Transmission Network Based Evaluation Of HIV Intervention And Treatment Strategy In Shenyang City

Objective:The epidemic of AIDS in China has been changed from the intravenous drug users(IDU)and paid people for blood supply to mainly by sexually transmitted people in recent years,in which men who have sex with men(MSM)of the HIV epidemic is growing rapidly.According to the report on the epidemic situation of AIDS and STDs in China,the proportion of MSM in the newly diagnosed HIV infection has been increased dozens of times.Because of the traditional marriage morality and social discrimination and other factors,China's heterosexual marriage MSM become the bridge population to spread HIV to ordinary women,which affects a wider range of people and their intervention is more difficult.We urgently need new technology guidance to carry out HIV preventive intervention scientifically.At the same time,China has adopted a lot of policies,including advance the time to start anti-retroviral therapy(ART)and a large number of AIDS intervention and intervention works carried out by disease control departments and research institutions.However the issue is how to effectively carry out treatment and intervention to improve the effect of controlling HIV transmission.All the above need new scientific means to guide.On the other hand,the effect of these treatments and interventions is mainly determined by the number of new diagnosed cases,which also needs more scientific evaluation methods.Due to the high specificity of HIV,and the influence of host immune response and the selection pressure of antiviral drugs,the sequences of HIV strains carried by each infected person have fingerprint characteristics.Therefore,HIV sequences can identify closely related virus strains.The HIV sequence is used to construct the molecular transmission network,and the identification of the transmission cluster is particularly useful in identifying the clusters that other methods can't find.Molecular transmission network are mainly used in the following aspects:Inform us about the spread of the virus between the region and the population.It reveals the critical population in the network and the demographic,behavioral and biological characteristics of the infection in the transmission network,and identifys the factors that drive the virus transmission.Guiding the development of targeted intervention.Evaluation of the effect of intervention measures.Forecast the development trend of the epidemic situation.The previous study of viral molecular networks abroad mainly aimed at the subtype B,the epidemic situation was relatively stable.However,in a complex situation with a variety of HIV strains common popularity and different transmission dynamics,the problem is how to use the HIV molecular transmission network to assess the effect of intervention and guide targeted intervention.Shenyang is the capital of Liaoning province.It is the first area in China to report the HIV epidemic in the MSM population.It is estimated that the number of MSM is about 140 thousand,second only to Beijing.Our team's early study found that a variety of virms subtype coexistence in Liaoning heterosexuals in 1991-2007,other ways of transmission have their own dominant strains.The epidemic of HIV in the MSM increased rapidly and the blood transmission decreased.CRF01_AE replaced the B subtype to become the dominant strain.Our follow-up study also found that the HIV genotype of MSM in Shenyang is complex.There are mainly three genotypes of CRF01_AE,CRF07 BC and B.In addition,there are a variety of recombinant viruses.Recently,our team has worked with the California University of America to analyze the molecular network of the local HIV infection and establish an optimized gene distance threshold for HIV epidemic strains with different transmission dynamics.On the above basis,this study uses the molecular transmission network method of optimizing parameters to carry out a retrospective study of large samples.The purpose is to reveal the dynamic change of expanding and new cluster,and review the effect of Shenyang HIV prevention,intervention and treatment on controlling HIV transmission based on the dynamic change of molecular transmission network in recent years.Methods:1.Study subjects:A total of 2268 HIV-1 positive