[Analysis on characteristics of HIV-1 molecular network in injected drug users in Guangzhou, 2008-2017]

Objective: To understand the characteristics of the HIV-1 molecular networks in injected drug users (IDUs) in Guangzhou from 2008 to 2017, and provide reference for the prevention and control of AIDS in this population. Methods: The serum samples of newly diagnosed HIV-1 positive IDUs in Guangzhou from 2008 to 2017 were collected, HIV-1 RNA was extracted for pol gene amplification and sequencing. The molecular cluster in the phylogenetic tree was identified by Cluster Picker 1.2.3 for cluster analysis. TN93 model in HyPhy2.2.4 was used to calculate the gene distance between the cluster sequences. Software Cytoscape3.8.2 was used to visualize the molecular network, and χ² test or exact probability method was used for cluster analysis and centrality analysis. Results: A total of 586 sequences were successfully amplified (73.9%, 586/793), and 80 molecular clusters were produced, with a clustering rate of 46.6% (273/586). In molecular clusters, the proportions of the strains from IDUs in Han ethnic group (48.4%, 260/537), IDUs who were local residents in Guangdong (52.7%, 146/277) and IDUs whose strain sequence subtype was CRF55_01B (93.3%, 14/15) were higher. In the molecular network, the degree range was 1-7, and nodes with degree ≥3 accounted for 12.8% (24/187), which was associated with another 81 nodes in the molecular network (43.3%, 81/187). The centrality analysis showed that the proportions of housework/unemployed with high degree centrality (19.0%, 19/100), high intermediary (22.0%, 22/100), and high proximity centrality (32.0%, 32/100) were higher in IDUs infected with HIV-1. Conclusion: The risk of HIV-1 clustering in IDUs in Guangzhou was high, suggesting that IDUs who were from both Guangdong and Guangxi and were house workers or unemployed should be viewed as the key targets, and precise intervention should be implemented to reduce the HIV-1 infection rate in this population.