Antiretroviral (ARV) Drug Resistance and HIV-1 Subtypes among Injecting Drug Users in the Coastal Region of Kenya

HIV-1 genetic diversity results into the development of widespread drug-resistant mutations (DRMs) for the first-line retroviral therapy. Nevertheless, few studies have investigated the relationship between DRMs and HIV-1 subtypes among HIV-positive injecting drug users (IDUs). This study therefore determined the association between HIV-1 genotypes and DRMs among the 200 IDUs. Stanford HIV Drug Resistance Database was used to interpret DRMs. The five HIV-1 genotypes circulating among the IDUs were A1 (25 (53.2%)), A2 (2 (4.3%)), B (2 (4.3%)), C (9 (19.1%)), and D (9 (19.1%)). The proportions of DRMs were A1 (12 (52.2%)), A2 (1 (4.3%)), B (0 (0.0%)), C (5 (21.7%)), and D (5 (21.7%)). Due to the large proportion of drug resistance across all HIV-1 subtypes, surveillance and behavioral studies need to be explored as IDUs may be spreading the drug resistance to the general population. In addition, further characterization of DRMs including all the relevant clinical parameters among the larger population of IDUs is critical for effective drug resistance surveillance.


Introduction
e introduction of highly active antiretroviral therapy (HAART) in 1996 has resulted in improved treatment outcome and survival rate in human immunodeficiency virus-1 (HIV-1) infected patients [1][2][3]. e success of these drug regimens is being challenged by the emergence of drugresistant mutations [1]. Indeed, transmitted drug resistance generally leads to a delay in virologic suppression [4,5] and results in an increased risk of treatment failure [6]. Drug resistance testing and monitoring of HIV subtypes can improve treatment outcomes in infected individuals [7,8]. However, limited studies have been done on the high-risk group such as injecting drug users (IDUs). ese mutant variants have become increasingly widespread, in drugtreated and untreated individuals infected with HIV, and have compromised the therapeutic options of drug-naïve infected people [6,9].
Studies in Kenya have reported the occurrence of HIV drug resistance upon ART failure in Mombasa [10], Mombasa and Nairobi as part of a multisite African study [11], and Burnt Forest, a rural Academic Model Providing Access to Healthcare (AMPATH) clinic [12]. HIV-1 infection is highly diverse with the circulation of subtypes A (50-80%), D (10-20%), and C (5-15%) and multiple recombinants (10-20%) [13,14]. Extensive genetic heterogeneity is driven by several factors, such as the lack of proofreading ability of the reverse transcriptase (RT) [15], the rapid turnover of HIV-1 in vivo [16], host-selective immune pressures [17], leading to drug resistance selection pressure, and recombination events during replication [18]. HIV type 1 is divided into groups M, N, O, and P, more than 90% of HIV infections are derived from HIV-1 group M, and the rest are minor groups [19,20]. e M group is subdivided further into clades, called subtypes, which are also given letters ranging from A to K. Subtype A has been subdivided into A 1 , A 2 , A 3 , A 4 , A 5 , and A 6 , while subtype F has been subdivided into F 1 and F 2 [21]. is study therefore assessed the relationship between HIV-1-circulating genotypes and drug-resistant mutations among IDUs.

Study Setting.
A cross-sectional survey was conducted among HIV-positive injecting drug users in Malindi Sub-County, coastal region of Kenya.
e study enrolled 200 IDUs who consented and fulfilled eligibility criteria of being HIV infected, 18 years old and above, actively injecting drugs for the past six months, and were able to respond to structured questionnaires during interviews.

Participant Recruitment and Administration of
Questionnaires. Snowball sampling was adopted since IDUs are hard to reach the population. A set of initial participants referred to as "seed" for an expanding chain of referrals. All the participants were provided with written informed consent, and only those who consented were recruited into the study. Each participant was assigned a confidential identification number. e reliability of the questionnaire was pretested on 20 respondents (10%) and revised based on their feedbacks. e structured questionnaires were then administered by a qualified counselor in a private room using face-to-face interviews for approximately 10 minutes. Sociodemographic information was also collected during the interviews.