patients and 2132 HIV negative MSM volunteer who had high risk behaviors from the First Affiliated Hospital of China Medical University in 2002-2016.We collected the HIV-1 positive plasma samples and demographic data such as demography(gender,age,nationality,occupation,education,marital status,etc.),geography(census register,etc.),clinical and biological data(viral load)and so on,forming a database.All infected persons signed the informed consent form.This study was approved by the ethics committee of the First Affiliated Hospital of China Medical University.2.Plasma virus RNA extraction and pol region amplification:First,collect HIV infected blood and separate plasma.Next,extract 60?l viral RNA from 140?l plasma according to QIAamp Viral RNAMini Kit standard protocol.HIV-1 pol 1.0Kb partial gene fragment(HXB2:2352-3352)was amplified by nested PCR method.The first round amplification primers were MAW-26:5'-TTGGAAATGTGGAAAGGAAGGAC-3'and RT-21:5'-CTGTATTTCTGCTATTAAGTCTTTTGATGGG-3'.The second round primers were PRO-1:5'CAGAGCCAACAGCCCCACCA-3' and RT-4R:5 '-CTTCTGTATATCATTGAC AGTCCAGCT-3'.3.Sequence quality assessment:The sequences were aligned using the HIVAlign program and were then manually edited.The retention length is 1000bp(HXB2:2253?3252),covering the protease area 1?99 amino acids and the reverse transcriptase region 1?234 amino acids.4.Virus subtype identification:Using FastTree 3.0 to construct approximate maximum likelihood tree,in which subtype N is the outgroup,the nucleotide substitution model is GTR+ G +1,the support value of the node is calculated by SH test(ShimodairaHasegawa-like test).According to the result of ML Tree,all HIV sequences of2002-2016 were divided into separate Fasta formats.5.HIV transmission network visualization:The HYPHY 2.2.4 TN93 model is used to calculate the pair distance between thepol gene sequences and be derived.We select the data that is less than the threshold of network to import Gephi(CRF01_AE selection gene distance<0.006,CRF07_BC selected gene distance<0.006,B subtype selection gene distance<0.016)to realize HIV molecular transmission network visualization.Different attributes(size,color,shape)are given to the node according to the analysis requirements.6.The correlation analysis of the molecular transmission network:The dynamic analysis of the molecular transmission network is carried out in a year.The population proportion of HIV molecular transmission network cluster and the clustering rate of HIV transmission network in each year were calculated.Classification of cluster according to the size and diffusion speed.Then the contribution of the cluster of different size/diffusion rate and the static/expanding/new transmission cluster of Shenyang HIV infection was identified.Based on the network,we observed the dynamic changes of different types of transmission cluster under ART and other intervention measures in recent years,and analyzed the effect of HIV prevention and treatment intervention in Shenyang area.At the same time,we selecting a large transmission cluster in molecular transmission network to explain the ART and other intervention measures 's effect on reducing the spread of HIV from the perspective of the expansion and vicissitude of large cluster.7.Statistical methods:Statistical analysis was carried out with SPSS 20.0.After removed the cases before 2009 and non MSM from 2002-2016 all clustered,large/rapid clustered or non clustered in the network,compared with 2009-2016 high-risk MSM respectively,the logistic regression model was used to analyze two groups of individuals.Univariate logistic regression analysis of variables with P value less than 0.05 was included in the multiple factor model.The P value was less than 0.05,indicating that the difference was statistically significant.Results:1.The characteristics of the molecular transmission network of HIV infection in Shenyang1.1 Main viral subtypes of HIV infection in Shenyang:CRF01_AE,CRF07_BC and B During the study period,a total of 2268 HIV infected people from the First Affiliated Hospital of China Medical University,the ML Tree constructed by their pol sequence showed that the most infection were CRF01_AE(75.8%).The CRF07_BC and B were less than CRF01_AE,which accounted for 12.2%and 7.8%respectively,and the other subtypes