Secondary Logo

Journal Logo

BASIC SCIENCE

Resistance profile of HIV-1 quasispecies in patients under treatment failure using single molecule, real-time sequencing

Monaco, Daniela C.a; Zapata, Lucasb; Hunter, Erica,c; Salomon, Horaciob; Dilernia, Dario A.a,c

Author Information
doi: 10.1097/QAD.0000000000002697

Abstract

Introduction

Antiretroviral therapy (ART) has transformed the HIV epidemic by improving life expectancy of people living with HIV [1–6], and reducing transmission [7–11]. Development of more diverse and effective antiretroviral drugs has facilitated a sustained control of viral replication and driven a switch in treatment guidelines towards earlier initiation of treatment to the point in which ART is currently recommended for all HIV-positive individuals, including those with early infection [12]. However, selection of drug resistance is still a major concern, both for treatment outcome as well as from a public health perspective [13,14]. In fact, the increase in prevalence of resistance mutations among the circulating viruses leads to changes in treatment guidelines [15,16], with the main concern around NNRTI resistance mutations that have been shown to be circulating at increasing frequencies [16–18].

HIV resistance to antiretroviral drugs is classified as primary or secondary, based on whether the mutations were acquired at the moment of infection or selected during ART, respectively. Primary resistance is a concern as it can impact efficacy of the first-line antiretroviral regimen, whereas secondary resistance mediates treatment failure. Although characterization of the transmitted virus in newly infected individuals provides an accurate idea of which resistance mutations can be expected as primary resistance [19–29], secondary resistance is significantly more complex as it depends on multiple factors, such as the antiretroviral drugs that were used in the regimen, the previous regimens that might have selected other mutations, and the time that the patient has been under suboptimal treatment [30–41]. In fact, while under suboptimal treatment, the persistence of drug levels keeps the resistance mutations selected at high frequency in the patient under treatment failure and might drive selection of new resistance mutations [42–47]. Considering this, a deeper understanding of the resistance profile of the viral quasispecies during treatment failure can provide valuable insights for a better management of HIV infection.

In the present study, we defined the resistance profile of the HIV variants present in patients under treatment failure by implementing single molecule, real-time (SMRT) sequencing technology. Next-generation sequencing (NGS) has been previously implemented to study the resistance profile of HIV as it provides a higher sensitivity compared with Sanger sequencing to detect low-frequency resistant variants, and the high-throughput nature of this technology makes it particularly suitable for resistance surveillance programs [48,49]. Virtually all the NGS-based HIV drug resistance tests implement short-read NGS technologies because of their higher throughput as well as their significantly lower error rate compared with long-read NGS technologies, such as SMRT sequencing (Pacific Biosciences Inc., Menlo: Park, California, USA). However, short-read NGS technologies have a major limitation for the study of HIV quasispecies in that they provide a consensus view of the viral population rather than an accurate reconstruction of the viral haplotypes present in the quasispecies, which can only be facilitated by using alternative approaches, such as unique sequence tags to generate single genomes [50,51]. Here, we implemented a method previously described by our group [52] based on SMRT sequencing, which allows for an accurate reconstruction of HIV sequences at the haplotype level, demonstrating the complex dynamics of the viral quasispecies during treatment failure.

Materials and methods

Sample collection and processing

Plasma samples from 38 de-identified HIV-positive individuals under treatment failure were included in the study. These samples were originally collected for drug resistance testing at the Institute of Biomedical Investigations in Retrovirus and AIDS (INBIRS), a national reference laboratory that provides testing services to several hospitals in Argentina. Viral RNA was extracted using the QIAamp viral extraction Kit (QIAGEN GmbH, Hilden, Germany) from 140 μl of the same plasma sample that previously tested positive for resistance. cDNA synthesis was carried out with Superscript IV (Life Technologies, Carlsbad, California, USA) using 1 μl of reverse primer (2 μmol/l), 1ul of dNTP (40 mmol/l), 1 μl of DTT, 1 μl of RNaseOUT, 1 μl of SSIV, and 11 μl of RNA, incubating 1 h at 50 °C, followed by 10 min at 70 °C. After cDNA synthesis, RNA was digested by incubating with RNaseH for 20 min at 37 °C. A first round of PCR was performed with the Q5 polymerase (NEB) using 5 μl of 5X Q5 Buffer, 5 μl of 5X Q5 Enhancer, 0.875 μl of dNTP (10 mmol/l), 0.625 μl of each primer (20 μmol/l), 0.25 μl of Q5 HS High-Fidelity enzyme, and 5 μl of cDNA, using the following cycling conditions: 3 min at 98 °C followed by 35 cycles of 30 s at 98 °C, 15 s at 55 °C and 5 min at 72 °C. The same conditions were implemented for the second round of PCR, but using 0.5 μl of product from the first round. For the second round of the PCR, barcoded primers were used in order to facilitate multiplexing of amplicons. The primers used for the RT-PCR were VifPolRev1 (5′-ATCATCACCTGCCATCTG-3′) for RT and first round PCR together with GagPolFw1 (5′-ACTTCCCTCAAATCACTC-3′); and VifPolRev2 (5′-CTGCCATCTGTTTTCCAT-3′) and GagPolFw2 (5′-CCTCAAATCACTCTTTGG-3′) for the second round PCR. These primers were designed to match conserved regions across all HIV complete genome reference sequences available at Los Alamos National Laboratory HIV Database (https://www.hiv.lanl.gov/content/index).

