Epidemiological surveillance of HIV-1 transmitted drug resistance among newly diagnosed individuals in Shijiazhuang, northern China, 2014–2015

Background The widespread use of antiretroviral therapy (ART) has led to considerable concerns about the prevalence of transmitted drug resistance (TDR). Sexual contact, particularly men who have sex with men (MSM) was the most prevalent form of HIV transmission in Shijiazhuang. Hence, we conducted an epidemiological surveillance study on TDR among newly diagnosed individuals who infected-HIV through sexual contact in from 2014–2015. Methods Genotypic resistance mutations were defined using the WHO-2009 surveillance list. Potential impact on antiretroviral drug was predicted according to the Stanford HIV db program version 7.0. The role of transmission clusters in drug resistant strains was evaluated by phylogenetic and network analyses. Results In this study, 589 individuals were recruited and 542 samples were amplified and sequenced successfully. The over prevalence of TDR was 6.1%: 1.8% to nucleoside reverse transcriptase inhibitors (NRTIs), 2.0% to non- NRTIs (NNRTIs) and 2.4% to protease inhibitors (PIs), respectively. We did not find significant differences in the TDR prevalence by demographic and clinical characteristics (p > 0.05). Using network and phylogenetic analysis, almost 60.0% sequences were clustered together. Of these clusters, 2 included at least two individuals carrying the same resistance mutation, accounting for 21.2% (7/33) individuals with TDR. No significant difference was observed in the clustering rate between the individuals with and without TDR. Conclusions We obtained a moderate level TDR rate in studied region. These findings enhance our understanding of HIV-1 drug resistance prevalence in Shijiazhuang, and may be helpful for the comprehensive prevention and control of HIV-1.


Results
In this study, 589 individuals were recruited and 542 samples were amplified and sequenced successfully. The over prevalence of TDR was 6.1%: 1.8% to nucleoside reverse transcriptase inhibitors (NRTIs), 2.0% to non-NRTIs (NNRTIs) and 2.4% to protease inhibitors (PIs), respectively. We did not find significant differences in the TDR prevalence by demographic and clinical characteristics (p > 0.05). Using network and phylogenetic analysis, almost 60.0% sequences were clustered together. Of these clusters, 2 included at least two individuals carrying the same resistance mutation, accounting for 21.2% (7/33) individuals with TDR. No significant difference was observed in the clustering rate between the individuals with and without TDR. PLOS

Introduction
The widespread use and increased coverage of ART has reduced significantly the risk of HIV transmission and decreased HIV-related morbidity and mortality [1]. Meanwhile, a major drawback of the global increase in ART access is a corresponding increase in HIV drug resistance, which can be transmitted to newly infected individuals. TDR of HIV has gradually become a particular concern because it has the potential to compromise the efficacy of combination ART regimens and may lead to the failure in first-line ART [2][3][4]. Previous published studies on TDR investigated in different countries showed that the rates of TDR differs between regions and time period, and the overall rate varies ranging from 0% to 23% in treatment-naïve individuals [5][6][7][8][9]. Although the prevalence of TDR among treatment-naïve HIV-1-infected individuals remains low in most areas of China, some recent reports indicated moderate levels in specific regions [10][11][12].
Our study focuses on Shijiazhuang city, which is the capital of Hebei province, located in north China, adjacent to Beijing and Tianjin, and Henan province. The first case of HIV-1 infection in Shijiazhuang was diagnosed in 1989, and 2347 HIV/AIDS cases had been cumulatively reported in this region by the end of 2015. The prevalence of HIV infection in the general population in Shijiazhuang city remains at a relatively low level [13]. Data from the sentinel surveillance showed that HIV epidemic among the general population was 0.04% in 2015, but it is spreading rapidly in one or more specific groups in recent years. Currently, sexual contact is the most prevalent form of HIV transmission in China. According to the data of the case registry system, the proportion of sexual transmission has been increasing in the newly reported HIV/AIDS cases, rising from 52.3% in 2005 to 96.7% in 2015 in Shijiazhuang. Furthermore, the trend was particularly apparent for it in men who have sex with men (MSM), which increased from 14.3% in 2005 to 65.7% in 2015.
In China, ART is provided to patients for free through the "Four Free One Care" policy since 2004 [14]. Hebei province was also among the first group of free ART provinces in that year. In 2015, the new guideline for ART has been changed to recommend all HIV-diagnosed individuals to receive ART [15]. Up to now, free ART has been available in Shijiazhuang for 13 years. However, there is little data on TDR and other molecular epidemiological data about the HIV epidemic among newly diagnosed, sexually infected individuals in this region. Therefore, we performed an epidemiological surveillance study on TDR among individuals newly diagnosed of HIV-1 infected through sexual contact in Shijiazhuang and aimed to provide these baseline resistance data to guide the choices of initial regimen so as to better clinical management and broader disease control effort.

