INTRODUCTION
Since 1996 combination antiretroviral therapy (cART) has offered HIV/AIDS-infected individuals the possibility of increased longevity, reduced HIV/AIDS-related complications, and improved quality of life [Reference Michaels1–4]. At the end of 2011, 108 000 (14%) out of the estimated 780 000 people living with HIV/AIDS in China were on cART made available by the National Free Antiretroviral Treatment Programme (NFATP) [5, Reference Zhang6]. The number of HIV/AIDS-infected individuals on cART had significantly increased to 170 655 (22%) by the end of 2012. In the past decade, the NFATP has significantly reduced HIV/AIDS-related mortality and morbidity in China [Reference Zhang3, Reference Zhang7, Reference Dou8]. Nevertheless, amidst the rapid scaling up of cART there are concerns such as treatment non-compliance which might lead to incomplete suppression of HIV replication which in turn might lead to HIV drug resistance (HIVDR) [Reference Hirsch9, Reference Tamalet10]. High prevalence of HIVDR would restrict therapy options, compromise the effect of current therapy regimens, and increase the risk of treatment failure. This is particularly problematic in resource-limited settings such as China where access to protease inhibitors (PIs) and new classes of antiretroviral drugs (ARVs) is still limited.
The World Health Organization (WHO) recommends surveillance of HIVDR in resource-limited settings where cART is being scaled up [Reference Bennett11]. China first conducted surveillance of HIVDR in ART-naive HIV-infected individuals in 2004 [Reference Si12]. By the end of 2012, the prevalence of HIVDR in ART-naive individuals ranged from 1·6% to 12·2% in China [Reference Si12–Reference Li24]. However, such surveillance data are not available for the most developed and populous areas in eastern China where the HIV epidemic is relatively new but rapidly spreading through unprotected heterosexual and homosexual behaviours.
To fill this significant gap, we conducted a molecular epidemiological survey during 2008–2011 to examine the prevalence of HIVDR in all newly diagnosed HIV-infected adults in a costal prefecture of Zhejiang province, eastern China. The knowledge gained from this study will be valuable for designing effective cART programmes for HIV/AIDS individuals in this region.
METHODS
Study site
This study was conducted in Taizhou prefecture of Zhejiang province, a coastal region in eastern China, which has a total of 5·9 million residents (National Bureau of Statistics of China, unpublished data). The first HIV case was reported in 1996. By the end of 2011, 886 HIV-infected individuals had been diagnosed and registered with the Chinese National Information System for AIDS Prevention and Control (CNISAPC). Of these, 69·3% were infected via heterosexual sex, 15·8% infected via homosexual sex, and 7·3% infected via injection drug use [Reference Lin25].
Study samples
A total of 671 adults were confirmed to be HIV-1 positive during the period of January 2008 to December 2011, of whom 513 (76·4%) had a minimum of 200 μl cryopreserved plasma sample at the time of their HIV diagnosis while naive to antiretroviral treatment (ART) and were eligible for HIV genotyping to determine HIVDR. The other 158 (23·6%) HIV-infected adults were not included for HIV genotyping and further analyses because of loss to follow-up after HIV diagnosis (n = 97) or lack of plasma samples (n = 61). The study was approved by the Institutional Review Board of Fudan University, China.
Ethical statement
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
RNA extraction and amplification
Plasma was separated from the whole blood within 2 h after collection and stored at −70°C until use. Viral RNA was extracted from plasma samples using the High Pure Viral Nucleic Acid kit (Roche Inc., Germany) according to the manufacturer's instructions. The HIV-1 pol gene (protease 1–99 amino acids and part of reverse transcriptase 1–300 amino acids) was amplified by using an in-house nested reverse transcription–polymerase chain reaction (RT–PCR) method. The target sequence was amplified with TaKaRa One-Step RNA PCR kit (TaKaRa Biotechnology, China) using primers MAW26 (5′-TTGGAAATGTGGAAAG GAAGGAC-3′) and RT21 (5′-CTGTATTTCTGCTATTAAGTCTTTTGATGGG-3′) in a 25 μl reaction. Cycling conditions were 50°C for 30 min and 94°C for 5 min in first-round RT–PCR, followed by 35 cycles at 94°C for 30 s, 55°C for 30 s, 72°C for 2 min, and an extension at 72°C for 10 min. The nested PCR was performed using the Takara Ex Taq PCR kit (TaKaRa Biotechnology), using primers PRO-1 (5′-CAGAGCCAACAGCCCCACCA-3′) and RT20 (5′-CTGCCAGTTC TAGCTCTGCTTC-3′) in a 50 μl reaction and the cycling conditions were 94°C for 5 min in first-round RT–PCR, followed by 30 cycles at 94°C for 30 s, 63°C for 30 s, 72°C 2·5 min, and an extension at 72°C for 10 min.