Multiplexed single molecule, real-time sequencing

Library preparation was performed following the same protocols as previously described [52]. Briefly, PCR products were purified separately using the Wizard SV Gel and PCR Clean-Up System (Promega, Madison, Wisconsin, USA) and DNA was quantified using the NanoDrop ND-1000 UV-Vis Spectrophotometer (Thermo Fisher Scientific, Waltham, Massachusetts, USA). Equal amounts of DNA from each of the 38 PCR products were pooled together and a SMRTbell library was generated according to protocols from the DNA Template Prep Kit 2.0 (Pacific Biosciences Inc.; cat 100–259-100), which involves a first step of Damage Repair, a second step of End Repair, an O.N. ligation to SMRTbell blunt adapters, and a fourth step of ExoIII and ExoVII digestion. The quality of the library was assessed using the Agilent 2100 Bioanalyzer system (Agilent Technologies, Santa Clara, California, USA). SMRT sequencing was performed on the PacBio RSII, using 2-h movies. Error correction and variant calling was performed by implementing the Multilayer Directed Phasing and Sequencing (MDPSeq) algorithm as previously described [52].

Multiplexed Illumina sequencing

Viral DNA was fragmented and appended with dual-indexed bar codes using the NexteraXT DNA Library Preparation kit (Illumina, San Diego, California, USA). Each sample was prepared in duplicate. Libraries were validated by capillary electrophoresis on an Agilent 4200 TapeStation, pooled at equimolar concentrations, and sequenced on an Illumina HiSeq3000 at 100SR to a depth of 10 000× coverage. Reads were aligned to the HXB2 reference sequence using bwa (v 0.7.17-r1188). Reference-based variant calling was performed with bcftools (v 1.9) and specific codons were examined for variants.

Resistance analysis

An in-house script was developed to search for mutations associated with reduced susceptibility to protease, reverse transcriptase, and integrase inhibitors in the final sequences. Resistance mutations considered in this analysis were those associated with major resistance as reported by the International AIDS Society-USA in 2017 [53]: protease inhibitors -- D30N, V32I, M46I, M46L, I47A, I47 V, G48 V, I50L, I50 V, I54L, I54 M, Q58E, T74P, L76 V, V82A, V82F, V82S, V82T, V82L, N83D, I84 V, N88S, and L90 M. NRTI -- M41L, A62 V, K65R, K65E, K65N, D67N, T69D, 69 insert t, K70R, K70E, L74 V, V75I, F77L, Y115F, F116Y, Q151 M, M184I, M184 V, L210W, T215F, T215Y, K219E, and K219Q. NNRTI -- L100I, K101E, K101P, K103N, K103S, V106A, V106 M, V108I, E138A, E138G, E138K, E138Q, E138R, V179L, Y181C, Y181I, Y181 V, Y188C, Y188H, Y188L, G190A, G190S, H221Y, P225H, F227C, M230L, and M230I. Integrase inhibitors -- T66I, E92Q, F121Y, Y143C, Y143H, Y143R, S147G, Q148H, Q148K, Q148R, and N155H.

Phylogenetic analysis

Viral subtype was determined using the Recombinant Identification Program (RIP, Los Alamos National Laboratory, https://www.hiv.lanl.gov/content/sequence/RIP/RIP.html). Neighbor Joining trees were constructed using Geneious 9.0.4, under the Tamura-Nei model. Support was assessed by running 100 bootstrap replicates. Intra-host HIV phylogenetic reconstruction was performed using a Bayesian approach as implemented in Mr. Bayes v3.2.6, using the GTR substitution model with gamma-distributed rate variation across sites, and a proportion of invariable sites (nst = 6, rate = invgamma), on four simultaneous independent runs with four chains each. The analysis was run until the standard deviation of the split frequencies was below 0.02.

