【作者】 林锦彦；

【导师】 陈清； 傅传喜；

【作者基本信息】 南方医科大学 ， 流行病学与卫生统计学， 2014， 博士

【摘要】 2009年席卷全球的新型甲型H1N1流感(甲流)给全球带来巨大社会和经济负担。由于病毒未发生关键性变异、人群的隐性感染及大规模疫苗接种等使得甲流在人群中的流行逐渐减弱。2013年来,其在我国人群中的阳性分离率又开始上升,我们应当警惕甲流因变异或重组产生致病性更强的毒株。迄今为止,人类仍无法有效预测与控制甲流的持续性变异,因此对甲流开展分子流行病学监测及抗原变异研究可为防控甲流、降低疾病负担和调整疫苗策略提供参考依据。甲流可引起全球范围的大流行或季节性流行,每次流行后对人群会造成不同程度的死亡负担,准确评估流感流行所带来的死亡负担,对全面了解甲流的健康危害程度、制定合理的预防控制措施具有重要的意义。然而,甲流增加的死亡不仅仅由“流感和肺炎”引起,还与其引起的呼吸循环系统疾病或诱发的其他基础疾病的恶化有关,导致很难直观获得甲流所带来的实际死亡负担。国际上常用数学模型估计流感流行造成的超额死亡率,以此间接评估流感相关死亡的真实情况。目前,我国有关超额死亡率的研究甚少,华南地区更是缺如。世界卫生组织(WHO)指出每年在流感流行高峰前期接种疫苗是预防流感最有效的方法,自2010-2011年起,WHO将甲流毒株(A/California/7/2009(H1N1) pdm09类似株)作为组份之一加入三价灭活流感疫苗(TIV),以期保护人群预防甲流的危害。然而,TIV能否在人群中发挥保护作用还与人群中的疫苗覆盖率密切相关。一种疫苗覆盖率至少达到60%才可形成有效的免疫屏障,为此,评估人群中TIV的覆盖水平也成为甲流预防免疫策略的重要环节之一。当前评估疫苗覆盖水平的方法通常是采用问卷调查或电话访问,这些方法容易出现信息偏倚,尤其当研究对象为儿童时,信息通常由家长提供,此时更难避免报告偏倚及回忆偏倚。儿童由于免疫系统尚未完全发育,罹患甲流的风险较高,我国每年超过40%的学龄前儿童及30%的学龄儿童罹患流感。为此,评估该人群TIV覆盖水平显得尤为重要。疫苗的保护作用可通过评估免疫原性、效力及效果三个方面来体现。国内对于疫苗的研究往往限于研制与开发过程,其是在严格试验条件下,以抗体水平作为研究终点的“理想”且快速的评价方法。然而,在疫苗实际接种过程中,还受到很多因素的影响,如疫苗存储运输条件、接种方法、疫苗覆盖程度、接种对象特征及疫苗与流行毒株是否匹配等众多因素的影响。因此,疫苗上市前保护作用评估无法真实反映疫苗在实际应用中的保护效果,只有以人群疾病为研究终点的流行病学保护效果评价才可得出疫苗的真实保护效果,即疫苗在实际使用过程中减少接种人群发病的实际保护能力。广州市作为中国的“南大门”,拥有超过794万的常住人口和467万的流动人口,是我国人口最多的省(广东省)中人口最多的城市,是全国交通及贸易重要枢纽；广州市属亚热带季风气候,全年平均气温21-23℃,平均相对湿度75%,这些社会与环境因素导致流感病毒甚为活跃。为此,本研究选取广州市作为研究现场,开展甲流分子流行特征、超额死亡及免疫预防策略等系列研究。具体研究结果如下：第一章新型甲型H1N1流感流行特征及血凝素基因变异特征研究通过分析广州市2009-2013年新型甲型H1N1流感(甲流)的分子流行病学特征及血凝素(HA)基因分子进化与蛋白变异特征,为甲流流行的预测及防控工作提供科学依据。方法收集广州市2009-2013年甲流流行病学及病原学监测资料,采用描述流行病学方法描述甲流的流行病学特征。选取2009-2013年广州市分离的18株甲流毒株,提取病毒核酸RNA,通过Realtime RT-PCR扩增血凝素(HA)片断并进行测序,用MEGA6等相关生物信息软件构建系统发育树,并进行序列比对、同源性分析和氨基酸变异分析。结果1.自2009年5月广州市出现首位甲流病例后发病数逐月增加,11月达最高峰,随后发病数逐渐减少,2011年5月开始无甲流病例报告,直至2013年1月,甲流重新出现并成为优势毒株；市区甲流感染构成比高于城乡结合部和郊区,男性感染构成比高于女性,儿童和青少年感染构成比高于老年人。2.广州市2009年以甲流和H3N2病毒为主要流行毒株,后者引发大流行；2010年以B型和H3N2为主；2011年甲流引发新的流行高峰,随后以B型为主；2012年以B型和H3N2为主,甲流未被检出；2013年全年以甲流为优势流行毒株。五个监测季节的流感流行期分别为4-12月、1-5月、1-6月、2-8月和2-6月,大部分监测季节都出现两个或以上的流行高峰。3.2009-2013年广州地区流行的甲流病毒在HA基因进化上和全球各地同时期毒株均位于同一分支,2009年和2010年甲流毒株与A/California/07/2009疫苗株高度同源；2011年分支毒株在进入2013年后近乎消失,但仍有一株保留下来；氨基酸变异分析显示,从2009-2013年共24个氨基酸位点发生变异,但发生明显变化趋势的位点只有160、214、391；有4个变异位点发生在抗原决定簇,潜在糖基化位点比较稳定,3年间没有发生变化。结论广州市2009-2013年期间甲流发病率在市区、儿童和青年人群中较高,甲流与H3N2、B型流感呈现相互交替的流行特征；2009-2013年广州市甲流病毒与全球同步流行；2009年和2010年与A/California/07/2009疫苗株具有很近的亲缘关系；甲流病毒高度同源,但某些氨基酸位点已发生突变。第二章新型甲型H1N1流感超额死亡的负二项回归模型研究通过评估新型甲型H1N1流感(甲流)及其他亚型流感的超额死亡率,为评估流感流行的严重程度和制定流感防控策略提供依据。方法使用负二项回归模型评估广州市2010-2012年的每周死亡数据和病毒学监测数据,估计广州市2010-2012年每年由甲流及其他亚型流感流行导致的5种死因(流感和肺炎AC、呼吸和循环系统疾病CRD、全死因P&I、缺血性心脏病IHD和慢性阻塞性肺疾病COPD)所引起的超额死亡率。结果1.2010-2012年3个流感流行季节全人群5种死因(AC、CRD、P&I、COPD和IHD)的年均超额死亡率分别为14.72/10万、11.40/10万、2.26/10万、2.81/10万和2.21/10万。≥165岁人群5种死因年均超额死亡率均是全人群的12-15倍,且同样是AC最高、CRD次之。2.2010-2012年,甲流引发的全人群5种死因的年均超额死亡率由高到低为AC、CRD、IHD、COPD和P&I(分别为3.99/10万、3.06/10万、0.72/10万、0.50/10万、和0.40/10万)。甲流引发的年均超额死亡率均集中于≥65岁人群。相比其他两个型别流感,甲流引发的各死因中除IHD外,其他死因的超额死亡率均最低。结论本研究通过拟合负二项回归模型评估广州市甲流及其他亚型流感5种死因的超额死亡率,阐明广州市甲流流行的危害程度。研究表明2010-2012年甲流均导致人群中出现超额死亡率；≥65岁人群的超额死亡率显著高于全人群；除IHD外甲流导致的超额死亡率均低于H3N2和B型流感；B型流感最高。本研究表明,应重视亚热带地区人群流感疫苗的接种工作,特别是≥65岁人群,以减少流感对人群的健康危害。第三章儿童流感疫苗(含甲流组分)覆盖水平及影响因素研究基于儿童免疫规划信息管理系统,调查分析广州市6-107月龄儿童接种流感疫苗(即三价灭活流感疫苗,简称TIV,其中含有新型甲型H1N1流感疫苗组分)的覆盖水平及影响因素,为甲流的免疫干预策略提供依据。