Genotyping of HIVDR
Bi-directional sequences were obtained in the region of interest for all samples analysed in this study. For quality control, mutations presented as amino acid mixtures were identified only if the corresponding nucleotide mixture was present in the sequences of both DNA strands. The sequences of those amplified fragments were edited using the Lasergene software (SafeNet Co., USA). HIVDR and the susceptibility of the viruses to ART were interpreted with both the Stanford HIV Drug Resistance Database (http://hivdb.stanford.edu) and the Surveillance Drug Resistance Mutations (SDRM) list recommended by WHO [Reference Bennett26]. Furthermore, for each HIV-infected individual who was identified as having HIV genetic mutations that are not listed in the SDRM list, a cumulated mutation score (i.e. designated as the ‘Total Scoring’ in the Stanford HIVDR Database) was calculated by adding mutation scores (i.e. designated as ‘Mutation Scoring’ in the Stanford HIVDR Database) of all identified genetic mutations. To evaluate the possible impact of HIVDR transmission on the efficacy of future therapy with PIs, nucleoside reverse transcriptase inhibitors (NRTIs) and non-nucleoside reverse transcriptase inhibitors (NNRTIs), we further analysed the HIV-1 pol gene sequences using the French National Agency on AIDS and Viral Hepatitis (ANRS) algorithm.
HIV-1 subtyping
The determination of HIV-1 subtypes was accomplished by submitting the obtained pol sequences including the entire PR gene and a part of the RT gene (1316 bp) to the HIV BLAST subtyping tool (http://www.hiv.lanl.gov) for phylogenetic analysis. The subsequent detailed phylogenetic analysis was performed using Molecular Evolutionary Genetics Analysis (MEGA) software, version 5.1. The recommended HIV-1 subtype reference sequences available at the Los Alamos HIV sequences database (www.hiv.lanl.gov) were used. The phylogenetic reconstruction was performed using the neighbour-joining method based on the Kimura two-parameter distance model with 1000 bootstrapped datasets.
Statistical analysis
Statistical analysis was performed using SPSS software v. 17.0 (SPSS Inc., USA). Continuous variables were presented as mean±s.d. or median (interquartile range). Categorical variables were compared using χ 2 test or Fisher's exact test where appropriate. All tests were two-tailed and a P value <0·05 was considered as significant.
RESULTS
Characteristics of study participants
Of the 513 eligible study subjects, 447 (87·1%) were successfully genotyped with the HIV-1 pol gene. The failure of genotyping was primarily due to low plasma viral load and the inability to amplify viral RNA. Those who were genotyped and those who were non-genotyped were not significantly different in demographic characteristics (data not shown). Table 1 presents demographic characteristics of those who were genotyped, and these individuals were subjected to further analysis. The majority of them was male (73·2%), aged 18–40 years (66·7%), ethnic Han (94·0%), married (51·9%), and infected via heterosexual sex (69·8%).
HIV-1 subtypes
More than half (53·7%) of the genotyped study participants were infected with CRF01_AE, 20·1% with CRF07_BC, 12·5% with subtype B, 11·6% with CRF08_BC, 1·8% with subtype C and 0·2% with subtype A (Table 1). The distribution of HIV-1 subtypes did not vary by gender, age, and HIV report year, but did vary by HIV transmission route (χ 2 = 64·90, P < 0·001). For the 312 heterosexually infected participants, 47·1% were infected with CRF01_AE, 20·5% with CRF07_BC, 14·1% with subtype B, 15·4% with CRF08_BC, 2·6% with subtype C and 0·3% with subtype A. For the 103 homosexually infected participants, 77·7% were infected with CRF01_AE, 12·6% with CRF07_BC, 8·7% with subtype B and 1·0% with CRF08_BC. For 21 infected injection drug users (IDUs), 28·6% were infected with CRF01_AE, 61·9% with CRF07_BC, 4·8% with subtype B, and 4·8% with CRF08_BC. For the eight participants who were infected via blood transfusion, four (50%) were infected with CRF01_AE, two (25%) with subtype B and two (25%) with CRF08_BC. All of the three individuals with transmission route unknown were infected with CRF01_AE.