accounted for only 4.1%.1.2 The overall composition of the three main subtypes of HIV molecular transmission networkMSM accounts for 88.2%,91.4%and 77.8%in the CRF01_AE,CRF07_BC and B network respectively.The proportion of HTS in B network transmission cluster reaches 21.4%.IDU is lower than 1.5%in three subtypes of molecular transmission network.In the 3 network,both MSM and HTS were interconnected.IDU is basically clustered individually in the CRF01_AE network,occasionally connecting to MSM.However in the CRF07_BC and B network,IDU can be seen to interconnect with HTS.1.3 The composition of the three main subtypes of molecular transmission cluster are MSM During the 2002-2016,the population of the three main subtypes of network transmission cluster gradually shifted from the majority of HTS to the MSM,and the proportion of IDU was less.Although the proportion of MSM and HTS fluctuates in the B network,it is worth noting that after 2008,the three main subtypes of molecular transmission cluster were mainly composed of MSM,and the proportion of MSM was significantly higher than that of HTS.1.4 Large clusters existed in the three main subtypes of HIV molecular transmission network In the three main subtypes of molecular transmission network,there are large cluster of more than 10 individuals,and cluster constituted by 2 individuals is the most.In CRF01_AE,CRF07 BC and B network,the number of large clusters is 7,1 and 2 respectively.2.Dynamic changes of molecular transmission cluster of HIV infection in Shenyang2.1 Dynamic changes of three main subtypes of HIV molecular transmission network In order to reveal the transition of HIV molecular network in Shenyang,we dynamically analyzed the expansion of clusters in three subtypes of molecular transmission network in a year and found that the three subtypes of molecular transmission network all have the expansion of existing cluster and the formation of new cluster.There are also situations where the original cluster is no longer or rarely expands to present a state of rest.2.2 Clustering rate changes of three main subtypes of HIV molecular transmission network We have focused on the dynamic changes of different subtypes of transmission clusters since 2008,the MSM has dominated the HIV epidemic in Shenyang.In 2008-2010,the clustering rate of CRF07 BC and B was significantly higher than that of CRF01_AE.Since 2011,the clustering rate of CRF01_AE and CRF07 BC decreased significantly,from 18%to 40.9%,while the clustering rate of B increased significantly,from 66.7%to 77.8%.2.3 The impact of large/rapid spread clusters on the epidemic gradually declineAccording to the size and growth rate of cluster,the large/rapid cluster is defined as the cluster has more than 10 nodes in 2016 and/or the growth rate of more than 3 nodes/year in 2009-2016 and small and medium/slow cluster is not satisfied with the condition of large/rapid cluster,but has more than 2 nodes.The results showed that the overall clustering rate of new diagnosed infection decreased gradually with the change of years,and the influence of large/rapid clusters declined gradually,and the influence of small and medium/slow cluster increased.2.4 The driving effects of different types of cluster on the growth of the HIV epidemic in Shenyang are gradually changingIn order to further analyze the main driving factors of HIV epidemic growth in different years,we defined clusters of expanding,static and new according to the characteristics of clusters.The static cluster is defined as the existing cluster of non new node in the next year;the expanding cluster is the existing cluster has new node in the next year;the new cluster is the cluster has no contact with the existing cluster in the next year.The results showed that as the year changed,the proportion of expanding cluster in the annual new infection had reduced,which means it has a diminishing impact on the epidemic,from 36%in 2009 to 20%in 2016.The changes in the new cluster were not significant.While the proportion of non clustered cases increased from 45%in 2009 to 65%in 2016.3 Retrospective evaluation of the effect of HIV prevention intervention in Shenyang area based on molecular transmission network3.1 The control effect of intervention measures based on ART in HIV transmission in different periodsWith