方法采用分层随机抽样方法,以年龄段作为分层标准,从广州市儿童免疫规划信息管理系统随机抽6-107月龄儿童,并从该系统收集基本信息及疫苗接种信息。结果1.共调查13399名6-107月龄儿童,TIV接种率(接种一剂及以上的概率)为37.7%,全程接种率为2.3%。2.年龄、性别、地区、户籍是TIV接种率的影响因素,36-59月龄和≥60月龄的女孩更有可能接种1剂或2剂以上TIV,市区及本地儿童接种1剂或2剂以上TIV的概率较郊区及流动儿童大。3.TIV首剂接种月龄中位数为14.0月,各个年份两剂疫苗接种时间间隔不符率从1.2%到14.3%。结论广州市儿童TIV接种水平尚未达到免疫屏障的要求,尤其是全程接种率极低；接种程序上存在首剂接种时间偏晚、接种时间间隔不符等质量问题。应采取有效措施提高TIV的接种率特别是全程接种率,并规范流感疫苗接种程序,从而更好地保护儿童免于感染甲流。第四章疫苗对儿童新型甲型H1N1流感流行病学保护效果的病例对照研究通过评价流感疫苗(即三价灭活流感疫苗,简称TIV,其中含有甲型H1N1流感疫苗组分)在流感流行期对6-107月龄儿童新型甲型H1N1流感(甲流)的上市后保护效果及其影响因素,为调整流感疫苗免疫策略提供参考依据。方法连续收集广州市2013年流感流行期报告的6-107月龄,经实验室检测甲流阳性的患者作为病例组,采用1：1匹配对照设计,以同期未出现流感样症状的社区健康儿童为健康对照组；采用成组对照设计,以实验室未检测出甲流病毒的儿童作为阴性对照组,对健康对照设计进行敏感度测试。从广州市免疫规划信息管理系统收集TIV接种信息,评估TIV的保护效果(1-调整OR)。结果1.研究共纳入2126名病例,健康对照在年龄、性别及地区均与病例一致；阴性对照组纳入200名儿童,除年龄外,性别、户籍和地区分布与病例组相比均无统计学差别。病例组TIV接种率为8.5%,低于健康对照组(17.2%)和阴性对照组(19.5%)。2.对于健康对照设计,2013年流感流行期TIV对甲流的保护效果为58.3%(95%CI,47.9-66.6),与阴性对照设计的保护效果(53.7%(95%CI,30.5-69.2))接近,两者无统计学差异。3.TIV对36-59月龄儿童(78.7%(95%CI,51.8-90.6))的保护效果高于6-35月龄(65.0%(95%CI,52.1-74.3))及60-107月龄(53.4%(95%CI,22.8-71.8));6-35月龄儿童中,完全免疫(74.5%(95%CI,61.6-83.0))的保护效果高于部分免疫(49.9%(95%CI,24.5-66.8))；对男性(63.4%(95%CI,50.6-72.9))的保护效果低于女性(73.3%(95%CI,50.7-85.6))；在甲流流行前期(56.6%(95%CI,31.9-72.3))的保护效果高于流行后期(45.3%(95%CI,15.7-64.5))。4.既往疫苗接种可影响当前流行季节TIV的保护效果,其中研究年度保护效果最高的是研究年度与上一年度(75.5%,(95%CI,57.8-85.8))以及研究年度与上上年度(84.9%(95%CI,53.8-95.1))的联合免疫；仅上一年度(2012-13年度)免疫的保护效果(46.3%(95%CI,26.6-60.8))低于研究年度。5.儿童接种TIV后1-2月内保护效果最高(76.8%(95%CI,66.0-84.1)),接种3-4月后保护效果衰减为50.3%(95%CI,31.0-64.2)。结论采用健康儿童作为对照的病例对照研究来评估疫苗保护效果具有可行性,可基本替代以实验确诊阴性的儿童为对照的研究结果。TIV对6-107月龄儿童患甲流具有中等程度的保护效果,TIV对36-59月龄儿童的保护效果高于6-35月龄；完全免疫的保护效果高于部分免疫；对男性的保护效果低于女性；在甲流流行前期的保护效果高于流行后期。既往疫苗接种可影响但无法替代当前流行季节接种TIV的保护效果,建议6-107月龄儿童每年在流感流行期前期全程接种TIV。更多还原

【Abstract】 Background&ObjectiveAfter the first description of a novel Influenza A (H1N1) pdm09(pH1N1) in Mexico and the United States in April2009, the virus rapidly spread worldwide, resulting in considerable human suffering and economic burden.Then the spread of pH1N1had been moderate since the virus did not undergo critical antigenic change and the host acquired immunity to pH1N1with the mass vaccination. However, in2013, the number of positive isolations of pH1N1began to rise again in Guangzhou, which caused concerns for possible epidemic. With the aim of providing the reference for the prevention and control of pH1N1, we need to carry incessancy surveillance about the pH1N1molecular epidemiological and the studies of pH1N1antigenic variation.Seasonal epidemic and pandemic influenza can caused substantial morbidity and mortality worldwide. The accurate evaluation of this mortality burden is important for the prevention of pH1N1. Because of that the flu diagnosis is usually based on clinical symptoms rather than the laboratory tests and cause death not only due to "flu and pneumonia" but also others serious complications. The death burden of the flu will be greatly underestimated only through analyzing the death of "flu and pneumonia". Foreign research commonly used mathematical models to estimate the influenza-associated excess mortality to accurately evaluate the mortality death. However, relative data were scarce in tropical regions. We aimed to estimate pH1N1-associated