Intra-host diversity estimation

Intra-host diversity was estimated using the Shannon index, modified to account for the level of viral diversification in the quasispecies, using the median distance between sequences within individuals (di):Diversity=(1)1ipiln(pi)1ln(di)

Results

The study of HIV quasispecies is a challenging task from a technical standpoint as the close genetic similarity between variants impairs the ability to resolve the viral haplotypes present in the sample under study. Sanger sequencing can only provide a consensus view of the quasispecies, unless performed in combination with cloning; although when implementing cloning the level of characterization that can be achieved depends on the number of clones that are sequenced. Next-generation sequencing (NGS) can solve this issue by facilitating massive parallel sequencing of every variant in the sample. However, short-read NGS technologies are limited in their capacity to fully phase the haplotypes, whereas long-read NGS suffer of a high background noise, which impairs the ability to distinguish true diversity from sequencing errors. We have previously shown that sequencing error can be corrected in SMRT sequencing by implementing a recursive clustering method (Multilayer Directed Phasing and Sequencing algorithm, MDPSeq), which allows accurate sequencing of individual variants [52]. In the present study, we use this approach to study the resistance profile of 38 HIV-positive individuals under treatment failure. According to the previously performed Sanger-based resistance tests, amplification and sequencing from the 38 plasma samples showed a total of 134 mutations associated with major resistance: specifically, 21 mutations associated with resistance to protease inhibitors, 63 to NRTIs, and 50 to NNRTIs (Supplementary Table 1, https://links.lww.com/QAD/B848). In order to implement our SMRT sequencing-based resistance test, we amplified the entire HIV pol gene using barcoded primers, and then sequenced the PCR products by pooling them into two mixtures of 19 amplicons each. The different HIV pol sequences obtained from each patient after phasing and assembly of the sequencing reads using our MDPSeq algorithm are depicted in the phylogenetic tree presented in Fig. 1. This phylogenetic analysis showed 38 differentiated clusters with no indication of recombination or cross-contamination between samples. Fourteen samples were identified as Subtype B variants, and the remaining 24 were identified as BF recombinants.

F1
Fig. 1:
Schematic view of the HIV pol sequences obtained for the 38 patients under study.

On the basis of sequencing coverage, we determine the relative frequency of each variant and estimated the intra-host HIV diversity index using a modified Shannon index. The classic Shannon index was weighted by the median genetic distance between variants in order to have a measure that accounts not only for the number of different sequences and their relative frequency but also for the level of diversification within the patient. Results from intrahost HIV diversity calculations are shown in Fig. 2. We observed a broad distribution of the modified Shanon index but with most samples exhibiting an intermediate level of diversity. Interestingly, although the number of viral variants found in each patient was variable, seven samples produced only one single viral sequence each (modified Shanon index = 0). In order to determine whether the lack of viral diversity in the seven samples that produced one single sequence was the consequence of primer selection during the RT-PCR, we re-amplified those samples using a different set of four primers and sequenced the new amplicons using Sanger sequencing. Results showed mixed peaks in the chromatogram of four samples, suggesting that primer selection could explain the lack of diversity observed in the SMRT sequencing-based approach. However, two samples showed no evidence of variability in any of the approximately 2300 sequenced nucleotides and another showed only one mixed peak, confirming the results from the SMRT sequencing-based approach and suggesting that the virus in these three patients had not diversified since treatment failure.

F2
Fig. 2:
Intra-host HIV pol diversity.

In order to determine the resistance profile of each sample, we implemented an in-house script to search for resistance mutations across each sequence. Our results showed that the SMRT sequencing-based test was able to identify seven additional major resistance mutations, all of them conferring resistance to NRTIs compared with the Sanger-based test. We found that the highest intra-host frequency of a mutation detected by the SMRT sequencing-based test but not detected by the Sanger-based test was 30%, ranging down to 2.5% (K70E: 2.5%, K65E: 3.2%, Y115F: 6.5%, K219Q: 6.7%, K65R: 27.3%, D67N: 28.6%, and M41L: 30.8%) (Fig. 3). These results suggest that our approach has a sensitivity to detect minor variants comparable with high-throughput short-read NGS workflows.

F3
Fig. 3:
Relative frequency of each drug resistance mutation.

At the same time, three resistance mutations reported in two samples by the Sanger-based resistance test were missing in the data generated with the SMRT-based approach. An inspection of the Sanger chromatograms from those two samples showed that in one case, the M184V mutation was present as a single nonambiguous chromatogram peak. In the other case, both the A62V and the K65R mutations showed up as minor variants in mixed peaks in the chromatogram. When these two samples were re-amplified using the additional set of four primers and Sanger sequenced, the M184V mutation was confirmed but mutations A62V and K65R were not detected.

