Introduction
Antiretroviral therapy (ART) has been rapidly scaled up over the last decade, and today almost 40% of HIV-infected individuals in sub-Saharan Africa are estimated to be on ART. This rapid ART rollout was possible because resource-limited countries (RLCs) adopted the 2010 WHO guidelines, which recommended standardized regimes and laboratory monitoring when available [1]. Nonetheless, new WHO guidelines recommend virological load to monitor treatment outcome, which has proven useful to deliver high-quality treatment in high-income countries [2–5].
Scaling up the quality of ART in sub-Saharan Africa is challenging because of the high costs of tests, lack of laboratory infrastructure, logistic constraints and need for trained staff [6,7–9]. However, there is growing evidence that a long-term failing treatment contributes to the development of resistance and increases the risk of transmitted drug resistance (TDR) [10–16]. Therefore, continuous expansion of ART increases the need to rapidly detect cases in which first-line treatment has failed, despite ongoing debate concerning the utility of routine virological monitoring [17,18].
Another peculiarity in RLCs is the high diversity of HIV-1 genetic variants, in particular of circulating recombinant forms (CRFs). It is now well established for subtype C that certain mutations differ from what is observed in individuals infected with subtype B under similar drug pressure [19–21], but our knowledge on the selection of variant-specific drug resistance mutations (DRMs) remains incomplete.
A better understanding of the emergence of DRM is important to evaluate the long-term effectiveness of the standardized regimens. Here, we thoroughly studied the impact of ART rollout without virological monitoring on the frequency of both known DRMs and additional selected mutations, using comprehensive genetic data from West and Central African countries. We also examined whether differences exist between the genetic variants that predominate in these regions.
Methods
Study design and HIV-infected individuals
Partial p51 reverse transcriptase sequences from reverse transcriptase inhibitor (RTI)-naive individuals and RTI-experienced individuals on first-line treatment failure were retrieved from cross-sectional studies on transmitted and acquired drug resistance in the database of UMI233 [10,22–45]. These studies were done in West and Central Africa. Experienced individuals received the same first-line regimen: zidovudine (ZDV)/stavudine (d4T) + lamivudine (3TC) + neviparine (NVP)/efavirenz (EFV). Virological failure was defined as HIV-1 RNA plasma viral load measurement more than 1000 copies/ml, as recommended by WHO for virological failure in RLCs [46].
HIV-1 subtype/circulating recombinant form designation in reverse transcriptase
A reference dataset containing each HIV-1 group-M genetic variant was retrieved from the Los Alamos curated database to confirm the subtype and CRF designations of our sequences. Alignments and maximum likelihood phylogenetic analyses were achieved using Muscle v3.8.31 and PhyML v3.0, respectively [47–49]. Clustering with particular reference sequences (approximate likelihood ratio ≥0.90) and recombination analyses with Simplot v3.5.1 were used for subtype/CRF assignment [50]. Sequences with unique mosaic structures composed of regions from multiple subtypes/CRFs were designated as unique recombinant forms (URFs).
Analysis of antiretroviral drug resistance mutations
Sequences were analyzed for DRMs using the HIVdb algorithm v7.0.1 (Drug Resistance Database from Stanford University,www.hivdb.stanford.edu). The resistance profile for each drug was categorized as high-level resistance, intermediate resistance and susceptible. The mutations from the International Antiviral Society-United States of America (IAS-USA) list of 2015 were used as a reference of DRM to RTI [51].
Statistical analysis
We selected the subtypes/CRFs from the RTI-experienced group with more than 60 occurrences to assess if DRMs were more frequently observed in some of them. We compared the prevalence of amino acid substitutions in RTI-experienced and RTI-naive individuals to identify additional nonpolymorphic RTI-selected mutations. Mutations that had a statistically significant association with treatment were defined as being nonpolymorphic when they occurred more frequently in RTI-experienced individuals (≥1%) than in RTI-naive ones (≤0.05%); mutations were independently selected (due to convergent evolution under drug pressure) when they were applicable to three or more countries and three or more major subtypes/CRFs (A, C, D, F, G, CRF01_AE and CRF02_AG); this criterion is similar to that outlined by Shahriar et al.[52]. We used the RTI-experienced dataset to evaluate if time on treatment affected the occurrences of DRMs. Here, the ART duration was discretized into three categories: 24 or less, 24–48 and more than 48 months. In all analyses, we distinguished between different substitutions for the same amino acid position.
