【作者】 林海江；

【导师】 何纳；

【作者基本信息】 复旦大学 ， 流行病与卫生统计学， 2011， 博士

【摘要】 研究背景浙江省台州地区位于我国东南沿海,是我国艾滋病相对低流行区,但近年来艾滋病发现病例数呈现快速上升态势,以性接触传播为主,特别是男男性接触传播比例上升明显。传统流行病学研究方法主要探索和确定个体发病的危险因素,将人群的发病频率和现象简单理解为人群中各个体发病情况的简单累加和合计。以人群为基础的艾滋病的发生,却不能只简单地孤立地看待人群中每个个体HIV暴露和感染与否,而忽视各个体之间艾滋病相关危险行为与危险行为网络(behavioral networks)以及这种危险行为网络的动态变化。行为网络理论在艾滋病防治领域的运用为艾滋病防控提供了新的视角。目前包括台州地区在内的以性接触为主要传播途径的艾滋病相对低流行区,缺乏准确的艾滋病传播行为动力学和分子机制数据,进而影响到艾滋病综合防治策略制定和实施的有效性和科学性。研究目的1.在台州地区运用行为网络理论,最大限度地早期发现常规工作中难以发现的HIV感染者,掌握HIV感染者与其高危接触者的艾滋病相关危险行为及行为网络特征；2.通过HIV感染者队列随访观察,掌握HIV感染者与其高危接触者的艾滋病相关危险行为及行为网络特征的动态变化,为制定旨在减少HIV感染者二代续发感染传播的预防干预措施和艾滋病综合防治策略提供依据；3.以感染HIV的男男性行为者(MSM)为核心,综合运用流行病学、行为网络理论方法及系统进化树分析技术,描述MSM人群中艾滋病相关危险行为网络特征和HIV优势病毒株,确定MSM人群中HIV传播的行为动力学和分子机制。研究方法1.台州地区新报告HIV感染者溯源调查以台州地区2008年1月1日至2010年12月31日期间新报告的HIV感染者为核心(指示病例),追踪调查其所提供的高危接触者,包括详细调查其艾滋病相关危险行为及行为网络特征以及HIV检测(简称“溯源调查”)。2.台州地区HIV感染者随访队列研究将台州地区2009年12月所有在访的HIV/AIDS病例纳入队列研究,分别于入组时及随访一年后对其艾滋病相关危险行为及行为网络进行流行病学调查,以比较该队列HIV感染者艾滋病相关危险行为及行为网络的动态变化。3.台州地区MSM人群艾滋病危险行为网络与分子流行病学研究以台州地区2008年1月1日至2010年12月31日期间新报告HIV阳性的MSM为核心开展溯源调查,详细调查其高危接触者的危险行为网络特征。综合运用流行病学、行为网络理论方法以及系统进化树分析技术,构建MSM人群HIV基因亚型网络和传播链。主要研究结果1.台州地区新报告HIV感染者溯源调查在研究期间(2008年1月1日至2010年12月31日)共新发现463例HIV感染者,其中462例符合研究入选标准,398例HIV感染者作为指示病例参与溯源调查和危险行为网络研究,参与率为86.1%。398例参与溯源调查的HIV感染者中,有396例有过性行为经历,累计报告5183名性伴,包括100例经男男性接触传播的HIV感染者共报告男男同性性伴1534名。86.7%(343/396)报告了包括配偶或固定性伴在内的2名及2名以上的性伴。HIV感染者与其报告的性伴共组成288个相对独立的危险行为子网络,其中,5.9%(17/288)性行为网络由两人组成,10.4%(30/288)由3-5人组成,19.1%(50/288)由6-10人组成,33.7%(97/288)由11-20人组成,30.9%(89/288)性行为网络规模在20人以上。其中,29个性行为网络同时含有男男同性性伴和异性性伴。398名指示病例共提及1403名最熟悉的高危接触者,其中320名前来接受HIV检测,检测参与率为22.8%。结果新发现125例HIV阳性,阳性检出率为39.1%,其中配偶或异性固定性伴HIV阳性检出率为34.9%(74/212),男男同性性伴HIV阳性检出率为54.3%(53/94)。2.台州地区HIV感染者随访队列研究2009年底,285例HIV/AIDS病例纳入队列研究,其中完成一年随访的有264例,随访率为92.6%。264例随访的HIV/AIDS病例中,有262例报告在确认感染HIV前有过性行为,他们累计提供性伴4012名,整体网络的规模为4274,平均每个研究对象报告15.3+15.9名(中位数为12名)性伴；经过为期一年的随访,有168例报告在近一年内有性行为,他们共提供246名性伴,整体性行为网络规模414,平均每个对象报告1.5±2.3名(中位数1名)性伴。经过为期一年的随访,经异性性接触传播的HIV感染者的高危接触者类别构成发生了较大的变化,确认感染HIV前主要以商业性性伴为主,占56.1%；在基线调查及最近一年随访期内,高危接触者类别转变成以配偶或固定性伴为主,分别占50.1%和84.4%。经男男同性性接触传播的HIV感染者的高危接触者类别,确认感染HIV前主要以非商业非固定性伴为主,占57.3%,在基线调查和最近一年随访期内,高危接触者构成分别以非商业非固定性伴、商业性性伴为主,分别占55.4%和52.4%。以确认感染HIV到基线调查不同时间长度为分层变量,经异性性接触为传播途径的HIV感染者,在确认感染HIV前,与配偶或固定性伴发生性行为的频率平均每月4次及以上的占(75.0-91.5)%,性行为过程中从未使用安全套的比例占(69.4-80.8)%；与非商业非固定性伴、商业性性伴发生性行为的频率平均每月4次及以上的分别占(13.3-35.7)%和(37.5-68.5)%,性行为过程中从未使用安全套的比例分别占(32.0-46.2)%和(21.8-41.4)%；在一年的随访期内,与配偶或固定性伴发生性行为的频率平均每月4次及以上的占(86.4-91.1)%,但性行为过程中每次均使用安全套者所占的比例为(90.9-98.2)%,高于基线调查时所占比例。与HIV确认之前和基线调查相比,随访时HIV/AIDS病例提及的商业性、非商业性异性性伴规模显著减小。同时,与商业性性伴、非商业性性伴发生性行为时绝大多数使用安全套,仅个别HIV感染者在性行为过程中不使用安全套。经男男同性性接触传播的HIV感染者,在确认感染HIV前,与高危接触者性行为的频率平均每月4次及以上的占100%,性行为过程中每次都使用安全套的仅占14.5%。在最近一年的随访期内,他们与高危接触者发生性行为的频率平均每月4次及以上的占94.7%,在性行为过程中每次使用安全套的比例上升,占54.5%,有40.9%的HIV感染者与高危接触者在发生性行为时有时使用安全套。3.台州地区MSM人群艾滋病危险行为网络与分子流行病学研究研究期间共发现MSM人群HIV阳性者111例,其中包括49例经以HIV阳性MSM为核心开展的第一轮溯源调查所发现,3例经第二轮溯源调查所发现。作为指示病例的HIV感染者报告他们最熟悉的高危接触者性行为网络规模中位数为8位(最小值1,最大值80),其中有23.8%超过21位,20.8%在11-20位之间。MSM人群HIV阳性者与他们报告的高危接触者共形成49个相对独立的子网络,其中29个子网络同时包含男男同性性伴和异性性伴。高危接触者类别以男男同性性伴为主,约占90%,在发生性行为时每次均使用安全套的比例较低,仅占6.6%。综合分析env和gαg基因序列,台州地区HIV阳性MSM感染的HIV病毒株存在CRF01_AE、CRF07_BC和B亚型等3种基因亚型,以CRF01 AE为主,以特异性env, gag基因扩增分别占73.3%和72.1%；其中有2例同时存在CRF01 AE与CRF07 BC两种亚型,2例同时存在CRF01 AE与B两种亚型,3例同时存在B与CRF07 BC两种亚型。