The comparison of results obtained with SMRT sequencing against those obtained with Sanger sequencing are important to compare our approach to the most frequently used approach for HIV drug resistance testing. However, minor mutations detected by the SMRT-based resistance test but missed by the Sanger-based approach could represent false-positives (specificity) and, similarly, mutations could exist that were not detected by either technology (false negatives, sensitivity). Therefore, in order to rule this out, we performed Illumina sequencing on a subset of 11 samples in which SMRT sequencing found drug resistance mutations at a prevalence at about or lower than 50%, including those that harbored the seven mutations detected by SMRT but not by Sanger. As shown in Fig. 4, there was a strong correlation (P < 0.0001) between the frequencies estimated by SMRT and Illumina sequencing, with only two mutations showing a large discrepancy between them (in red letters in Supplementary Table 1, https://links.lww.com/QAD/B848). All but one of the mutations detected by Illumina were also detected by SMRT and all but three of the mutations detected by SMRT were also detected by Illumina. In three out of the four cases, it corresponded to mutations found at frequencies of 5% or lower but, interestingly, one of them was found in 30% of the reads for the SMRT sequencing. Importantly, five out of the seven mutations identified by the SMRT but not by the Sanger-based resistance test were confirmed by Illumina sequencing, and seven out of the 11 mutations identified by SMRT at a frequency lower than 20% were identified by Illumina at that same frequency range. This result shows an inter-platform consistency between short-read and long-read NGS approaches, and indicates that both NGS methods have a higher sensitivity than Sanger sequencing.

F4
Fig. 4:
Comparison of the relative frequency of mutations as determined by single molecule, real-time and Illumina sequencing.

As the SMRT-based resistance test, in contrast with the Sanger sequencing approach, allows reconstructing the viral haplotypes, we performed a closer examination of the minor resistant variants deciphering the actual phylogenetic relationship between HIV variants in the viral quasispecies. In order to do this, we performed a Bayesian phylogenetic analysis of the sequences obtained from each patient and we mapped the resistance mutations across the HIV variants in the phylogenetic reconstruction of the quasispecies. We applied this approach to the six samples that, according to our SMRT test, harbored at least one resistance mutation present at a frequency between 20 and 80%. Results presented in Fig. 5 show that the resistance profile of viral variants in the quasispecies of multiresistant patients is complex. Two of the six samples (Emory ID 05 and 09) harbored susceptible wild-type variants among the resistant ones. While the susceptible variant in sample Emory ID 05 might be explained as a reversion from a resistant variant, the profile in sample Emory ID 09 is more complexed: the susceptible variants were present both in the most ancestral branches as well as mixed among resistant variants, suggesting a dynamic selection/reversion process. In two of the six patients, evidence suggested that the treatment failure was linked to a specific resistance mutation present in the most ancestral variant to the clade: K65R in patient Emory ID 05, and K103N in patient Emory ID 29. Overall, a trend toward higher numbers of resistance mutations in the more derived sequences was observed in samples Emory ID 05, Emory ID 25, and Emory ID 29, suggesting that additional resistance mutations might be selected after treatment failure.

F5
Fig. 5:
Bayesian analysis of the viral quasispecies.

Discussion

In the present study, we validated a SMRT sequencing technology approach for the detection of HIV drug resistance. In addition, we show that the greater sequencing resolution achieved through haplotype phasing can help gain insights into the dynamics of the viral quasispecies during treatment failure.

According to our results, the SMRT-based resistance test was able to identify 131 of the 134 resistance mutations previously identified with the Sanger-based resistance test, although two of the three resistance mutations not detected by the SMRT sequencing approach could not be confirmed when the Sanger sequencing was repeated. In fact, the SMRT-based test found seven additional resistance mutations that were not detected by the Sanger-based test, all of them at a frequency lower than 30%. Importantly, five out of these seven mutations were confirmed implementing Illumina sequencing, which suggest that these mutations were in fact present at low frequency in these samples and that both methods are able to detect low frequency variants at higher sensitivity than Sanger sequencing. In fact, the strong correlation between the relative frequencies estimated from the SMRT and Illumina sequencing indicates that both long-read and short-read NGS-based tests are sensitive methods to accurately estimate the frequency of drug resistance mutations.

It is interesting also to note that among the resistance mutations identified by the Sanger-based resistance test, seven of them were present at a frequency lower than 10% according to the SMRT-based test, a frequency lower than the expected sensitivity threshold for Sanger sequencing. This can be explained by the fact that both tests were performed on the same plasma samples but they were not performed on the same PCR product. Therefore, it is not possible to ascertain that the mutations found by both tests were present in the amplicon sequenced by Sanger at the same frequency as they were in the amplicon sequenced using the SMRT technology. In other words, these apparent inconsistencies in sensitivity threshold between methods could be explained by PCR amplification bias. Overall, our results are consistent with a detection limit for the Sanger-based test of 20–30% and they also indicate that the SMRT-based test would be able to detect resistance mutations at a frequency as low as 2% (M41L in Emory 05), when 19 samples are multiplexed in a single SMRT cell, and sequenced in an RSII instrument. Higher levels of multiplexing per SMRT cell could potentially be achieved if the higher throughput Sequel instrument is used for sequencing.