Moreover, Fisher exact tests were conducted to evaluate the significance of the associations. The null hypothesis of no association was rejected if the P value of the test was less than a significance level of 0.05 after adjusting for multiple comparisons. We used the Benjamini–Hochberg correction that controls the expected proportion of false-positives among all rejected hypotheses and performs well in scenarios of independence or positive dependency [53]. We also used Bonferroni and Holm procedure corrections that sought to control the familywise error rate; these procedures reduce the probability of spurious findings, but are overly conservative by not accounting for the correlation structure of the test statistics [54].
Results
Dataset characteristics
A total of 3736 sequences were included in the analysis; 1599 sequences were obtained from RTI-experienced individuals failing first-line ART (ZDV/d4T + 3TC + NVP/EFV) and 2137 were obtained from RTI-naive individuals. The datasets of RTI-experienced individuals were reduced to 1303 by removing sequences that were susceptible to the complete first-line regimen: This was done to diminish the effects of nonadherent individuals. Similarly, the number of sequences from RTI-naive individuals was reduced to 2039 sequences to diminish the effect of TDR, and this was accomplished by removing strains with resistance to any of the drugs.
Sixty-five percent (n = 2176) of the sequences were from Central African countries (Burundi, Cameroon, Chad, Democratic Republic of Congo and Equatorial Guinea), and 35% (n = 1166) were from West African countries (Benin, Burkina Faso, Cote d’Ivoire, Senegal and Togo) (Table 1). HIV-1 strains were highly diverse and included eight subtypes, 18 CRFs and numerous URFs. The dominant subtypes/CRF in the RTI-experienced group were CRF02_AG (45.7%, n = 595), A (6.7%, n = 83), CRF06_cpx (6.3%, n = 82), C (5.0%, n = 65) and G (4.9%, n = 64), whereas in the RTI-naive group they were C (27.0%, n = 550), CRF02_AG (26.5%, n = 541), A (9.0%, n = 184), CRF06_cpx (5.7%, n = 116) and D (3.4%, n = 70). The disparate proportion of treated versus naive populations for each subtype and CRFs do not reflect a particular epidemiological setting, but rather the random outcome of aggregating data from independent studies. URFs accounted for 17.8% (n = 232) and 16.6% (n = 338) of sequences from RTI-experienced and RTI-naive individuals, respectively.
;)
Table 1: Dataset characteristics.
Resistance in the reverse transcriptase inhibitor–experienced group
Among the 1367 RTI-experienced individuals included, 88.9% were resistant to 3TC (n = 1158), 26.9% were resistant to ZDV (n = 351), 27.2% were resistant to d4T (n = 354), 97.5% were resistant to NVP (n = 1270) and 75.2% were resistant to EFV (n = 980). ZDV and d4T levels of resistance were similar because nearly the same number of mutations confers resistance to both drugs. Likewise, EFV and NVP select for the same DRMs, but differential penalty scores for the mutations to these drugs result in different drug resistance interpretations (e.g. mutation Y181C/I/V scores 60 for NVP but 30 for EFV, whereas both K103S and G190A score 60 for NVP and 45 for EFV): 22.6% (n = 294) of the sequences were associated with an intermediate level of resistance to EFV, whereas the same sequences showed high-level resistance to NVP. Resistance to other RTI drugs currently approved by the Food and Drug Administration, USA, was also observed: 20.9% (n = 273) to abacavir, 19.5% (n = 254) to didanosine, 8.4% (n = 110) to tenofovir (TDF), 6.1% (n = 80) to etravirine and 16.3% (n = 213) to rilpivirine.
All RTI mutations from the IAS-USA list were present in the dataset, except 227C (Fig. 1 and Table S1, https://links.lww.com/QAD/A972). The most frequent nucleoside reverse transcriptase inhibitor (NRTI) mutation was M184V/I (88.9%), with M184V (86.9%) being the predominant mutation. Thymidine analog mutations (TAMs) were most frequent at position T215Y/F (33.6%). The Q151M mutation was only detected in 27 individuals (2.1%). Insertions at codon 69 were observed in four sequences (0.3%). The K103N mutation was the most common non-nucleoside reverse transcriptase inhibitors (NNRTI) (49.7%) mutation, followed by Y181C (31.9%).