excess mortality in Guangzhou, China and assess the different excess mortality due to other subtypes of viruses.The World Health Organization (WHO) contends that annual vaccination prior to the influenza epidemic peak is currently the principal strategy for preventing influenza. Since2010-2011, the pH1N1strain have added in the seasonal trivalent inactivated influenza vaccine(TIV) to protect against pH1N1. The protective role of TIV is also closely related to the population vaccine coverage. Regular surveys of flu vaccine coverage can make a better understanding of the situation about the population’s level of vaccination and immunity population. Most of the current domestic coverage investigation mainly adopts questionnaire survey, which is difficult to avoid selection bias and information bias. Children are susceptible to be infected by influenza, about40%preschool children and30%school child infected by influenza each year in China. Serious outcomes of influenza infection can result in serious influenza complications or even death, especially those with certain health conditions. Therefore, it is necessary to assess the level of TIV coverage in this population.Vaccine effectiveness can be assessed by immunogenicity, efficacy and effectiveness. Postmarketing vaccines have been assessed the safety and validity by preclinical study and phase Ⅰ, Ⅱ, Ⅲ clinical trials. However, these studies are not possible to assess epidemiological effect, immune persistence and the aspect of health economics. Because of that the factors that affect the protection of the actual effect after vaccination is numerous such as the matching degree between influenza virus vaccine strains and popular plant, the time interval of vaccination and epidemic period and the inoculation rate of population, it is necessary to conduct a evaluation of the post-marketing epidemiologic effectiveness based on the endpoint of disease. At present serological study are overwhelming and the observation time too short to be able to make a exact evaluation for the epidemiological effect of vaccine, the immune persistence and benefit and risk of health economics. In conclusion, the influenza vaccine effectiveness is not completely assessed, and further postmarketing studies among children are needed.Guangzhou is located in south China with a subtropical monsoon climate, viral influenza strains that persist in this region could potentially serve as year-round reservoirs of global genetic diversity. On account of the present situation and the problems listed above, study on Excess Mortality of Influenza, relationship between the epidemiological feature of pH1N1and HA genetic variation characteristics, vaccine effectiveness against pH1N1and inoculation rate investigation of children based on Vaccine registration information have been conducted. The concrete results are as following:Chapter1Study on molecular characteristics for influenza A (H1N1) pdm09and genetic characteristics of hemagglutinin.Through epidemiologic and molecular bioinformatics methods, we analyze the molecular epidemiological characteristic, the gene molecular evolution and the protein variation of pH1N1in Guangzhou to provide a