Ethical Consideration.
is study sought ethical approval from the Scientific Ethical Review Unit (SERU)-KEMRI (SSC no. 1438). Furthermore, permission was also sought from Jaramogi Oginga Odinga University of Science and Technology (JOOUST). Voluntary and written informed consent was obtained from the study participant before being allowed to take part in this study. Furthermore, the data collected from this study were confidential and only used for the purpose explained in the consent forms. Participation in this study presented no life-threatening risks.

Sample
Collection. Blood samples were collected from HIV-positive IDUs with and without prior exposure to firstline antiretroviral therapy for HIV-1 subtypes and drugresistant mutation analysis. About 5 ml of whole blood was drawn from each participant, separated into plasma, and transported to Kenya Medical Research Institute (KEMRI) under the cold chain for serological HIV-1 testing and molecular analysis. HIV colloidal gold, a rapid test for antibody to HIV, was used according to guidelines by the Ministry of Health for adult HIV testing [22,23]. Whole blood was spun at 3,000 rpm for 3 minutes, and plasma aspirated aseptically and stored at −80°C for subsequent RNA extraction.

RNA Extraction and
Genotyping. RNA was extracted from 140 µl of plasma using a QIAmp viral RNA kit according to the manufacturer's instructions (Qiagen Inc., USA). A nested polymerase chain reaction (PCR) was performed using AmpliTaq Gold (Roche Molecular Systems, Branchburg, NJ) [24]. PCR products of correct size were confirmed by agarose gel electrophoresis, purified, and sequenced by dideoxynucleoside-based analysis using a Big-Dye terminator kit (Applied Biosystems) and ABI Prism 3100 equipment (Applied Biosystems, Foster City, US) [25].

Statistical Methods.
Chi-square (χ 2 ) tests were done to establish the relationship between HIV-1 subtypes and drug resistance using Stata software version 13.0. Statistical significance was established at the 95% confidence limit within a marginal error of 0.05. Demographic data were analyzed using descriptive statistics; frequency (n) and percentage occurrence (%) of variables were generated using crosstabulations.

Sociodemographic Characteristics of Study Participants.
is study involved 200 injecting drug users with 120 (60%) males and 80 (40%) females. 55% (n � 111) of the study participants were born in Malindi, 38% (n � 76) were from Kilifi, while 99% (n � 198) of the study participants were selfemployed. A high proportion of this study population was unmarried (70.5%, n � 141), and 85% (n � 171) had a primary level of education. 70.5% (n � 141) were single with only 23.5% married (n � 47); the rest were divorcees (6.0%, n � 12) (Table 1). e mean age of the study population was 33.36 years with a range of 19 to 82 years. Out of the extracted 200 RNA samples, 79 DNA samples were amplified, and 47 samples were successfully sequenced. e drug-resistant mutations were observed on 23 out of the 47 samples (48.9%).

HIV-1 Genotypes
Circulating among the IDUs. As shown in Figure 1, a total of 5 HIV-1 subtypes, namely, A 1 , A 2 , B, C, and D, were observed. e HIV-1 subtype A 1 was the most abundant (25 (53.2%)). Both subtypes C and D had a similar frequency of 19.1% compared to A 2 and B in which both were 4.3%.