Fig. 1: Frequency of mutations.
Emergence of additional variation
We compared the frequencies of each amino acid at each sequence position between RTI-experienced individuals and RTI-naive individuals to identify which mutations were associated with treatment. We validated our approach after confirming that only positions and/or substitutions from the IAS-USA list with very low frequency in the ART-experienced group (≤0.8%) and/or multiple occurrences in the ART-naive group did not have a statistically significant association with ART in our dataset. These DRMs were K65E/N, K70E, F77 K101P, E138G/K/R, V179, Y181I, Y188/C/H, F227C and M230I; V179D, V179T, V106I and E138A were present in the ART-naive at frequencies 0.2, 0.4, 1.0 and 2.6%, respectively. Moreover, some of these substitutions (K70E, E138A/G/K/Q/R, V179D/F/T/L, F227C and M230I) were not associated with resistance to the drugs taken by the RTI-experienced individuals under study [51].
In addition to the known DRMs, a total of 62 mutations were significantly more frequently observed in RTI-experienced individuals (Fisher exact test), but only 14 of them qualified as independently selected mutations as outlined in methods (Table 2). Among these 14 mutations, nine are not included in the IAS-USA list but are included in at least one of the three expert lists of DRMs (HIVdb: Stanford University, Stanford, California, USA; ANRS: French National Agency for AIDS Research, and REGA: Rega Institute for Medical Research and University Hospitals Leuven, Leuven, Belgium): D67G (2.5%), T69D (1.9%), T69N (4.1%), L74I (1.5%), V75M (1.8%), I132L (2.2%), D218E (1.7%), F227L (3.6%) and K238T (5.5%) (Fig. 1) [55–58]. The majority of these mutations are scored in at least one of the lists with the exception of I132L and D218E, which are only listed as poorly characterized RTI-DRMs in HIVdb. Mutations T69D/N, L74I and V75M are present in the three lists.
Table 2: Mutations selected by first-line nonnucleoside reverse transcriptase inhibitor–based antiretroviral therapy.
We thus identified five additional selected mutations associated with RTI-experienced individuals, which are not taken into account by any of the expert lists of DRMs: I94L (1.1%), L109I (2.4%), V111L (1.1%), T139R (2.4%) and T165L (2.6%). I94L and T165L were frequently observed along TAMs, whereas L109I and T138R were associated with both TAMs and NNRTI-DRM. The frequency of V111L was not associated with any known mutation (Table S2, https://links.lww.com/QAD/A972). The ‘RTI-experienced versus RTI-naive’ archive from HIVdb substitutions informs that these mutations are very rare in RTI-naive individuals (≤0.06%, except V111I 0.4%, but mostly on subtype B and CRF01_AE) but have been documented in RTI-experienced individuals from different subtypes (subtypes A, B, C, D, F, G, CRF01_AE and CRF02_AG), although in lower frequencies: 0.3% I94L, 0.6% L109I, 1.0% V111L, 0.6% T139R and 0.7% T165L (HIVdb was last accessed in February 2016).
Resistance and time on antiretroviral therapy
The median time on ART was 28 months (interquartile range 22–55). Mutations M184V, all TAMS, certain NNRTIs (A98G, K101H, V108I, P225H and the additional selected mutation K238T from the Stanford and REGA lists) and two newly identified selected mutations (I94L and L109I) were significantly more frequently observed in the groups treated for longer periods of time (P < 0.05) (Table 3). K65R was the only mutation significantly more frequently observed in the category of patients 24 months or less on ART.
Table 3: Reverse transcriptase inhibitor–selected mutations associated with time on treatmenta.
Mutation M184V, which alone confers high level of resistance to 3TC, was the NRTI-resistant mutation observed in highest proportion (82.4, 90.4 and 96.7% for ≤24, >24–48 and >48 months, respectively) and is among the first NRTI mutations to emerge; consequently, sequences had on average at least one mutation of this class (Fig. 2). TAMs develop more slowly than other NRTI. However, the proportion of sequences with three or more TAM resistance mutations increased when patients were treated for longer periods of time (Fig. 2). The frequency of any TAM profile did not have a statistically significant association with time on treatment.
Fig. 2: Distribution of the number of known reverse transcriptase inhibitor resistance mutations by sequence and time category.