Our results also show how multiplexing of samples is facilitated by our approach. This is a challenging task as every sample has an unknown number of genetic variants, which can range from one to hundreds. However, our algorithms are able to search the raw data and determine, based on the reading of the sequencing noise, how many variants are present and can be reconstructed with high confidence. It is important to highlight that all the HIV variants sequenced in this study clustered in monophyletic clades that matched the barcode assigned to each sample. This indicates that our algorithms are highly accurate on phasing sequencing reads as not one single sequence was assigned to a wrong sample.

Our study is limited by the lack of information of the particular drug regimen and treatment history of the study subjects. As samples were obtained from a reference laboratory where they were originally collected for drug resistance testing as patients were derived there from several different hospitals, clinical information linked to these patients was unavailable. The inability to access viral load measurements on these samples prevented us also from establishing template sampling depth and template recovery. However, as samples under study were collected to perform actual drug resistance testing ordered by a physician following up the response to treatment in these patients, the viral load measurements should have been at least one log higher than the last recorded measurement, as such increment is a condition to request a drug resistance test.

In spite of the lack of clinical data, the SMRT-based approach allowed us to explore the composition of the viral quasispecies at the moment of treatment failure, and to evidence a complex pattern of resistance among viral variants in the quasispecies. Although six of the 38 samples did not exhibit evidence of drug resistance mutations in either test, 11 of the 32 samples with evidence of resistance mutations had at least one resistance mutation at low frequency suggesting that some resistance could have been selected after treatment failure, during suboptimal treatment. In fact, the results from the Bayesian reconstruction of the phylogenetic relationships between variants in the viral quasispecies suggest a trend toward accumulation of resistance mutations, with variants in more derived branches of the tree harboring more mutations than those located in more ancestral branches. However, we also found evidence of the same resistance mutation being independently selected in different branches of the tree, variable numbers of resistance mutations among viral variants in the quasispecies, or mixtures of fully susceptible variants and resistant variants with a single mutation. Although the reasons behind these observations could remain highly speculative, these findings indicate that viral dynamics and resistance profiles during treatment failure are more complex than the simple selection of a resistant variant that rise to dominate the viral quasispecies, and remain selected during suboptimal treatment.

It is important not to disregard the fact that we were able to confirm that in two of the 38 samples under study, there was only 1 single HIV variant detected. In those two cases, both the SMRT test performed on population whole-pol RT-PCR amplicons, and a Sanger test performed on subsegments of pol using entirely different sets of primers, showed no evidence of more than one HIV pol sequence. These cases could represent an early treatment failure, in which the resistant variant was recently selected and is dominating the quasispecies. However, considering the well-known capacity of HIV to mutate and rapidly diversify, these findings are intriguing.

In summary, in the present study, we validated SMRT technology as an accurate method to assess the presence of HIV drug resistance mutations to a sensitivity equivalent to that of Illumina and higher to that of Sanger sequencing. We also demonstrated the utility of this approach as a tool to gain useful insights into viral quasispecies genetic composition. In particular, we uncovered a complex pattern of resistance profile in the quasispecies during treatment failure. Considering the clinical relevance of understanding emergence of drug resistance during antiretroviral treatment, and the capability of the SMRT test to account for otherwise undetected minor resistant variants, we consider that extending this study to larger cohorts of patients under treatment failure can provide relevant insights to inform decisions of treatment regimen changes in this type of patients.

Acknowledgements

We are grateful for the support from amfAR (Mathilde Krim Fellowship in Basic Biomedical Research; 109716-63-RKVA to D.C.M., and 108672-5-RKGN to D.D.) and Georgia Research Alliance (GRA.VI.1.6.C8 and GRA.VL17.C9 to D.D. and E.H.).We thank Kathryn Pellegrini and Gregory Tharp (Yerkes Genomics Core) for technical assistance on Illumina sequencing data processing.

Contributions: D.A.D. and D.C.M. conceived and designed the experiments, performed the experiments, analyzed the data, and wrote the manuscript. L.Z. performed the Sanger tests, collected samples and data necessary for validation. E.H. contributed reagents/materials/analysis tools, and contributed to data interpretation. H.S. designed the sample panel to perform the study, and facilitated sample collection and sharing.

Sources of funding: The Foundation for AIDS Research (amfAR - Mathilde Krim Fellowship in Basic Biomedical Research; 109716-63-RKVA to D.C.M., and 108672-5-RKGN to D.D.), and Georgia Research Alliance (GRA.VI.1.6.C8 and GRA.VL17.C9 to D.D. and E.H.).

Conflicts of interest

There are no conflicts of interest.