Drug-Resistant Mutations (DRMs) among Different HIV-1 Subtypes.
Of the 47 study participants, 23 (48.9%) had the drug resistance as shown in Table 2. HIV-1 subtype A 1 had the highest proportion of drug-resistant mutations (52.2%, n � 12), followed by both C and D which had 5 (21.7%) mutations. However, there was no significant association between the HIV-1 genotype and drug-resistant mutations (χ 2 � 2.9752; P � 0.704). Other HIV-1 subtypes such as A 2 and B had very low frequency of drug-resistant mutations.
E138A mutant genes had a high frequency of occurrence (3 times) in both NRTIs and NNRTIs. However, K103N mutations were more implicated in NNRTIs (4 times) than in NRTIs (once), while the individuals taking NNRTIs had also high chances of developing K103N, K219KN, and Y181C mutations (Table 3).
ese results further showed that HIV-1 subtype B exhibited no drug-resistant mutation.
It was observed that 34.8% (n � 8) of the 23 samples that exhibited major resistant mutations were ART-naïve individuals; they were drug-resistant de novo IDUs. ese individuals were still on Septrin at the time of sample collection. Among the drug-naïve participants, it was established that HIV-1 subtype A 1 exhibited most of mutant genes affecting both NRTIs (K70KN, D67N, K70R, V75VI, M184LV, K219HQ, Y115YF, K65KE, and L74L * W) and NNRTs (V106VI, V179IL, and G190AS). NTRIs were observed to be the most resisted group of ART in this study. Subtype B had no mutant gene among the drug-naïve IDUs (Table 5).
Among ART-experienced IDUs, this study established 65.2% (n � 15) samples that showed major resistant mutations. It was revealed that those who had been on ART between 1 and 5 years had more mutations than those less than one year and also above five years. It was also observed that HIV-1 subtype A 1 had the most abundant drug-resistant mutations. Mutant gene E138A was commonly observed in subtype C (Table 6).

Discussion
e study shows five HIV-1 genotypes circulating with the high proportion of drug-resistant mutations observed in HIV-1-A 1 subtype among the IDUs. As much as our study shows a number of drug-resistant mutations in HIV-1-A 2 , a study conducted by Songok et al. [26] indicated that this subtype is rare in Kenya. Our study has also shown the presence of HIV-1 subtype B among the IDUs that had previously been reported to be predominantly found in the USA, Europe, Australia, ailand, and Brazil [27,28].
Moreover, earlier studies documented that HIV-1 subtypes are not randomly distributed among the globe and show distinct geographical distribution [29]. Subtypes A and D are the most dominant in Africa; subtype B in the USA, Europe, Australia, ailand, and Brazil; subtype C in South Africa, Ethiopia, and India; F in some regions of Central Africa and Eastern Europe; and HIV-1-circulating recombinant form consisting of CRF01_AE in Southeast Asia [27]. e findings suggest a possible importation of rare subtypes, and this can be attributed to the commercial sex       [31]. However, contrary results were reported in Uganda which established HIV-1 seroconverts infected with subtype A (15%) and D (59%), suggesting predominance of subtype D [32]. In Brazil, among 69 IDUs' samples amplified, 52 (75%) were identified as HIV-1 subtype B, 15 (22%) as subtype C, and 2 (3%) as subtype F (de Martínez et al. [32]). Kiwanuka et al. [33] reported that, globally, subtype C is the most successful of the HIV-1M lineages and accounts for >50% of infections, whereas subtypes A and B each account for over 10% of worldwide HIV infections. Subtypes D and G, CRF01_AE, and CRF02_AG account for only between 2% and 6% each. e difference in the proportions of HIV-1 subtypes suggests that it may be a driver of HIV-1-resistant mutation distribution in the population [34]. It is interesting that one HIV subtype can exhibit regional predominance. Perhaps, this subtype A has evolved to be more virulent than other strains due to improved replication fitness which may explain why it is more abundant.