NNRTI mutations are drugs with low genetic barriers to resistance. Consistently, sequences from RTI-experienced individuals had on average two mutations conferring resistance to NNRTIs, but results differed again depending on time under treatment: the longer the treatment, the higher the proportion of sequences with more mutations (Fig. 2). Common NNRTI DRMs (e.g. K103N and Y181C) did not have a statistically significant association with time on treatment. They were equally frequent regardless of time category and further supported the rapid selection of these mutations.
Resistance and HIV-1 subtype/circulating recombinant form
We evaluated whether DRMs from the IAS-USA list and the 14 additional selected mutations were more or less frequently observed in certain HIV-1 variants. We only considered subtypes/CRFs for which more than 60 sequences were available in the RTI-experienced group: subtypes A, C, G, CRF02_AG and CRF06_cpx. The frequency of one NRTI and two NNRTI mutations differed according to subtype/CRF (Fig. 3 and Table S3, https://links.lww.com/QAD/A972). Our study confirmed higher frequency of NNRTI mutation V106M in subtype C (12.3%). We observed that (NRTI) G190A was less frequent in CRF02_AG (14%) and M41L was more frequent in CRF06_cpx (37%). We also observed higher frequencies of K65R in subtype C, but the latter was not statistically significant in our study (P = 0.076 after Benjamini–Hochberg correction).
Fig. 3: Reverse transcriptase inhibitor resistance mutations associated with subtype or circulating recombinant forms.An asterisk indicates a frequency value that differed significantly from the corresponding values for other subtypes/circulating recombinant forms.
Discussion
We retrieved and analyzed more than 3500 partial reverse transcriptase sequences from RTI-naive and RTI-experienced individuals from 10 countries in West and Central Africa. All RTI-experienced individuals were on a standardized first-line RTI-based regimen without virological monitoring. ART failure was primarily detected during cross-sectional studies, which aimed to document ART efficacy at the country or program level, and the majority of these individuals did not have clinical signs of failure [28]. Drug-resistant profiles evidenced that the drugs 3TC and the NNRTIs from the regimen were no longer effective for a considerable proportion of patients. Moreover, multiresistance was revealed, involving drugs not included in the administered first-line regimens.
Using a representative dataset of individuals failing a first-line treatment, we confirmed the accumulation and increased frequency of DRMs over time, stressing the impact of expanding ART without scaling virological monitoring. Moreover, following a thorough statistical analysis, we identified five selected mutations that are not included in any expert list of DRM and nine mutations from the HIVdb, ANRS and/or REGA lists. These mutations had a statistically significant association with treatment, they were common in the RTI-experienced group (≥1%), rare in the RTI-naive group (≤0.05%) and independently selected (present in at least three countries and at least three main subtypes/CRFs). Finally, we found one major NRTI and two major NNRTI mutations that differed significantly in frequency across subtypes/CRFs.
We argue that our approach for identifying treatment-selected mutations was satisfactory given that it identified most positions and substitutions from the IAS-USA list and additional mutations from other expert lists of DRM. The accurate identification of additional treatment-related mutations in databases is challenging, because misclassification can result from TDR, population stratification, sample size and the many statistical comparisons required [59]. In this respect, we excluded isolates from RTI-naive persons with evidence of drug resistance prior to the analyses and verified the phylogenetically independent selection of additional mutations by confirming that they were present in different countries and subtypes/CRFs. We used sound significance thresholds for individual comparisons following controls of the false discovery rate and the familywise error rate, so as to compensate for the number of inferences being made, and the corrected P values of the additional selected mutations suggest that there is a low chance that our results are hampered by false positives.
Following our statistical approach, we did not find novel substitutions associated with resistance at known positions from any expert lists of DRM, but we observed the emergence of five additional independently selected mutations at different positions of the reverse transcriptase that are not documented in any of these lists: I94L, L109I, V111I, T139R and T165L. In the HIVdb database, these mutations are very rare in RTI-naive individuals but have been documented at low levels in RTI-experienced individuals from different subtypes and treatments. T165L has been pointed out as an NRTI-DRM, I94L and T139R as NNRTI-DRMs, and L109I has been pointed out as an undifferentiated RTI-DRM [52,60]. T139R and T165L have also been associated with reduced susceptibility to NNRTIs in vitro[61]. These observations come from worldwide data, which clarifies that these mutations arose independently [60]. Although these mutations are unlikely to strongly impact HIV-1 drug resistance, which in this study was already high, more studies are necessary to unveil the association between novel treatment-selected mutations and established DRM especially with regard to fitness. In our study, mutations I94L and L109I were significantly associated with time on treatment and in higher frequencies than previously reported. Therefore, the emergence of additional variation may be a consequence of long periods of time in treatment and the absence of adequate individual monitoring.