综合艾滋病相关危险行为流行病学调查和分子生物学检测结果,有18个行为子网络内各个体之问不仅存在流行病学意义上行为学关联,而且子网络内至少有2个及以上个体感染的HIV基因亚型一致,基因序列聚集在一起,提示他们之间存在直接HIV传播关系。6个行为子网络尽管他们存在着流行病学意义上的行为学关联,但系统进化树分析结果提示他们之间不存在直接HIV传播关系。另外还有11个子网络,由于各种原因无法检测出子网络内某些个体的HIV基因亚型,目前无法准确判断他们之间HIV传播关系。研究结论1.以HIV感染者为核心的溯源调查工作能高效发现常规工作中难以发现的HIV感染者,这种方法不仅适用于艾滋病高流行区,同样适用于以性接触传播途径为主的艾滋病相对低流行区。2.台州地区HIV感染者与其高危接触者性行为及行为网络关系复杂,且危险行为方式隐蔽。特别是MSM人群该特征尤为显著,存在着HIV广泛传播风险。3.在确认感染HIV前后及经过为期一年的随访,HIV感染者的性行为及行为网络发生显著变化。即由原先密集复杂的性行为网络变为稀疏、松散的不易HIV传播的行为网络结构。高危接触者类别以非配偶或非固定性伴为主向以配偶或固定性伴二人组成为主转变。危险行为及行为网络结构的改变可有效减少其流行病学传染源意义和HIV续发二代传播。4.台州地区HIV阳性MSM人群中HIV亚型以CRF01 AE为主,约占75%,同时存在B和CRF07 BC亚型,且存在同时感染两种HIV亚型的个体。综合运用行为网络理论、分子流行病学方法及系统进化树分析技术,为精确构建以HIV感染者为核心的HIV传播网络链提供了更为科学的方法,为HIV传播动力学与防治策略的制定调整提供了更为科学准确的参数。更多还原

【Abstract】 BackgroundTaizhou prefecture is a relatively low HIV prevalence area in China. The reported HIV infection in Taizhou prefecture continues to increase rapidly, especially during the recent 3 years. And sexual transmission is the primary route of transmission with homosexual transmission increasing rapidly. Traditional epidemiological researches on diseases mainly focus on independent individuals rather than as systems of interacting individuals. In recent years, researches applied risky behavioral network theory to the HIV control and prevention, which provide valuable insights into HIV transmission that might otherwise go unrecognized by routine practices. Presently, dynamic behavioral and molecular data of HIV transmission is still facing shortage in Taizhou prefecture and other relatively low HIV prevalence areas with sexual transmission as the primary route, which has seriously affected the development of comprehensive HIV prevention and control strategies effectively and scientifically.ObjectivesThe study was conducted in Taizhou prefecture, a low HIV-prevalence area in China with specific aims:(1) to trace and HIV-test those having risky contacts with HIV-infected individuals at various levels and characterize the risky behavioral networks among HIV infections and their contacts (with and without HIV infection) and its dynamic risky networks of HIV infections. (2) to reveal the dominant HIV subtypes and the reconstruction of the HIV transmission networks using a dated phylogeny approach and provide information of the HIV transmission dynamics among men who have sex with men (MSM). (3) to provide information of the parameterization of epidemiological and molecular models to design and improve intervention strategies for HIV control and prevention.Methods1. Tracing the sources of newly reported HIV infections in Taizhou prefecture All newly reported HIV infections from January 2008 through December 2010 in Taizhou prefecture were invited to participate in a cross-sectional survey which requested numbers and contact information of individuals with whom they had ever had high risk contacts as well as risky behavioral acts. Persons having had risky contacts with HIV-infected cases were thereof approached to participate in this contact tracing survey and receive HIV testing. Those who tested positive for HIV were subject to further round of contact tracing surveys.2. Cohort study on the HIV infections in Taizhou prefectureAll reported HIV infections before December of 2009 meet the admission criterion of the study were included to characterize the dynamic