Study population
Participants were enrolled if they were diagnosed as HIV seropositive for the first time and infected through sexual contact between January 2014 and December 2015. On obtaining a participant's informed consent, peripheral blood sample was drawn in an EDTA-added vacutainer tube. CD4 + T cell counts were detected using whole blood and plasma samples were aliquoted and stored at -80˚C until use. Demographic and behavioral information of participant was collected via counselor's face-to-face interview.

CD4 + T cells count and HIV-1 viral loads detection
Blood samples were collected within 3 months after infection had been confirmed. According to the manufacturer's instructions, absolute CD4 + T cell count was performed by flow cytometry (FC500, Beckman, USA) within 24 h after blood sample processing using the TRITEST three-color CD4/CD8/CD3 reagent and TRUCOUNT tubes (Becton Dickinson, San Jose, CA). Plasma HIV-1 RNA viral load was determined with NucliSens kit (Biomerieux, France) according to the manufacture's recommendation.

Drug resistance mutations analysis
All sequences were screened utilizing Calibrated Population Resistance (CPR) tool version 6.0 available at Stanford University HIV Drug Resistance Database (http://cpr.standford.edu/cpr/ index.html), and then resistance mutations to NRTIs, NNRTIs, and PIs were defined as the detection of at least one mutation in any drug class according to the WHO surveillance drug resistance mutations (SDRMs) list updated in 2009 [17,18]. The overall prevalence was defined as the percentage of patients harboring any mutations indicative TDR. To predict the susceptibility to available NRTIs, NNRTIs, and PIs, sequences were analyzed using the Stanford HIV db algorithm version 7.0 [19]. Three levels of resistance were scored: high (R), intermediate (I, intermediate and low), sensitive (S, potentially resistant and sensitive).

Phylogenetic analyses and network construction
To avoid the effect of homoplasy of drug resistance mutations on the phylogenetic analysis, all 43 codons associated with major DRM in PR and RT were removed from all of the sequences within the alignment. phylogenetic sequences analysis was estimated via a maximum likelihood (ML) approach with a bootstrap analyses with 1000 replicates using the general time reversible + Gamma (GTR + γ) model of nucleotide substitution in FastTree version 2.1. Transmission cluster were defined as those comprising virus strains from 2 or more individuals whose sequences clustered in phylogenetic trees with a bootstrap value ! 70%. The trees were edited and visualized using FigTree version 1.4.2.
To construct the transmission network, sequences were codon-aligned to an HXB2 reference sequence (positions 2253-3749). Forty-eight codons associated with drug resistance were removed from the alignment. Tamura-Nei 93 (TN93) pairwise genetic distance was calculated for all pairs of sequences. The TN93 genetic distance between two sequences 1.5% was identified as the potential transmission partners. For visualizing and analyzing network, the network data were processed using software Cytoscape 3.5.2. We described characteristics of the network, including the number of sequences (nodes), links (edges) and clusters (groups of linked sequences) [20][21][22][23].

Statistical analysis
Categorical variables were reported in both numbers and proportions, while continuous variables were calculated as medians ± standard deviation (SD), respectively. χ 2 tests and logistic regression were used for categorical variables. A two-tailed p values less than 0.05 was considered significant. All data were analyzed using Statistical Analysis System version 9.1 (SAS Institute Inc., USA).

Characteristics of the study population
All of 589 patients in the study were newly diagnosed in the year of 2014-2015 and infected through sexual contact. Of all subjects, the majority were male (92.9%, 547/589) and MSM (77.6%, 457/589). The patients studied were of ages ranging from 16 to 78 years with age of 36.5±13.7 years old. 96.9% individuals in this study were Han ethnicity and 82.7% were permanent residents in Shijiazhuang. At enrollment, the mean baseline of CD4 + T cell counts was 328±202.3 cells/μL, and the mean of HIV viral RNA load was 4.33±0.9 log copies/mL. The proportional distribution of 589 patients analyzed by sex, age, transmission routes, permanent residence from sociodemographic data were presented in Table 1.

Factors associated with TDR
Comparisons between patients infected with HIV-1 strains harboring drug resistant mutations and those without were shown in Table 2. No differences were observed between the prevalence of TDR and sex, age at diagnosis, nationality, marriage status, education level, permanent residence, Risk behavior, and occupation (all p values > 0.05). Both the mean CD4 cell count and the logarithmic mean of viral load in TDR group were slightly lower than those in without TDR group, although the difference was not statistically significant (p values > 0.05). In addition, no significant different prevalence of TDR were observed among CRF_01AE, CRF_07BC, subtype B, and others subtypes (P = 0.68).