In high-income countries, routine virological monitoring has allowed the prompt detection of treatment failure and use of individual-based fully active ART regimens [2–4]. The overall picture appears to indicate a decrease of drug resistance, as more individuals on therapy become completely suppressed [5]. For example, Scherrer et al.[62] showed that in Switzerland the prevalence of drug resistance steadily decreased over 15 years as a result of individual virological monitoring and the use of highly effective first-line regimens. In contrast, major questions remain on the long-term and population-level impact of managing ART without adequate individual monitoring and the use of drugs with low genetic barriers to resistance in RLC. In addition to the rapid selection of two NNRTI-DRM (K103N and Y181C), which confer intermediate or high degree level of resistance to NVP and EFV, we evidenced a trend of increased DRM accumulation with time on ART. We clearly showed that the emergence of certain NNRTI mutations, all TAMs and mutation M184V were positively associated with time on treatment. The selection of multiple NNRTI-DRM and the accumulation of TAMs over time, associated with cross-resistance to all RTIs [63,64], reduces the clinical utility of several of the currently approved RTIs and, consequently, of drug options for second-line and subsequent regimens. For example, second-generation NNRTIs, which are only available as part of third-line regimens, may lose their ability to perform well because of the previous accumulation of mutations. Also, Ciaffi et al.[28] pointed out that late detection of treatment failure might lead to patients with high viral load, which has been significantly associated with shorter timeframe to second-line failure and may deserve special treatment strategies. The consequences of a delay until modification of a regimen may differ according to the original regimen combination. The administration of a regimen containing a protease inhibitor drug, as compared with a regimen that contains a NNRTI, may result in a lower prevalence of DRMs after virological failure [65–67]. The integrase inhibitor Dolutegravir (DTG) has shown a high barrier to resistance, and, in this context, the WHO now recommends its use as an alternative first-line option [68–70]. Nonetheless, even with the use of highly potent antiretroviral drugs, clinical monitoring would continue to be a major barrier to optimal management as treatment failure is often detected late [71]. Therefore, both adequate regimens (such as first-line DTG-based regimens) and virological monitoring are necessary to better control resistance and to guarantee the use of RTIs in a durable manner.
Graham et al.[72] evidenced the continuous shed of resistant virus and the steady accumulation of mutations in the female genital tract after virological failure, stressing that undetected treatment failure increases the risk for TDR. Treatment-naive individuals with resistant virus have also been shown to drive TDR [73,74], and baseline resistance has been significantly associated with a shorter timeframe to virological failure in individuals receiving standard first-line regimens [75]. The impact on drug susceptibility by NNRTI-TDR has been shown to be substantial in both high-income and resource-limited settings [16,76]. We also documented high prevalence of K103N and increased frequency of TAMs, which upon transmission persist longer in the absence of drug pressure [74,77]. Consequently, the scale-up of viral load testing to promptly detect virological failure will prevent the emergence of complex drug resistance and will contribute to a reduction of TDR.
Previous reports showed that certain DRMs are differentially selected by standard regimens according to HIV-1 subtype [20,78]. Moreover, infections with subtype C have a shorter timeframe to secondary virological failure when compared with subtype B [79]. Similar to other studies, we also observed a higher frequency of V106M in subtype C [80,81]. We also observed an increased presence of K65R in subtype C, but in contrast to the study from Theys et al.[82], the difference was NS in our study. However, our number of subtype C samples was lower. Overall, given the high prevalence of this subtype for certain counties of sub-Saharan Africa, it will be important to continue surveillance of the preferential selection of DRMs by this subtype.