stream of basis for influenza.MethodsInfluenza surveillance data were collected and analyzed to describe the epidemiologic feature of pH1N1in Guangzhou during2009-2013.The nucleotide sequences of HA gene of pH1N1viruses were obtained by using RT-PCR and gene sequencing techniques. The characteristics of HA sequences were analyzed with bioinformatics software.Results1. The pH1N1epidemiological characteristics of time showed that the number of pH1N1case increased month by month since May2009when the first imported case was reported, and reached the peak in November, since then the number of case decreased with most of them to be sporadic cases, and there was no case reported after2011May. However pH1N1became predominant strain again since January2013. The regional distribution showed that the infection Component ratio of urban was higher than that of urban-rural fringe and suburb region. The population distribution showed that the infection Component ratio of male was higher than that of female while that of children and adolescents higher than that of elderly group.2. In the respect of subtypes monitoring, H3N2and pH1N1were mainly pandemic strains and in Guangzhou during2009-2010the latter remained a potential pandemic threat, while H3N2and B (H1N1) during2010-2011. And during2011-2012, pH1N1took hold in the first quarter of a year and then pH1N1afterwards. H3N2and B (H1N1) were predominant and pH1N1Had not been detected during2012-2013. pH1N1became a predominant train during2013for a whole year. The five influenza epidemic monitoring season were respectively April-December, January-May, January-June, February-August and February-June and more than2peaks of the pandemic had been detected in most of the monitoring seasons.3. The HA of pH1N1during2009-2013of Guangzhou had multiple branches on the gene evolutionary process and shared the same branch with the other strains all over the world at the same time. The strains of2009and2010shared the same branch and had the high homo logy with A/California/07/2009. The strains of2011was a nearly ones after2013while one of those preserved. Amino acid sequence variation analysis showed15sites of the amino acid residues were not completely consistent from2009-2013which were respectively site4,5,6,7,8,9,10,11,12,65,103,145,421,499and526.ConclusionThe pH1N1virus strains of2009-2013were so closely related with A/California/07/2009in2009and2010worldwide. pH1N1virus was highly homologous but some amino acids sites of that mutated. H3N2, pH1N1, B (H1N1) were alternating with each other in Guangzhou during2009-2013. pH1N1had a higher morbidity in urban region and children and adolescents population. Chapter2Study on influenza A (H1N1) pdm09-associated excess mortality in GuangzhouWe aimed to estimate the influenza-associated excess mortality due to pHlNl and other subtypes of viruses across Guangzhou, which contribute to make a better understanding of the influenza and mortality burden of that.MethodsWe estimated influenza-associated excess mortality due to5-outcome including pneumonia and influenza (P&I), circulatory and respiratory disease (CRD), all cause deaths (AC), ischemic heart disease (IHD) and chronic obstructive