risky behavioral networks through a one-year follow up. Baseline and dynamic data on AIDS-related risky behaviors including numbers, type of the contacts and usage of the condom during the sexual intercourse was collected and compared when the HIV infections enrolled the cohort and were over a one-year follow-up.3. AIDS-related risky behavioral networks and molecular epidemiology study of newly reported HIV infected MSM in Taizhou prefectureAll newly reported HIV infected MSMs and their risky contacts during the study period from 2008 through 2010 were included into the study to identify the HIV infections and characterize the egocentric risky behavioral networks of the HIV infections with their contacts through contact tracing methods. Combined with behavioral network theory, reconstruction of HIV transmission egocentric risky behavior and viral genotypic networks among HIV-infected MSM was established using modern epidemiology and recently developed phylogenetics techniques.Main results1. Tracing the sources of newly reported HIV infections in Taizhou prefecture There were a total of 463 newly reported HIV infections and 462 HIV infections met the study criterion during the study period. Among those,398 newly reported HIV infections, with the paticipation rate of 86.1%, served as the’index cases’were included into the contact tracing and risky behavioral network study. Of those,396 cases with sexual experiences reported a total of 5,183 individuals with whom they had ever had sexual contacts, including HIV-infected MSM reporting a total of 1,534 male sexual partners. And 86.7% (343/396) of them reported multiple sexual partners (two or more sexual partners).There were a total of 288 independent egocentric sexual networks, including 29 egocentric sexual networks with both heterosexual and homosexual partners, were identified and 5.9%(17/288) of networks are dyads which included 2 members, 10.4% (30/288) included 3 to 5 members,19.1%(50/288) included 6 to 10 members, 33.7%(97/288) included 11 to 20 members,30.9%(89/288) included 21 members or more.398 HIV cases served as’index cases’reported totally 1,403 risky contacts with risky behavioral information. Among those,22.8%(320/1403) of contacts received HIV testing and 39.1% of the tested were positive for HIV. The study indicated that the highest HIV prevalence of the contacts those who come to receive HIV test was among spouses or regular heterosexual partners (34.9%,74/212) and male sexual partners of HIV-infected MSM (54.3%,53/94).2. Cohort study on the HIV infections in Taizhou prefecture In the study, a total of 285 HIV/AIDS infections enrolled the baseline study and 92.6%(264/285) of them completed the one-year follow-up. Among them,99.2% (262/264) reported that they have sexual behaviors before confirmation of HIV infection and reported a total of 4012 risky sexual contacts. The size of the whole network is 4274 and each participant had an average of 15.3±15.9 sexual partners.During the one-year follow-up period,63.6%(168/264) cases reported they had sexual behaviors and reported a total of 246 sexual partners and each participant had an average of 1.5+2.3 sexual partners. Of those infected with HIV through heterosexual contacts, their main types of sexual partners were commercial