Higher frequencies of infections caused by recombinant viruses have been observed over the last decades, and the cocirculation of multiple variants increases the emergence of more diverse and complex recombinant forms [83–86]. Given so, the effect of CRF-specific responses on ART may be diminished in regions where many variants cocirculate. Here, using a representative sample from West and Central African countries, we documented 18 CRFs and a substantial proportion of URFs in both the RTI-naive (16.6%) and experienced groups (17.8%). We documented two mutations differentially associated with CRF02_AG and CRF06_cpx that are less studied but have a high prevalence in some of these countries. However, these mutations only differ in proportions and therefore have no major impact on actual predictions of ART outcome. Nonetheless, it cannot be excluded that the use of other drug regimens can have different impacts on certain CRFs. In particular, the long-term use of the actual recommended TDF-based first-line regimens has to be evaluated. Higher TDF resistance rates were already observed in sub-Saharan countries where many people continue to be on ART with suboptimal monitoring [19,87–90].
In conclusion, the absence of adequate monitoring in West and Central African countries leads to lack of awareness of both virological failure (with substantial acquired resistance to drugs from first-line standard regimens and to drugs not yet taken by the patients) and the selection of rare mutations (though more studies are necessary to attribute their selection to resistance or fitness). Failing to recognize virological failure under NNRTI-based regimens leads to a rapid selection of resistance to this drug class and to the accumulation of known RTI-DRM over time. The accumulation of mutations limits future treatment options and increases the risk of transmission of HIV-1 drug-resistant strains; these observations stress the importance of rapidly scaling-up virological tests and using more potent first-line drugs to preserve the long-term use of ART treatment programs in RLC. In addition, to confirm a higher frequency of V106M in subtype C, we did not find CRF-specific mutations associated with drug resistance. Therefore, recombinants with high prevalence in West and Central African countries, in particular CRF02_AG, do not differentially impact prediction of treatment outcomes. This validates the global robustness of the actual drugs resistance algorithms.
Acknowledgements
Role of each of the authors: C.J.V.A. designed the project, consolidated and analyzed the data, wrote the R-scripts, prepared the figures and tables, interpreted the data and wrote the manuscript. N.V. processed HIV samples, generated sequence data and determined HIV subtype. E.G. and L.S. processed HIV samples, generated sequence data and determined HIV subtype. E.D. conceived the project, provided supervision, interpreted the data and participated in the writing of the manuscript. O.G. designed the project, provided supervision, performed statistical analysis, interpreted the data and participated in the writing of the manuscript. M.P. conceived and designed the project, provided supervision, interpreted the data and participated in the writing of the manuscript.
This work was supported by IRD (Institut de recherche pour le développement). CJVA was supported by a fellowship from the Labex EpiGenMed, via the National Research Agency, Programme for Future Investment “ANR-10-LABX-12-01” and University of Montpellier.
Conflicts of interest
There are no conflicts of interest.
References
1. WHO.
Antiretroviral therapy for HIV infection in adults and adolescents. 2010; Geneva, Switzerland: World Health Organization, 156.
2. Dragovic G, Smith CJ, Jevtovic D, Dimitrijevic B, Kusic J, Youle M, et al
Choice of first-line antiretroviral therapy regimen and treatment outcomes for HIV in a middle income compared to a high income country: a cohort study.
BMC Infect Dis 2016; 16:106.
3. Sullivan PS, Jones JS, Baral SD.
The global north: HIV epidemiology in high-income countries.
Curr Opin HIV AIDS 2014; 9:199–205.
4. Richardson ET, Grant PM, Zolopa AR.
Evolution of HIV treatment guidelines in high- and low-income countries: converging recommendations.
Antiviral Res 2014; 103:88–93.
5. Vermund SH, Leigh-Brown AJ.
The HIV epidemic: high-income countries.
Cold Spring Harb Perspect Med 2012; 2:a007195.
6. Roberts T, Cohn J, Bonner K, Hargreaves S.
Scale-up of routine viral load testing in resource-poor settings: current and future implementation challenges.
Clin Infect Dis 2016; 62:1043–1048.
7. Fokam J, Elat JB, Billong SC, Kembou E, Nkwescheu AS, Obam NM, et al
Monitoring HIV drug resistance early warning indicators in Cameroon: a study following the revised World Health Organization recommendations.
PLoS One 2015; 10:e0129210.
8. Jonas A, Sumbi V, Mwinga S, DeKlerk M, Tjituka F, Penney S, et al
HIV drug resistance early warning indicators in Namibia with updated World Health Organization guidance.
PLoS One 2014; 9:e100539.
9. Meresse M, March L, Kouanfack C, Bonono RC, Boyer S, Laborde-Balen G, et al
Patterns of adherence to antiretroviral therapy and HIV drug resistance over time in the Stratall ANRS 12110/ESTHER trial in Cameroon.