pulmonary disease (COPD) according to weekly numbers of deaths and influenza surveillance data through negative binomial regression model during2010-2012.Results1.The annual excess mortalities of5-outcomes (AC, CRD, P&I, COPD and IHD) during3influenza seasons are retrospectively14.72(95%CI,12.12-17.31) per100,000,11.40(95%CI,9.38-13.42) per100,000,2.26(95%CI,1.86-2.67) per100,000,2.81(95%CI,2.31-3.32) per100,000and2.21(95%CI,1.78-2.64) per100,000. Annual excess mortality of5-outcome in≥65group are12-15times higher than in total population, that of AC is the highest, and CRD secondly.2.During2010-2012, the excess mortality of total population for AC, CRD, IHD, COPD and P&I resulted from pH1N1is3.99per100,000,3.06per100,000,0.72per100,000,0.50per100,000and0.40per100,000. People aged>65years made a major contribute to the excess mortality of5-outcome caused by pH1N1. Except IHD(12.01,95%CI,6.98-17.04per100,000), the excess mortality for AC, CRD, P&I and COPD caused by pH1N1is the lower than other two subtypes.Conclusion Influenza pH1N1led to marked excess mortality in every season during2010-2012, which posted the greatest threat to≥65group. Influenza virus Type B accounted for the highest excess mortality than pH1N1. The study indicates that it attachs importance to vaccinating in subtropical regions during influenza season, especial for≥65group in order to reduce the excess mortality. Chapter3Trivalent inactivated influenza vaccination coverage among children aged6-107months in Guangzhou, China.We aim to estimate the influenza vaccine (TIV) coverage rate and factors of children aged6-107months selected from the Children’s Expanded Programmed Immunization (EPI) Administrative Computerized System.MethodsInfluenza vaccine information and basic information of children aged6-107months up to investigation date were collected from the EPI by stratified random sampling.Results1. The TIV vaccination rate of13399children aged6-107months was37.7%, and the fully vaccination rate was was2.3%only.2. Age, gender, region and household factors affected the TIV vaccination rate. Those who are≥60months, female, downtown or local performed higher rate of one or two dose vaccination.3. The median month of first-dose vaccination was14months. The ratio of the incorrect time-lag between two dose of TIV ranged from1.2%to14.3.ConclusionThe coverage rate of TIV in Guangzhou was not enough to emerge herd immunity effects, especially the fully vaccination rate is extremely low; There are lingering problems about the delayed first-dose time and incorrect time-lag between two dose. More action should be taken in order to improve the coverage rate of TIV especial for fully vaccination rate. It’s of importance to strengthen the immunization affairs aiming at low-rate population and standardize the vaccination program for the health of children in Guangzhou. Chapter4Effectiveness of trivalent inactivated influenza vaccine against influenza A (H1N1) pdm09in children during2013in Guangzhou, China, a case-control studyWe aimed to estimate the effectiveness of TIV for children aged6-107months during influenza epidemic period of2013seasons and to provide the scientific basis for control