乙型肝炎肝硬化并发肝性脑病患者血氨水平与认知障碍关系研究*

目的 研究乙型肝炎肝硬化并发肝性脑病(HE)患者血氨水平变化与认知障碍程度的关系。方法 2018年3月～2020年9月我院诊治的乙型肝炎肝硬化并发HE患者94例,根据临床和蒙特利尔认知评估量表(MoCA)评分评估认知障碍程度。常规检测血氨,计算计算胰岛素抵抗指数(HOMA-IR)。采用Logistic多因素回归模型分析发生昏迷的影响因素,建立预测模型。应用MedCalc1 5.1统计学软件绘制受试者工作特征曲线(ROC),计算曲线下面积(AUC),判断模型预测发生昏迷的效能。结果 在本组94例患者中,发生昏迷32例(34.0%);昏迷组血氨水平为(120.7±18.2)μmol/L,显著高于轻中度认知障碍者【(76.4±18.3)μmol/L,P<0.05】;昏迷组MoCA评分为(13.7±1.2)分,显著低于轻中度认知障碍者【(18.2±2.9)分,P<0.05】;昏迷组视空间执行能力、命名、注意、语言、抽象、延迟记忆和定向力各维度评分分别为(1.5±0.6)分、(1.6±0.7)分、(2.1±1.0)分、(1.7±0.6)分、(0.9±0.3)分、(2.0±0.8)分和(4.9±0.8)分,显著低于轻中度认知障碍者【分别为(2.7±1.0)分、(2.2±0.9)分、(3.4±1.2)分、(2.5±0.7)分、(1.5±0.4)分、(4.4±1.1)分和(5.4±1.0)分,P <0.05】;乙型肝炎肝硬化并发HE患者血氨水平与MoCA评分呈显著负相关(r=-0.511,P<0.001);Logistic多因素分析结果显示血氨(β=-0.529,95%CI=0.432～0.803,P<0.001)、高敏C反应蛋白 (β=-0.345,95%CI=0.518～0.968,P=0.030)和胰岛素抵抗指数(β=-0.761,95%CI=0.289～0.755,P=0.002)是患者发生昏迷的独立影响因素;根据Logistic多因素分析结果建立预测模型P=1/【1+e^{(0.242+0.980X1+0.529X2+0.345X3+0.761X4)}】,经受试者工作曲线(ROC)分析显示该模型判断患者发生昏迷的曲线下面积(AUC)为0.843(SE=0.042,95%CI=0.760～0.926,P<0.001),其敏感度为0.806,特异度为0.781。结论 乙型肝炎肝硬化并发HE患者认知功能与血氨水平密切相关,监测血氨水平变化有助于预测发生昏迷的风险。

Serum ammonia level changes and its correlation to cognitive impairment in patients with hepatitis B cirrhosis and hepatic encephalopathy

Objective The aim of this study was to investigate serum ammonia level changes and itscorrelation to cognitive impairment(CI) in patients with hepatitis B cirrhosis and hepatic encephalopathy(HE).Methods There were 94 patients with hepatitis B liver cirrhosis and HE encountered in our hospital between March 2018 and September 2020,the cognitive impairment was estimated by Montreal cognitive assessment scale(MoCA) and homeostasis model assessment-insulin resistance(HOMA-IR) wascalculated.The Logistic regression model was applied to analyze the risk factors impacting cognitive impairment,and the area under receiver operating characteristic(AUC) was used to predict the occurrence of hepatic coma.Results There were 32 cases of hepatic coma out of 94 patients(34.0%) in our series;the blood ammonia level in patients with hepatic coma was(120.7±18.2) μmol/L,significantly higher than (76.4±18.3) μmol/L,P<0.05],and the total score of MoCA was(13.7±1.2),significantly lower than (18.2±2.9),P<0.05] in patients with mild to moderate CI;the scores of visuospatial executive ability,naming,attention,language,abstraction,delayed memory and orientation were(1.5±0.6),(1.6±0.7),(2.1±1.0),(1.7±0.6),(0.9±0.3),(2.0±0.8)and(4.9±0.8),all significantly lower than (2.7±1.0),(2.2±0.9),(3.4±1.2),(2.5±0.7),(1.5±0.4),(4.4±1.1) and(5.4±1.0),respectively,P<0.05] in patients with mild to moderate CI;there was a significant negative correlation between serum ammonia levels and total MoCA scores in patients with hepatitis B cirrhosis and HE(r=-0.511,P<0.001);the Logistic multivariate analysis showed that blood ammonia(β=-0.529,95% CI=0.432-0.803,P<0.001),serum high sensitivity C-reactive protein levels(β=-0.345,95% CI=0.518-0.968,P=0.030)and HOMA-IR(β=-0.761,95% CI=0.289-0.755,P=0.002) were the independent influencing factors for patients with hepatic coma;the performance of our established prediction model:P=1/1+e^{(0.242+0.980X1+0.529X2+0.345X3+0.761X4})] was exchellent with the AUC of 0.843(SE=0.042,95% CI=0.760-0.926,P<0.001),and the sensitivity and specificity of 0.806 and 0.781,respectively in predicting the occurrence of hepatic coma.Conclusion The cognitive functions of patients with hepatitis B cirrhosis and HE is closely related to blood ammonialevels,and it might be helpful to predict the risk of hepatic coma by monitoring blood ammonia levels.