HIV Med 2014; 15:478–487.
10. Konou AA, Dagnra AY, Vidal N, Salou M, Adam Z, Singo-Tokofai A, et al
Alarming rates of virological failure and drug resistance in patients on long-term antiretroviral treatment in routine HIV clinics in Togo.
AIDS 2015; 29:2527–2530.
11. Sigaloff KC, Ramatsebe T, Viana R, de Wit TF, Wallis CL, Stevens WS.
Accumulation of HIV drug resistance mutations in patients failing first-line antiretroviral treatment in South Africa.
AIDS Res Hum Retroviruses 2012; 28:171–175.
12. Wamalwa DC, Lehman DA, Benki-Nugent S, Gasper MA, Gichohi R, Maleche-Obimbo E, et al
Long-term virologic response and genotypic resistance mutations in HIV-1 infected Kenyan children on combination antiretroviral therapy.
J Acquir Immune Defic Syndr 2013; 62:267–274.
13. Barth RE, Aitken SC, Tempelman H, Geelen SP, van Bussel EM, Hoepelman AI, et al
Accumulation of drug resistance and loss of therapeutic options precede commonly used criteria for treatment failure in HIV-1 subtype-C-infected patients.
Antivir Ther 2012; 17:377–386.
14. Hamers RL, Sigaloff KC, Kityo C, Mugyenyi P, de Wit TF.
Emerging HIV-1 drug resistance after roll-out of antiretroviral therapy in sub-Saharan Africa.
Curr Opin HIV AIDS 2013; 8:19–26.
15. Magambo B, Nazziwa J, Bbosa N, Gupta RK, Kaleebu P, Parry CM.
The arrival of untreatable multidrug-resistant HIV-1 in sub-Saharan Africa.
AIDS 2014; 28:1373–1374.
16. Rhee SY, Blanco JL, Jordan MR, Taylor J, Lemey P, Varghese V, et al
Geographic and temporal trends in the molecular epidemiology and genetic mechanisms of transmitted HIV-1 drug resistance: an individual-patient- and sequence-level meta-analysis.
PLoS Med 2015; 12:e1001810.
17. Jourdain G, Le Coeur S, Ngo-Giang-Huong N, Traisathit P, Cressey TR, Fregonese F, et al
Switching HIV treatment in adults based on CD4 count versus viral load monitoring: a randomized, noninferiority trial in Thailand.
PLoS Med 2013; 10:e1001494.
18. Okoboi S, Ekwaru PJ, Campbell JD, Egessa A, King R, Bakanda C, et al
No differences in clinical outcomes with the addition of viral load testing to CD4 cell count monitoring among HIV infected participants receiving ART in rural Uganda: long-term results from the Home Based AIDS Care Project.
BMC Public Health 2016; 16:101.
19. Brenner BG, Oliveira M, Doualla-Bell F, Moisi DD, Ntemgwa M, Frankel F, et al
HIV-1 subtype C viruses rapidly develop K65R resistance to tenofovir in cell culture.
AIDS 2006; 20:F9–F13.
20. Brenner B, Turner D, Oliveira M, Moisi D, Detorio M, Carobene M, et al
A V106M mutation in HIV-1 clade C viruses exposed to efavirenz confers cross-resistance to nonnucleoside reverse transcriptase inhibitors.
AIDS 2003; 17:F1–5.
21. Loemba H, Brenner B, Parniak MA, Ma’ayan S, Spira B, Moisi D, et al
Genetic divergence of human immunodeficiency virus type 1 Ethiopian clade C reverse transcriptase (RT) and rapid development of resistance against nonnucleoside inhibitors of RT.
Antimicrob Agents Chemother 2002; 46:2087–2094.
22. Nzeyimana SD, Baramperanye E, Mouacha F, Nindagiye E, Ntahobari S, Ndayiragije E, et al
Monitoring of HIV-1 drug resistance and associated programmatic factors in patients initiating antiretroviral therapy at two ART sites in Bujumbura, Burundi.
Antivir Ther 2010; 15:A153.
23. Vergne L, Diagbouga S, Kouanfack C, Aghokeng A, Butel C, Laurent C, et al
HIV-1 drug-resistance mutations among newly diagnosed patients before scaling-up programmes in Burkina Faso and Cameroon.