Advances in Virology
Our study further shows that HIV-1 subtype A had significantly higher drug resistance followed by both C and D mutations. e finding is in agreement with other studies that reported having either HIV-1 subtype A or D is not associated with the acquisition of drug-resistant mutations [35]. However, M184V/I was significantly more common in subtype A as compared to subtype D. e absence of resistance mutation in subtype B is in contrast with other results [36] which established that the distributions of M184V/I were significantly associated with subtype B.
In the recent past, antiretroviral drugs were developed and efficacy tested with HIV-1 subtype B, and clinical effectiveness and pattern of drug resistance among subtype-Binfected individuals were established (Chaplin B. et al., 2011). Genetic differences between subtypes might impact the drug resistance pathways (Ode H. et al., 2007). It has been demonstrated that the mutational pathway to drug-resistant mutations to NRTI drugs may vary among different HIV-1 subtypes. However, the mechanisms and reason as to why this happens are yet to be fully assessed (Dumans A. T. et al., 2009). is could explain partly the absence of drugresistant mutations in the HIV-1 B subtype in Kenya.
ere was frequent occurrence of A62AV mutant genes in NRTIs followed by E138A mutation and A98G mutant genes in NNRTIs followed by K103N. ese observations suggest that individuals on NRTIs are highly likely to develop A62AV and E138A, while those on NNRTIs are more likely to develop A98G and K103N mutant genes. Similarly, a study in Ghana among women with a history of prophylaxis recognized K103N and A98G as the dominant NNRTI mutations [36]. Furthermore, NNRTI mutations observed among drug-inexperienced individuals were K103N, V106A, and E138A with one minor drug resistanceassociated mutation as A98G [36,37]. Furthermore, HIV-1 subtype A exhibited most of mutant genes among drugnaïve participants affecting both NRTIs (K70KN, D67N, K70R, V75VI, M184LV, K219HQ, Y115YF, K65KE, and L74L * W) and NNRTs (V106VI, V179IL, and G190AS). ese findings show that NTRIs were the most resistant type of ART. However, studies in China revealed high percentages of NRTI (M184I/V) and NNRTI (K103N/S and Y181C/ I) mutations in subtype B. High percentages of M184I/V (26.3%) and K103N/S (39.5%) were found in subtype B strains in ART-naïve individuals [38][39][40].
We have observed DRMs among the naïve and those using the drugs, and we have also observed more DRMs among those who have used ART for a long time. ese suggest that the DRMs in ART-naïve IDUs could be due to acquisition of these mutations from patients failing treatment with resistant strains, prior exposure to ART, or undisclosed ART. is observation is in agreement with a recent study by Barik et al, (2021). Characterization of drug resistance mutations in RT gene of HIV-1 subtype C-infected individuals revealed that mutation M184V (63.15%) associated with lamivudine and abacavir and K103N (36.84%) identified in first-line ART failure in patients could be due to acquisition of these mutations from those failing treatment with resistant strains, prior exposures to ART, or Advances in Virology undisclosed ART (41). is study is among the few that has investigated antiretroviral drug-resistant mutations and HIV-1 subtypes among hard-to-reach population of injecting drug users. However, snowball sampling approach used in this study is not representative and is prone to biases. Also, due to failure to consider the clinical parameters such as CD4 + cells and viral load, some samples failed to amplify. is is perhaps due to low copies of the virus that could not be detected by the nested RT-PCR. We however note that as much as the failure to amplify reduced the sample size, the findings are consistent with other studies.

Conclusion
All the HIV-1 subtypes have the drug-resistant mutations except subtype B. With drug-resistant mutations across all the HIV subtypes, there is a need to enhance the surveillance and more behavioral studies as IDUs may act as a source of drug resistance to the general population.

Data Availability
All data generated and analyzed are included within this research article.

Ethical Approval
Approval was obtained from the Scientific Ethical Review Unit (SERU)-Kenya Medical Research Institute (SSC no. 1438). e permission for the study was sought from Jaramogi Oginga Odinga University of Science and Technology (JOOUST). e procedures used in this study adhered to the standards of the Declaration of Helsinki.

Consent
Participation was solely voluntary upon consent. e research project obtained written informed consent from all participants after clearly introducing the purpose of the study and the right of the participant to refuse to participate or withdraw from the study at any time. No gifts or inducement was used to influence participation. Information regarding procedures of the study was provided to the participants including how the data were to be handled and used. e participants were taken through the informed consent document to ensure that they understood. Participation in this study presented no life-threatening risks. e authors affirm that participants provided informed consent for the publication regarding their individual data.

Conflicts of Interest
e authors declare that they have no conflicts of interest.