Prevalence of HIVDR
Twelve patients were determined to have low to high HIVDR according to WHO SDRM criteria. The overall prevalence of HIVDR using WHO SDRM criteria in the study participants was 2·7% (12/447). Six (50%) of participants were only resistant to NRTIs, four (33·3%) were only resistant to NNRTIs, one(8·3%) was resistant to both PIs and NNRTIs, and one (8·3%) was resistant to all three classes of ART.
Table 1 presents the prevalence of HIVDR by demographic characteristics of the participants. The prevalence was significantly higher in farmers than other occupations combined (4·4% vs. 1·2%, χ 2 = 4·15, P = 0·042). No other significant difference was observed.
Predicted phenotypic HIVDR
According to the Stanford HIV Drug Resistance Database, 38 (8·5%) patients were predicted to be potentially resistant to ART, including the above 12 patients determined to have low to high HIVDR according to WHO SDRM criteria. Of these individuals, 25 (65·8%) were aged 21–40 years, 36 (94·7%) were ethnic Han, 23 (60·5%) were married, and 32 (84·2%) were infected with HIV via heterosexual sex. The median CD4 cell count of the 38 patients was 355 cells/mm3.
Table 2 presents the detailed information of the 38 patients including their potential resistance to ART. Four (10·5%) harboured major PI mutations – M46I, V82A and L90M; 11 (28·9%) harboured minor PI mutations – A71 T, A71 V, L10F, L10I and L33F; 10 (26·3%) harboured NRTI-related mutations – A62 V, M184 V, T69N, T69S and T215S; and 23 (60·5%) harboured NNRTI-related mutations – G190A, K103N, V108I, V179D, V179E and Y188L, with V179D and V179E mutations occurring most frequently. As shown in Table 2, nevirapine (NVP; 21/38, 55·3%), efavirenz (EFV, 15/38, 39·5%) and rilpivirine (RPV, 8/38, 21·0%) were the drugs most likely to be less effective for treatment due to built-up HIVDR.
PIs, Protease inhibitors (ATV/r, atazanavir/r; FPV/r, fosamprenavir/r; IDV/r, indinavir/r; SQV/r, saquinavir/r; LPV/r, lopinavir/r); NRTIs, nucleoside reverse transcriptase inhibitors (DDI, didanosine; ABC, abacavir; 3TC, lamivudine; FTC, emtricitabine; AZT, zidovudine; D4 T, stavudine; TDF, tenofovir; NFV, nelfinavir); NNRTIs, non-nucleoside reverse transcriptase inhibitors (EFV, efavirenz; NVP, nevirapine; ETR, etravirine; RPV, rilpivirine).
* Samples nos. 1–12 were determined to have low to high level of phenotypic drug resistance, i.e. HIVDR.
Minor amino acid substitutions associated with HIVDR
Most of study participants were found to have one or more minor genetic mutations or amino acid substitutions in the HIV pol gene that are associated with drug resistance (Table 3). Of them, 65·8% had I93L, 60·2% had M36I and 48·3% had L63P. Such amino acid substitutions occurred differentially by HIV subtypes (Table 3).
DISCUSSION
The present study, for the first time, examined HIVDR in newly diagnosed HIV-infected cases in Zhejiang province, eastern China where the HIV epidemic is rapidly spreading through sexual transmission. The study indicates low prevalence (<5%) of HIVDR in the study participants according to WHO classification [Reference Myatt and Bennett27], thus suggesting that current cART regimens are likely to remain effective in the study area. Such a low prevalence of HIVDR has also been observed in many regions in China [Reference Si12–Reference Liao23]. The highest prevalence of HIVDR (12·2%) in ART-naive HIV patients was observed in Henan province where the HIV/AIDS epidemic started from commercial plasma donors in the early 1990s and large-scale cART was available earliest in the country [Reference Li24]. A confluence of factors may contribute to such a relatively low prevalence of HIVDR in newly diagnosed or ART-naive HIV patients, for example, the relatively low coverage and short duration of NFATP as well as a success of first-line cART. It should be noted that the prevalence of HIVDR in the present study area was lower than an earlier study (2006–2007) of 145 treatment-naive HIV patients (7·6%) [Reference Ye15]. One possible explanation is that the 145 HIV patients in that study were from 22 provinces where the HIV epidemic as well as the coverage and duration of cART varied substantially.
