摘要目的 系统评价美金刚胺治疗中重度阿尔茨海默病(AD)的疗效与安全性. 方法检索中国期刊全文数据库(CJFD)、美国国立医学生物信息中心PubMed数据库(PubMed)、美国国立.卫生院临床试验数据库(Clinicaltrials.gov)、美金刚胺生产商麦氏(Merz)、灵北(Lundbeck)和森林实验室(Forest laboratories)的出版物以及森林实验室临床试验中心(Forest clinical trials registry)关于美金刚胺治疗中莺度AD的随机、双肓、安慰剂对照研究.对符合条件的研究结果用Revman 5.0软件进行Meta分析.以美金刚胺组相对于安慰剂组在总体功能、认知功能、日常生活能力和精神行为4个方面的加权均数差(WMD)为指标进行疗效评价,以两组在不良事件和严重不良事件两方面的相对危险度(RR)为指标进行安全性评价. 结果共纳入4项研究,1683名患者被随机分配,其中美金刚胺组849名,安慰剂组834名.Meta分析结果显示,美金刚胺组相对于安慰剂组在总体功能(CIBIC-PIus)、认知功能(SIB)、日常生活能力(ADCS-ADL19)和精神行为(NPI)4个方面的WMD分别为-0.29(P<0.00001)、2.79(P<0.00001)、1.03(P=0.002)、-2.85(P=0.002);美金刚胺组相对于安慰剂组在不良事件和严重不良事件的RR值分别为1.01(P=0.66)和0.98(P=0.90).除认知功能(P=0.05)外,4项研究在其他方面均未见明显统计异质性. 结论美金刚胺同安慰剂相比可以改善患者总体功能、认知功能、日常生活能力及精神行为症状,美金刚胺的安伞性和耐受性同安慰剂相比差异无统计学意义.

abstractsObjective To analyze the efficacy and safety of memantine in treating moderate to severe Alzheimer's disease(AD) from published and unpublished randomized clinical trials. Methods A search for randomized, placebo-controlled clinical trials of memantine in treating moderate to severe AD using Chinese journal full-text database (CJFD), PubMed and Clinical trials, gov was carried out. The clinical trials registry and press releases from Merz, Lundbcck and Forest Laboratories were examined. A meta-analysis of included clinical trials was performed with Revman 5.0 software. Efficacy was assessed with the weighted mean difference (WMD) of global status, cognition, activity of daily living and behavior. Safety and tolerability were assessed with the relative risk (RR) of the adverse event (AE) and the severe adverse event (SAE). Results A total of 4 clinical trials were included in the meta-analysis, of which 1683 patients were randomly assigned (849 on memantine;834 on placebo). Patients'records were pooled. Overall WMD were -0.29 (P<0. 00001) for the global domain (CIBIC-Plus), 2. 79 (P<0. 00001) for the cognition domain (SIB), 1. 03 (P = 0. 002) for the activity of daily living domain (ADCS-ADL19) and -2. 85 (P = 0. 02) for the behavior domain (NPI). The overall relative risk of memantine versus placebo were 1.01 (P = 0.66) for AE and 0.98 (P = 0. 90) for SAE. There was no sign of heterogeneity except for the cognition domain (P=0. 05). Conclusions There is a significant effect in favour of memantine treatment in four domains: global status, cognition, activity of daily living and behavior. The safety and tolerability between memantine and placebo do not have statistical differences.