References

1. Hamers RL, Sawyer AW, Tuohy M, Stevens WS, Rinke de Wit TF, Hill AM. ART-A Consortium. Cost-effectiveness of laboratory monitoring for management of HIV treatment in sub-Saharan Africa: a model-based analysis. AIDS 2012; 26 (13):1663–1672.
2. Lorenzana SB, Hughes MD, Grinsztejn B, Collier AC, Luz PM, Freedberg KA, et al. Genotype assays and third-line ART in resource-limited settings: a simulation and cost-effectiveness analysis of a planned clinical trial. AIDS, 2012; 26:1083–1093.
3. Dilernia DA, Monaco DC, Cesar C, Krolewiecki AJ, Friedman SR, Cahn P, Salomon H. Estimation of HIV-testing rates to maximize early diagnosis-derived benefits at the individual and population level. PLoS One 2013; 8:e53193.
4. Lohse N, Hansen AB, Pedersen G, Kronborg G, Gerstoft J, Sørensen HT, et al. Survival of persons with and without HIV infection in Denmark, 1995–2005. Ann Intern Med 2007; 146:87–95.
5. Mocroft A, Ledergerber B, Katlama C, Kirk O, Reiss P, d’Arminio Monforte A, et al. EuroSIDA study group. Decline in the AIDS and death rates in the EuroSIDA study: an observational study. Lancet 2003; 362:22–29.
6. Mocroft A, Phillips AN, Gatel l J, Ledergerber B, Fisher M, Clumeck N, et al. EuroSIDA study group. Normalisation of CD4 counts in patients with HIV-1 infection and maximum virological suppression who are taking combination antiretroviral therapy: an observational cohort study. Lancet 2007; 370:407–413.
7. Baeten JM, Kahle E, Lingappa JR, Coombs RW, Delany-Moretlwe S, Nakku-Joloba E, et al. Partners in Prevention HSV/HIV Transmission Study Team. Genital HIV-1 RNA predicts risk of heterosexual HIV-1 transmission. Sci Transl Med 2011; 3:77ra29.
8. Bavinton BR, Pinto AN, Phanuphak N, Grinsztejn B, Prestage GP, Zablotska-Manos IB, Jin F, et al. Opposites Attract Study Group. Viral suppression and HIV transmission in serodiscordant male couples: an international, prospective, observational, cohort study. Lancet HIV 2018; 5:e438–e447.
9. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy N, et al. HPTN 052 Study Team. Antiretroviral therapy for the prevention of HIV-1 transmission. N Engl J Med 2016; 375:830–839.
10. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy N, et al. HPTN 052 Study Team. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med 2011; 365:493–505.
11. Eshleman SH, Hudelson SE, Redd AD, Swanstrom R, Ou SS, Zhang XC, et al. Treatment as prevention: characterization of partner infections in the HIV Prevention Trials Network 052 Trial. J Acquir Immune Defic Syndr 2017; 74:112–116.
12. Panel's Recommendations for Acute and Recent (Early) HIV Infection. Considerations for Antiretroviral Use in Special Patient Populations - Acute and Recent (Early) HIV Infection. 2017. Available at: https://aidsinfo.nih.gov/guidelines/html/1/adult-and-adolescent-arv/20/acute-and-recent--early--hiv-infection.
13. Gupta RK, Gregson J, Parkin N, Haile-Selassie H, Tanuri A, Andrade Forero L, et al. HIV-1 drug resistance before initiation or re-initiation of first-line antiretroviral therapy in low-income and middle-income countries: a systematic review and meta-regression analysis. Lancet Infect Dis 2018; 18:346–355.
14. Günthard HF, de Wolf F, Dunn D, Cozzi-Lepri A, de Luca A, Wittkop L, et al. Effect of transmitted drug resistance on virological and immunological response to initial combination antiretroviral therapy for HIV (EuroCoord-CHAIN joint project): a European multicohort study. Lancet Infect Dis 2011; 11:363–371.
15. Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection. Recommendations for a public health approach. Second Edition. Geneva: World Health Organization. Available at: http://www.who.int/hiv/pub/arv/arv-2016/en/. 2016. [Accessed 20 February 2020].
16. World Health Organization. HIV Drug Resistance Report 2017. Geneva: World Health Organization; 2017.
17. Rodriguez-Rodrigues N, Duran A, Bouzas MB, Zapiola I, Vila M, Indyk D, et al. Increasing trends in primary NNRTI resistance among newly HIV-1-diagnosed individuals in Buenos Aires, Argentina. J Int AIDS Soc 2013; 16:18519.