sexual partners (56.1%) before confirmation of HIV infection. However, at the baseline and during the recent one-year follow-up period, their main types of sexual partners changed to spouse or regular partners, about 50.1% and 84.4%, respectively. Of those infected with HIV through homosexual contacts, their main types of sexual partners were noncommercial causal sexual partners (57.3%) before confirmation of HIV infection. At the baseline and during the recent one-year follow-up survey, their main types of sexual partners changed to spouse or regular partners, about 55.4% and 52.4%, respectively. Of those infected with HIV through heterosexual contacts stratified by the time from baseline survey to HIV confirmation, before confirmation of HIV infection,75.0%-91.5% of them had sex with spouse or regular partners four times or above per month,69.4%-80.8% of them had never used condoms with their spouse or regular partner during their sexual intercourse.13.3%-35.7% and 37.5%-68.5% of them had sex with their noncommercial causal and commercial sexual partners four times or above per month, respectively. During their sexual intercourse,32.0%-46.2% and 21.8%-41.4% of them had never used condoms with their noncommercial causal and commercial sexual partners, respectively. During the recent one-year follow-up period,86.4%-91.1% of them had sex with their spouse or regular partners four times or above per month, and the proportion of participants who had consistent condom use with their spouse or regular partners increased to 90.9%-98.2% during their sexual intercourse. And they had consistent condom use with non-commercial causal and commercial sexual partners although they reported few noncommercial causal and commercial sexual partners. However, few HIV infections did not have consistent condom use during their sexual intercourse after their HIV confirmation. Of those infected with HIV through homosexual contacts, before the confirmation of HIV infection,100% of them had sex with high risky contacts four times or above per month and only 14.5% of them had consistent condom use during sexual intercourse. During the recent one-year follow-up period, 94.7% of them had sex with high risky contacts four times or above per month and 54.5% of them had consistent condom use. However, there were 40.9% of HIV-infected MSM who sometimes used condom during their intercourse in the recent one year of the study.3. AIDS-related risky behavioral networks and molecular epidemiology study among MSM in Taizhou prefectureTotally 111 HIV cases were newly reported during the study period. Among those, 101 cases participated in the contact tracing survey and served as’index cases’, including 49 cases identified from the first round contact tracing and 3 cases identified from the second round. Epidemiological study indicated that the median number of the high risky contacts that index cases reported was 8 (min:1, max:80),23.8% of the ’index cases’reported they had more than 21 high risky contacts and 20.8% reported that they had 11 to 20 high risky contacts.90% of them reported their main type of sexual partner was male sexual partners, and 6.6% of them had consistent condom use during their sexual intercourse. There are three HIV-subtypes including CRF01_AE, CRF07_BC