efforts of pH1N1.MethodsChildren aged6-107months diagnosed as pH1N1virus positive and confirmed by laboratory (Real-time PCR positive) during influenza season in2013across Guangzhou were collected and assigned to case group, and cases which are pH1N1negative were assigned to negative control group. Non-paired design was conducted between case group and negative control group. The healthy controls without influenza symptoms were confirmed by researchers, and cases and health control groups were1:1matched by birth date, gender, and outpatient services of vaccination. The pH1N1epidemic periods were defined by the surveillance of circulating viral strains. The pH1N1cases were obtained from the surveillance system. Negative controls were randomly collected from pH1N1-negative patients confirmed by Guangzhou CDC laboratory. Health controls were randomly sampled from the Children’s Expanded Programmed Immunization (EPI) Administrative Computerized System. Non-conditional logistic regression was conducted for case-negative design, and conditional logistic regression was applied for1:1paired case-health design. Confounding factor (age, gender, region, census register and vaccination age) were adjusted for VEs and95%confidence interval.Results1. We analyzed the vaccination information for2126sets of subjects, whose average age of onset was38.7±25.3months and average age47.0±24.9months. Cases mainly were local children and male and Baiyun district own the most cases. Health control followed the same pattern.200children of negative group were analyzed and there was no difference between case and health groups on gender, region and census register except average age (42.2±20.6months). The coverage rate of pH1N1influenza vaccine was8.5%, which lower than health control (17.2%) and negative control (19.5%).2. The VE for pH1N1of health control during2013was58.3%(95%CI,47.9-66.6) and familiar with that of negative control (53.7%95%CI,30.5-69.2).3. Vaccine for36-39months group had a higher VE than6-35months group and60-107months group (78.7%95%CI,51.8-90.6VS65.0%95%C,52.1-74.3and53.4%95%CI,22.8-71.8). Full vaccination provided more protection than partial vaccination among children aged6-35months (74.5%(95%CI,61.6-83.0)vs.49.9%95%CI,24.5-66.8). VE of male was lower than that of female (63.4%95%CI,50.6-72.9VS73.3%5%CI,50.7-85.6). Vaccine provided more protection in pre-epidemic period than in post-epidemic period (56.6%95%CI,31.9-72.3VS45.3%95%CI,15.7-64.5).4. The effect of current vaccine might be influenced by the previous vaccine, and the combined immunization of2013and2012-2013,2013and2011-2012increased most obviously (75.5%95%CI,57.8,85.8AND84.9%95%CI,53.8-95.1). However, VE of2012-2013(46.3%95%CI,26.6-60.8)) was lower than that of2013.5. The VE decreased gradually from76.8%(95%CI,66.0-84.1) in1-2months after vaccination to50.3%(95%CI,31.0-64.2) in3-4months after vaccination.ConclusionIt is feasible to adopt healthy children as a control to assess the VE of TIV. The TIV had moderate protection against pH1N1for children aged6to107months. Annual,full and timely vaccination prior to the influenza epidemic peak should be encouraged for children aged6to107months.更多还原