the expansion of ART coverage,in the molecular network,the number of new infection increased in different years due to the different source channels of referral,especially since 2014.But it is worth noting that the number of recent infection did not increase,and there is a downward trend.The proportion of cases from large/rapid cluster continue to shrink,falling from 39%in 2009 to 14%in 2016,while the proportion of cases from small and medium/slow cluster increased from 16%in 2009 to 21%in 2016 in new infection,suggesting that the impact of large/rapid cluster on the epidemic situation is narrowing.On the other hand,the rate of virus control in case of molecular network has increased significantly.The proportion of expanding cluster and new clusters in the total transmission cluster has decreased,while the proportion of static cluster has increased.All of these phenomena suggest that the active preventive intervention in recent years has a significant impact on the virus molecular transmission network.3.2 ART and other intervention measures,s effect on reducing the spread of HIV from the perspective of the expansion and vicissitude of large cluster In order to further clarify the effect of prevention and intervention in recent years,we chose a CRF01_AE rapid/large cluster AECL2 to independent observation and analysis:with the expansion of the coverage of ART,the proportion of virus controllers in the transmission cluster has been increasing.Although the number of new cases has increased in recent 2013-2015 years due to the addition of the unit transfer treatment cases,after it has dropped significantly in 2016.The expansion of recent infection occurred only in 2009.4 Analysis of the related factors of HIV transmission in Shenyang area under the intervention of active preventive treatment4.1 Analysis of related risk factors of clustering spread of 2008-2016 HIV in Shenyang In order to identify the related factors of HIV clustered transmission in Shenyang,we used logistic regression analysis to compare the clustered group of three different genotypes of molecular networks with high-risk group respectively.In CRF01_AE network,we used whether the research object is in the CRFOI_AE network cluster as the outcome variable.In the multiple factor logistic regression analysis,after correcting education,marital status,occupation,age,domicile place mutually,we found that risk factors associated with clustered spread of MSM of CRFO1_AE included married(VS.unmarried)(AOR=1.76,95%CI=1.20-2.58,P=0.004),divorced(VS.unmarried)(AOR=1.88,95%CI=1.13-3.14,P=0.016);at the age of 25-29(VS.?24)(AOR=1.54,95%CI=1.10-2.15,P=0.001),at the age of 30-39(VS.?24)(AOR=1.99,95%CI=1.38-2.87,P<0.001),at the age of 40-49(VS.?24)(AOR=2.10,95%CI=1.30-3.39,P=0.003),age?50(VS.?24)(AOR=2.13,95%CI=1.19-3.84,P=0.012);Shenyang(VS.non Shenyang)(AOR=11.32,95%CI=8.44-15.18,P<0.001).In CRF07 BC network we used whether the research object is in the CRF07 BC network cluster as the outcome variable.In the multiple factor logistic regression analysis,after correcting education,occupation,age,and domicile place mutually,we did not find risk factors associated with clustered spread of MSM of CRF07_BC.In B network we used whether the research object is in the B network cluster as the outcome variable.In the multiple factor logistic regression analysis,after correcting education,occupation and domicile place mutually,we found that risk factors associated with clustered spread of MSM of B included farmer(VS.other)(AOR=9.51,95%CI=1.49-60.74,P=0.017);Shenyang(VS.non Shenyang)(AOR=29.11,95%CI=10.36-81.79,P<0.001).4.2 Analysis of related risk factors of 2008-2016 large/rapid cluster in Shenyang In order to further distinguish the risk factors of large/rapid cluster in Shenyang,we used logistic regression analysis to compare the three subtypes of large/rapid cluster group with high-risk group.In CRFO1_AE network we used whether the research object is in the CRF01_AE large/rapid network cluster as the outcome variable.In the multiple factor logistic regression analysis,after correcting education,marital status,occupation,age,and domicile place mutually,we found that risk factors associated with large/rapid clustered spread of MSM of CRF01_AE included individual business(VS.other)(AOR=1.88,95%CI=1.06-3.31,P=0.030);at