The majority of HIVDR in our study are associated with reverse transcriptase inhibitors (RTIs). This finding reflects the fact that RTIs, usually two NRTIs + one NNRTI such as AZT + 3TC + EFV or AZT + 3TC + NVP, are the most widely used regimen in China as first-line cART. Previous reports demonstrate that HIV can develop a wide range of mutations resistant to RTI if viral suppression is incomplete [Reference Casado28–Reference Cheung, Wynhoven and Harrigan30]. On the other hand, some of the HIVDR in the study are associated with PIs, such as atazanavir (ATV), nelfinavir (NFV), fosamprenavir/r (FPV/r), indinavir/r (IDV/r) and saquinavir/r (SQV/r), which had not been introduced into the study area. These data suggest that HIV drug-resistant isolates from other countries may have ‘slipped through’ into eastern China or there are polymorphisms in subtypes prevalent in the region, which has implications for these individuals relying on the national ART programme. Moreover, six patients were identified to be resistant to multiple ART in the study. One of whom was found to have multiple HIV drug-resistant mutations (L90M, T215S, Y188L) that are associated with all the three classes of ART. The transmission of multiple HIVDR may have severe public health consequences and deserves more close scrutiny in the near future.
HIV mutations and amino acid substitutions associated with resistance to PIs and RTIs have been extensively characterized in subtype B [Reference Metzner31–Reference Hammond33], which is predominant in developed countries. However, such information is still relatively lacking in non-B subtypes, particularly in China. In this study, more than half of the study participants were infected with CRF01_AE, only a few (1·7%) of whom were infected with HIVDR. However, a relatively high prevalence of HIVDR was observed for CRF08_BC (3/52, 5·8%) and subtype B (3/56, 5·4%), although these two subtypes only accounted for 24% of the participants. CRF08_BC mainly conferred resistance to NRTIs, whereas subtype B mainly conferred resistance to NNRTIs.
Some minor amino acid substitutions associated with HIV drug resistance were identified in the study, which occurred differentially by HIV subtypes. Amino acid substitutions such as V77I and A71 V/T in protease were frequently found in subtype B. This observation was consistent with previous reports in China [Reference Han17]. M36I accounted for 95·0% of the amino acid substitutions occurring in CRF 01_AE in this study. The improved viral replication offered by M36I might favour a more rapid spreading of non-B subtypes of HIV-1 [Reference de Felipe34–Reference Tee, Kamarulzaman and Ng36]. I93L was primarily associated with non-A subtypes [Reference Tee, Kamarulzaman and Ng36, Reference Liu37], which was also observed in the present study. A62 V and T69N/S, both conferring resistance to NRTIs, were only found in CRF08_BC. Given the complexity of HIV subtypes circulating in this area and substantial differences in the profile of drug resistance mutations between the subtypes, it is of particular importance to implement HIV molecular surveillance programmes, especially HIVDR surveillance in newly HIV-infected individuals in the study area as well as other areas in China where multiple HIV genotypes are circulating. Knowledge gained from such programmes would be very useful for understanding HIV dynamics and designing optimal cART programmes in targeted programme sites.
Some limitations should be noted. First, all study participants were tested for HIVDR at the time of HIV diagnosis. However, due to low practice of HIV testing and delayed HIV diagnosis, many of the study participants might have been living with HIV for several years at the time of their HIV diagnosis. Thus, the possibility that the HIVDR identified in the study was not transmitted from other HIV-infected individuals with existing HIVDR but was developed by themselves after exposure to ART could not be ruled out, although such a possibility might be low given that the HIVDR was detected at the time of their HIV diagnosis and the availability of ART has been extremely limited for individuals without HIV infection in China. Second, given the relatively low prevalence of HIVDR in newly diagnosed HIV-infected individuals in the study area, the present study sample is not large enough to allow for more specific analysis of HIVDR. Future larger studies especially large longitudinal cohort studies are warranted.
ACKNOWLEDGEMENTS
This work was supported by the Shanghai Municipal Health Bureau (grant no. XBR2011043) and the Chinese National Major Science & Technology Project of Infectious Diseases (grant no. 2012ZX10001001-004). Preparation of this article was supported in part by the Emory Center for AIDS Research (P30 AI050409; Wong).
DECLARATION OF INTEREST
None.