and B among the HIV infected MSM in Taizhou prefecture. CRF01_AE was the dominant subtype in Taizhou prefecture, which accounted for 73.3% and 72.1% respectively, when using the phylogenic analysis of the nucleotide sequences of the env and gag gene. In addition,2 cases were found to have recombinant strain related to both CRF01_AE and CRF07_BC,2 cases were found to have CRF01_AE and B, and another 3 cases were found to have CRF07_BC and B.On the basis of phylogenic analysis of the nucleotide sequences of the env and gag gene and epidemiological survey on AIDS-related risky behaviors, at least two or more epidemiologically linked HIV infected individuals among each of the 18 clusters were identified to have direct HIV transmission by viral genetic evidence of close match with sequence and genetic distance. Whereas, there were 6 clusters of discordance between epidemiologically linked with HIV sequence isolated, which reflected that individuals HIV sequence isolated rejected direct HIV transmission on the basis of genetic distance and phylogenetic tree. And for another 11 clusters, direct HIV transmission can not be identified due to negative PCR for the low HIV virus load less than the limit of detection of some HIV infections among the risky behavioral networks or other reasons. In addition, some epidemiologically unlinked clusters had isolated with matching HIV sequence, which suggested that some HIV infected individuals among the clusters may share the same source of HIV infection.Conclusion1. Contact tracing, the practice of identifying HIV infections has been confirmed to be successful in both high and low HIV-prevalence areas. Contact tracing, as a supplement to routine contact investigations to identify unrecognized HIV infections, adds a comprehensive public health approach to the epidemic that will be successful in reducing the ever-escalating numbers of new HIV infections.2. Newly reported HIV infections in Taizhou prefecture reported a large and complicated risk behavioral networks and low rate of condom use, suggesting a potential risk of HIV transmission among these connected people, especially among MSM.3. After their HIV confirmation and comprehensive intervention, the dynamic behavioral networks of the HIV infections changed from the large, complicated networks to a small, simple one over time and the main types of sexual partner changed from non-regular, non-spouse sexual partner to regular or spouse sexual partner, which do not facilitate the HIV transmission.4. Reconstruction of the HIV transmission network using modern epidemiology and a dated phylogenetic tree analysis approach has revealed a large and complicated risk behavioral networks and HIV transmission among MSM in Taizhou prefecture. CRF01_AE was the dominant epidemic CRF strains among the MSM of Taizhou prefecture while the subtype B and CRF07_BC were also circulating among MSM in Taizhou prefecture. The updated phylogenetic tree analysis approach combined modern epidemiology provides valuable insights into HIV transmission that might otherwise go unrecognized by routine practices. The quantitative description of the transmission dynamics among MSM will be important for parameterization of epidemiological models and designing intervention strategies.更多还原