the age of 30-39(VS.?24)(AOR=1.90,95%CI=1.17-3.06,P=0.009),at the age of 40-49(VS.?24)(AOR=2.67,95%CI=1.46-4.87,P=0.001),age?50(VS.?24)(AOR=2.57,95%CI=1.20-5.50,P=0.015);Shenyang(VS.non Shenyang)(AOR=9.76,95%CI=6.58-14.47,P=0.001).In CRF07_BC network we used whether the research object is in the CRF07_BC large/rapid network cluster is the outcome variable.In the multiple factor logistic regression analysis,after correcting education,occupation and domicile place mutually,we did not find risk factors associated with large/rapid clustered spread of MSM of CRF07_BC.In B network we used whether the research object is in the B large/rapid network cluster is the outcome variable.In the multiple factor logistic regression analysis,after correcting education,occupation,age,and domicile place mutually,we found that risk factors associated with large/rapid clustered spread of MSM of B included farmer(VS.other)(AOR=10.37,95%CI=1.42-75.80,P=0.021);Shenyang(VS.non Shenyang)(AOR=43.07,95%CI=10.17-182.34,P<0.001).4.3 Analysis of related risk factors of 2008-2016 non clustering individuals inShenyang HIV molecular transmission network In recent years,the proportion of new infection is increasing.We use logistic regression analysis to explore the related risk factors of the cases of non clustered.In CRF01_AE network we used whether the research object is individual that outside the CRF01_AE network cluster is the outcome variable.In the multiple factor logistic regression analysis,after correcting education,marital status,occupation,age and domicile place mutually,we found that risk factors associated with sporadic spread of MSM of CRF01_AE included married(VS.unmarried)(AOR=1.51,95%CI=1.03-2.23,P=0.037),divorced(VS.married)AOR=1.97,95%CI=1.20-3.23,P=0.007);at the age of 25-29(VS.?24)(AOR=1.64,95%CI=1.18-2.28,P=0.003),at the age of 30-39(VS.?24)(AOR=2.68,95%CI=1.87-3.84,P<0.001),at the age of 40-49(VS.?24)(AOR=2.57,95%CI=1.61-4.11,P<0.001),age)50(VS.?24)(AOR=4.27,95%CI=2.45-7.45,P<0.001);Shenyang(VS.non Shenyang)(AOR=23.04,95%CI=16.26-32.65,P<0.001).In CRF07_BC network we used whether the research object is individual that outside the CRF07_BC network cluster is the outcome variable.In the multiple factor logistic regression analysis,after correcting education,age and domicile place mutually,we found that risk factors associated with sporadic spread of MSM of CRF07 BC included at the age of 25-29(VS.<24)(AOR=2.06,95%CI=1.08-3.91,P=0.028),at the age of 30-39(VS.?24)(AOR=3.76,95%CI=1.98-7.15,P<0.001),at the age of 40-49(VS.?24)(AOR=3.42,95%CI=1.56-7.52,P=0.002),age)50(VS.?24)(AOR=2.99,95%CI=1.03-8.70,P=0.045);Shenyang(VS.non Shenyang)(AOR=21.52,95%CI=9.24-50.11,P<0.001).In B network we used whether the research object is individual that outside the B network cluster is the outcome variable.In the univariate logistic regression analysis,we found that risk factors associated with sporadic spread of MSM of B is Shenyang(VS.non Shenyang)(OR=22.38,95%CI=5.26-95.19,P<0.001).Conclusions:1.In the study,we use method of retrospective study of regional,large sample and long-term observation,found that there are large cluster in three subtypes of HIV transmission network in Shenyang area.In recent years,the overall clustering rate of new diagnosed infection has been decreasing.The clustering rate of CRF01_AE and CRF07 BC virus strains have decreased,while the clustering rate of B subtype virus strains has increased.2.The analysis of the molecular transmission network suggests that from the perspective of the narrowing proportion of infection in the large/rapid cluster among the newly diagnosed infection,the reducing influence of the expanding cluster,the stable contribution of new cluster,and the increasing number of static cluster,China's active AIDS treatment and intervention measures have controlled the rapid spread of the HIV epidemic in Shenyang to a certain extent in recent years.3.Under the positive treatment intervention,the factors related to the further diffusion of HIV in Shenyang area include:married,divorced,individual business,at the age of 25 years old or above,Shenyang residents infection.We should focus on such people.4.For the analysis of related factors,more behavioral factors should be added to get more effective direction of intervention.