18. Phillips AN, Cambiano V, Nakagawa F, Revill P, Jordan MR, Hallett TB, et al. Working Group on Modelling Potential Responses to High Levels of Pre-ART Drug Resistance in Sub-Saharan Africa. Cost-effectiveness of public-health policy options in the presence of pretreatment NNRTI drug resistance in sub-Saharan Africa: a modelling study. Lancet HIV 2018; 5:e146–e154.
19. Dilernia DA, Lourtau L, Gomez AM, Ebenrstejin J, Toibaro JJ, Bautista CT, et al. Drug-resistance surveillance among newly HIV-1 diagnosed individuals in Buenos Aires, Argentina. AIDS 2007; 21:1355–1360.
20. Avila-Rios S, García-Morales C, Garrido-Rodríguez D, Ormsby CE, Hernández-Juan R, Andrade-Villanueva J, et al. Mexican HIV Molecular Epidemiology Project Group. National prevalence and trends of HIV transmitted drug resistance in Mexico. PLoS One 2011; 6:e27812.
21. Charpentier C, Bellecave P, Cisse M, Mamadou S, Diakite M, Peytavin G, et al. High prevalence of antiretroviral drug resistance among HIV-1-untreated patients in Guinea-Conakry and in Niger. Antivir Ther 2011; 16:429–433.
22. de Medeiros RM, Junqueira DM, Matte MC, Barcellos NT, Chies JA, Matos Almeida SE, et al. Co-circulation HIV-1 subtypes B, C, and CRF31_BC in a drug-naive population from Southernmost Brazil: analysis of primary resistance mutations. J Med Virol 2011; 83:1682–1688.
23. Hofstra LM, Sauvageot N, Albert J, Alexiev I, Garcia F, Struck D, et al. Transmission of HIV drug resistance and the predicted effect on current first-line regimens in Europe. Clin Infect Dis 2016; 62:655–663.
24. Muwonga J, Edidi S, Butel C, Vidal N, Monleau M, Okenge A, et al. Resistance to antiretroviral drugs in treated and drug-naive patients in the Democratic Republic of Congo. J Acquir Immune Defic Syndr 2011; 57: (Suppl 1): S27–33.
25. Myers JE, Taylor BS, Rojas Fermín RA, Reyes EV, Vaughan C, José L, et al. Transmitted drug resistance among antiretroviral-naive patients with established HIV type 1 infection in Santo Domingo, Dominican Republic and review of the Latin American and Caribbean literature. AIDS Res Hum Retroviruses 2012; 28:667–674.
26. Ndembi N, Hamers RL, Sigaloff KC, Lyagoba F, Magambo B, Nanteza B, et al. Transmitted antiretroviral drug resistance among newly HIV-1 diagnosed young individuals in Kampala. AIDS 2011; 25:905–910.
27. Sigaloff KC, Mandaliya K, Hamers RL, Otieno F, Jao IM, Lyagoba F, et al. Short communication: high prevalence of transmitted antiretroviral drug resistance among newly HIV type 1 diagnosed adults in Mombasa, Kenya. AIDS Res Hum Retroviruses 2012; 28:1033–1037.
28. Soria J, Bull M, Mitchell C, La Rosa A, Dross S, Kraft K, et al. Transmitted HIV resistance to first-line antiretroviral therapy in Lima, Peru. AIDS Res Hum Retroviruses 2012; 28:333–338.
29. Stekler JD, Milne R, Payant R, Beck I, Herbeck J, Maust B, et al. Transmission of HIV-1 drug resistance mutations within partner-pairs: a cross-sectional study of a primary HIV infection cohort. PLoS Med 2018; 15:e1002537.
30. Boyd M, Emery S, Cooper DA. Antiretroviral roll-out: the problem of second-line therapy. Lancet 2009; 374:185–186.
31. Boyd MA, Cooper DA. Second-line combination antiretroviral therapy in resource-limited settings: facing the challenges through clinical research. AIDS 2007; 21: (Suppl 4): S55–63.
32. Chaplin B, Imade G, Onwuamah C, Odaibo G, Audu R, Okpokwu J, et al. Distinct pattern of thymidine analogue mutations with K65R in patients failing tenofovir-based antiretroviral therapy. AIDS Res Hum Retroviruses 2018; 34:228–233.
33. De Luca A, Sidumo ZJ, Zanelli G, Magid NA, Luhanga R, Brambilla D, et al. Accumulation of HIV-1 drug resistance in patients on a standard thymidine analogue-based first line antiretroviral therapy after virological failure: implications for the activity of next-line regimens from a longitudinal study in Mozambique. BMC Infect Dis 2017; 17:605.
34. Fily F, Ayikobua E, Ssemwanga D, Nicholas S, Kaleebu P, Delaugerre C, et al. HIV-1 drug resistance testing at second-line regimen failure in Arua, Uganda: avoiding unnecessary switch to an empiric third-line. Trop Med Int Health 2018; 23:1075–1083.
35. Kanters S, Socias ME, Paton NI, Vitoria M, Doherty M, Ayers D, et al. Comparative efficacy and safety of second-line antiretroviral therapy for treatment of HIV/AIDS: a systematic review and network meta-analysis. Lancet HIV 2017; 4:e433–e441.
