA level decreased though the treatment. This effect was constant in consective three ALSS treatments. This result clarified the effect of ALSS in improving the internal environment on small molecule level; at the same time, the dynamic processes of biomarkers during multi-ALSS treatment were analysed. Also, we analysed the corelations between these dynamic changes of bimarkers and the pronosis.更多还原