摘要：

目的 探讨帕金森病与脑肿瘤发病风险的相关性。方法 以帕金森病，肿瘤和癌症，以及PD，Parkinson's disease，tumor，cancer，neoplasm 等中英文检索词，计算机检索1965 年1 月-2016 年7 月美国国立医学图书馆生物医学信息检索系统、Web of Science、荷兰医学文摘、Cochrane 图书馆，以及中国知网中国知识基础设施工程、万方数据库、维普数据库等数据库收录的关于帕金森病与脑肿瘤发病风险的前瞻性队列研究或病例对照研究，同时辅助手工检索，采用Newcastle-Ottawa 量表和Stata 12.1 软件进行文献质量评价和Meta 分析。结果 共获得1832 篇文献，经剔除重复和不符合纳入标准者，最终纳入11 项临床研究共350 632 例帕金森病患者。Meta 分析显示：帕金森病患者脑肿瘤发病风险高于健康人群（OR = 1.370，95%CI：1.120 ~ 1.690；P = 0.003），剔除2 篇低质量文献后结论仍一致（OR = 1.360，95%CI：1.080 ~ 1.720；P = 0.008）；根据发病顺序和地域分层分析，仅帕金森病后脑肿瘤发病风险高于健康人群（OR = 1.430，95%CI：1.120 ~ 1.830；P = 0.004），欧洲（OR = 1.420，95%CI：1.290 ~ 1.560；P = 0.000）和亚洲的中国台湾地区（OR = 2.590，95%CI：1.730 ~ 3.880；P = 0.000）帕金森病患者脑肿瘤发病风险均高于健康人群。漏斗图以及Begg 法（P = 0.583）和Egger 法（P = 0.985）证实发表偏倚较小。结论 帕金森病患者脑肿瘤发病风险增加。

关键词: 帕金森病, 脑肿瘤, Meta分析

Abstract:

Objective To assess the correlation between Parkinson's disease (PD) and brain tumor. Methods Taking Parkinson's disease, tumor, cancer in Chinese, and PD, Parkinson's disease, tumor, cancer, neoplasm in English as the key words, prospective cohort studies and case -control studies on relation between PD and brain tumor were searched by using PubMed, Web of Science, EMBASE/SCOPUS, Cochrane Library, China National Knowledge Infrastructure (CNKI), Wanfang Data, and VIP database from January 1965 to July 2016, assisted by manual searching. Quality assessment and Mete-analysis were made by using Newcastle-Ottawa Scale (NOS) and Stata 12.1 software. Results A total of 11 studies with a total number of 350 632 PD patients were included in the overall analysis after excluding duplicate ones and those which did not meet the inclusion criteria from 1832 articles. Meta-analysis showed that compared with healthy people, PD patients had an increased risk of brain tumor (OR = 1.370, 95%CI: 1.120-1.690; P = 0.003), and the result was consistent after excluding 2 low-quality articles (OR = 1.360, 95% CI: 1.080-1.720; P = 0.008). Further stratified analyses according to disease onset time and regional difference showed that brain tumor occurrence in PD patients was significantly higher than healthy people only after the diagnosis of PD (OR = 1.430, 95% CI: 1.120-1.830; P = 0.004). The geographical subgroup analyses showed a higher risk of brain tumor among PD patients in Europe (OR = 1.420, 95%CI: 1.290-1.560; P = 0.000) and Taiwan area of China in Asia (OR = 2.590, 95% CI: 1.730-3.880; P = 0.000) compared with healthy people. Funnel plot, Begg test (P = 0.583) and Egger test (P = 0.985) showed there was no bias. Conclusions PD patients have a higher risk of brain tumor.

Key words: Parkinson disease, Brain neoplasms, Meta-analysis