Phylogenetic and drug resistant-associated transmission networks analysis
Phylogenetic analysis showed that 325 (60.0%) of 542 sequences of this study grouped in 72 transmission clusters composed of 2 or more individuals. 15 (45.5%) out of 33 sequences with TDR were included in 8 clusters. of which, 2 clusters included at least two individuals carrying the same resistance mutation, the M46I/L was one of the sharing mutation, and K101E was the other (Fig 2). The prevalence of TDR was not significantly different among individuals who were part of clusters and who were not [4.6% (15/325) vs. 8.3% (18/217), respectively; P = 0.1].
To explore the effect of TDR on viral transmission based on the transmission network, we also construct transmission network. Similar to the phylogenetic analysis, 57 clusters were identified, of which 11 contained at least one individual with TDR. Among 542 individuals in the network, 301 (55.5%) were potential transmission partners with another, and 18 (54.5%) individuals harboring mutations of drug-resistance were included (Fig 3). No significant difference was observed in the clustering rate between the individuals with and without TDR (P = 1.0).

Discussion
In the present study, we analyzed the HIV-1 pol gene sequences of 542 patients with newly diagnosed HIV-1-infection through sexual contact and treatment-naïve in Shijiazhuang City Results of phylogenetic analysis illustrated the diversity of HIV-1 subtype among the studied subjects. A total of nine HIV-1 subtypes, including CRF01_AE, CRF07_BC, B, CRF08_BC, CRF02_AG, CRF55_01B, CRF59_01B, and CRF65_cpx, and URFs, were detected. CRF01_AE, CRF07_BC, and subtype B represented the three most common subtypes among sexual contact population, similar to what has been shown in other reports [24][25][26][27]. In addition, CRF59_01B and CRF65_cpx were firstly identified in our samples in Shijiazhuang. These finding indicated the highly genetic heterogeneity and subtypes/CRFs diversity among sexual The prevalence of transmitted drug resistance in Shijiazhuang, China contact population epidemic in this city, which may facilitated the more generation of new recombination forms and made intervention more complicated.
The overall prevalence among 2014-2015 studied samples was 6.1% in Shijiazhuang city, a moderate level (5%-15%) according to the WHO categorization method. The TDR rate was slightly higher than that of studies in other region in China [26,28,29], while lower than that previous published studies in some industrialized countries. The higher prevalence of TDR in developed countries may be due to the fact that the patients there have been subject to more prolonged periods of ART in these countries [30,31]. When examining the presence of resistance by drug class, we found that there was no significant difference in the prevalence of drug-resistant mutations among the three antiretroviral classes (NRTI, 1.8%; NNRTI, 2.0%; and PI, 2.4%). For PIs mutations, 91.7% of M46I/L (11/12) was detected in CRF01_AEinfected MSM in our study. Although the M46L mutation can decrease susceptibility to a number of protease inhibitors when present with other mutations, it is also a natural polymorphism in CRF_01 AE. It is likely that the occurrence of M46I/L may be the characteristics of specific viral, and may not have any particular impact on the success of current ART [32][33][34]. In NNRTIs class, high-level resistant to Efavirenz and Nevirapine were mainly due to the The prevalence of transmitted drug resistance in Shijiazhuang, China mutations K103N, V106M and Y181C. For NRTIs, 0.6% of patients were predicted to be resistant to Lamivudine and Emtricitabine, which is mostly attributable to the presence of M184V. Lamivudine, Efavirenz and Nevirapine have been prescribed in the first-line regiment of ART in China [10,35]. We therefore considered that the emergence of these mutations may relate to the usage of first-line regimen.
No significant differences were observed between the prevalence of TDR and demographic characteristic, which was similar with a previous study [36,37]. Many studies reported that the higher prevalence of TDR among MSM compared to heterosexuals; interestingly, the opposite results also were found in others reports [38][39][40]. In this study, the overall prevalence of TDR was not significantly different between MSM and heterosexual. Since the majority of newly diagnosed HIV-1-infected individuals were MSM (78%) in Shijiazhuang, TDR surveillance in this group should be performed to better inform treatment decisions.
This study had a clustering rate with 60% of sequences segregating 72 clusters using phylogenetic analysis. with regard to the transmission of resistant variants in transmission clusters, only 2 transmission clusters were identified that contained drug-resistant individuals having a transmission relationship with each other. The remaining individuals carrying TDR in the network were independent. There was no closely correlation between the clustering rate and the spread of TDR. In previous study [41], the contribution of transmission cluster to the spread has been reported at frequency of 20% in newly diagnosed patients, similar to our study results. The prevalence of transmitted drug resistance in Shijiazhuang, China In summary, we focused on the prevalence of TDR among individuals who were newly diagnosed and infected-HIV through sexually contact in Shijiazhuang in 2014-2015. The overall prevalence of TDR was 6.1%, which distributed almost equally among the three drug classes (NRTI, NNRTI and PI) as a moderate level. Although TDR was identified in the transmission clusters, most of TDR were independent rather than converged together. These findings enhance our understanding of HIV-1 drug resistance prevalence and subtype distribution in Shijiazhuang.
Supporting information S1