36. Kantor R, DeLong A, Schreier L, Reitsma M, Kemboi E, Orido M, et al. HIV second-line failure and drug resistance at high- and low-level viremia in Western Kenya. AIDS 2018; 32:2485–2496.
37. Lam EP, Moore CL, Gotuzzo E, Nwizu C, Kamarulzaman A, Chetchotisakd P, et al. Antiretroviral resistance after first-line antiretroviral therapy failure in diverse HIV-1 subtypes in the SECOND-LINE Study. AIDS Res Hum Retroviruses 2016; 32:841–850.
38. Matsuda EM, Coelho LPO, Romero GF, de Moraes MJ, Lopez-Lopes GIS, Morejon K, et al. High prevalence of drug resistance mutations among patients failing first-line antiretroviral therapy and predictors of virological response 24 weeks after switch to second-line therapy in Sao Paulo State, Brazil. AIDS Res Hum Retroviruses 2017; 34:156–164.
39. Meloni ST, Onwuamah CK, Agbaji O, Chaplin B, Olaleye DO, Audu R, et al. Implication of first-line antiretroviral therapy choice on second-line options. Open Forum Infect Dis 2017; 4:ofx233.
40. Pujades-Rodriguez M, O’Brien D, Humblet P, Calmy A, et al. Second-line antiretroviral therapy in resource-limited settings: the experience of Medecins Sans Frontieres. AIDS 2008; 22:1305–1312.
41. Sawe FK, McIntyre JA. Monitoring HIV antiretroviral therapy in resource-limited settings: time to avoid costly outcomes. Clin Infect Dis 2009; 49:463–465.
42. Badri M, Lawn SD, Wood R. Utility of CD4 cell counts for early prediction of virological failure during antiretroviral therapy in a resource-limited setting. BMC Infect Dis 2008; 8:89.
43. Marconi VC, Sunpath H, Lu Z, Gordon M, Koranteng-Apeagyei K, Hampton J, et al. Prevalence of HIV-1 drug resistance after failure of a first highly active antiretroviral therapy regimen in KwaZulu Natal, South Africa. Clin Infect Dis 2008; 46:1589–1597.
44. Orrell C, Walensky RP, Losina E, Pitt J, Freedberg KA, Wood R. HIV type-1 clade C resistance genotypes in treatment-naive patients and after first virological failure in a large community antiretroviral therapy programme. Antivir Ther 2009; 14:523–531.
45. Petersen ML, van der Laan MJ, Napravnik S, Eron JJ, Moore RD, Deeks SG. Long-term consequences of the delay between virologic failure of highly active antiretroviral therapy and regimen modification. AIDS 2008; 22:2097–2106.
46. Wallis CL, Mellors JW, Venter WD, Sanne I, Stevens W. Varied patterns of HIV-1 drug resistance on failing first-line antiretroviral therapy in South Africa. J Acquir Immune Defic Syndr 2010; 53:480–484.
47. Levison JH, Orrell C, Losina E, Lu Z, Freedberg KA, Wood R. Early outcomes and the virological effect of delayed treatment switching to second-line therapy in an antiretroviral roll-out programme in South Africa. Antivir Ther 2011; 16:853–861.
48. Brumme CJ, Poon AFY. Promises and pitfalls of Illumina sequencing for HIV resistance genotyping. Virus Res 2017; 239:97–105.
49. Inzaule SC, Hamers RL, Paredes R, Yang C, Schuurman R, Rinke de Wit TF. The evolving landscape of HIV drug resistance diagnostics for expanding testing in resource-limited settings. AIDS Rev 2017; 19:219–230.
50. Keys JR, Zhou S, Anderson JA, Eron JJ, Rackoff LA, Jabara C, Swanstrom R. Primer ID informs next-generation sequencing platforms and reveals preexisting drug resistance mutations in the HIV-1 reverse transcriptase coding domain. AIDS Res Hum Retroviruses 2015; 31:658–668.
51. Boltz VF, Rausch J, Shao W, Hattori J, Luke B, Maldarelli F, et al. Ultrasensitive single-genome sequencing: accurate, targeted, next generation sequencing of HIV-1 RNA. Retrovirology 2016; 13:87.
52. Dilernia DA, Chien JT, Monaco DC, Brown MP, Ende Z, Deymier MJ, et al. Multiplexed highly-accurate DNA sequencing of closely-related HIV-1 variants using continuous long reads from single molecule, real-time sequencing. Nucleic Acids Res 2015; 43:e129.
53. Wensing AM, Calvez V, Günthard HF, Johnson VA, Paredes R, Pillay D, et al. 2017 update of the drug resistance mutations in HIV-1. Top Antivir Med 2017; 24:132–133.
Keywords:

HIV drug resistance; long-read next-generation sequencing; phasing; single molecule, real-time sequencing

Supplemental Digital Content

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.