Antivir Ther 2006; 11:575–579.
24. Aghokeng AF, Kouanfack C, Laurent C, Ebong E, Atem-Tambe A, Butel C, et al
Scale-up of antiretroviral treatment in sub-Saharan Africa is accompanied by increasing HIV-1 drug resistance mutations in drug-naive patients.
AIDS 2011; 25:2183–2188.
25. Somda A, Sangare L, Soro M, Yameogo S, Bazie B, Bigirimana F, et al
Surveillance of transmitted drug-resistant HIV among young pregnant women in Ouagadougou, Burkina Faso.
Clin Infect Dis 2012; 54 (suppl 4):S317–319.
26. Vergne L, Stuyver L, Van Houtte M, Butel C, Delaporte E, Peeters M.
Natural polymorphism in protease and reverse transcriptase genes and in vitro antiretroviral drug susceptibilities of non-B HIV-1 strains from treatment-naive patients.
J Clin Virol 2006; 36:43–49.
27. Vidal N, Niyongabo T, Nduwimana J, Butel C, Ndayiragije A, Wakana J, et al
HIV type 1 diversity and antiretroviral drug resistance mutations in Burundi.
AIDS Res Hum Retroviruses 2007; 23:175–180.
28. Ciaffi L, Koulla-Shiro S, Sawadogo A, le Moing V, Eymard-Duvernay S, Izard S, et al
Efficacy and safety of three second-line antiretroviral regimens in HIV-infected patients in Africa.
AIDS 2015; 29:1473–1481.
29. Aghokeng A, Vergne L, Mpoudi-Ngole E, Mbangue M, Deoudje N, Mokondji E, et al
Evaluation of transmitted HIV drug resistance among recently-infected antenatal clinic attendees in four Central African Countries.
Antivir Ther 2009; 14:401–411.
30. Vidal N, Mulanga C, Bazepeo SE, Mwamba JK, Tshimpaka J, Kashi M, et al
HIV type 1 pol gene diversity and antiretroviral drug resistance mutations in the Democratic Republic of Congo (DRC).
AIDS Res Hum Retroviruses 2006; 22:202–206.
31. Aghokeng AF, Mpoudi-Ngole E, Chia JE, Edoul EM, Delaporte E, Peeters M.
High failure rate of the ViroSeq HIV-1 genotyping system for drug resistance testing in Cameroon, a country with broad HIV-1 genetic diversity.
J Clin Microbiol 2011; 49:1635–1641.
32. Djoko CF, Rimoin AW, Vidal N, Tamoufe U, Wolfe ND, Butel C, et al
High HIV type 1 group M pol diversity and low rate of antiretroviral resistance mutations among the uniformed services in Kinshasa, Democratic Republic of the Congo.
AIDS Res Hum Retroviruses 2011; 27:323–329.
33. Ayouba A, Lien TT, Nouhin J, Vergne L, Aghokeng AF, Ngo-Giang-Huong N, et al
Low prevalence of HIV type 1 drug resistance mutations in untreated, recently infected patients from Burkina Faso, Cote d’Ivoire, Senegal, Thailand, and Vietnam: the ANRS 12134 study.
AIDS Res Hum Retroviruses 2009; 25:1193–1196.
34. Yaotse DA, Nicole V, Roch NF, Mireille PD, Eric D, Martine P.
Genetic characterization of HIV-1 strains in Togo reveals a high genetic complexity and genotypic drug-resistance mutations in ARV naive patients.
Infect Genet Evol 2009; 9:646–652.
35. Djoko CF, Wolfe ND, Vidal N, Tamoufe U, Montavon C, LeBreton M, et al
HIV type 1 pol gene diversity and genotypic antiretroviral drug resistance mutations in Malabo, Equatorial Guinea.
AIDS Res Hum Retroviruses 2010; 26:1027–1031.
36. Vergne L, Kane CT, Laurent C, Diakhate N, Gueye NF, Gueye PM, et al
Low rate of genotypic HIV-1 drug-resistant strains in the Senegalese government initiative of access to antiretroviral therapy.
AIDS 2003; 17 (suppl 3):S31–38.
37. Monleau M, Faihun F, Affolabi D, Afangnihoun A, Boillot F, Anagonou S, et al
Antiretroviral drug resistance in HIV-1 infected patients receiving antiretroviral treatment in routine clinics in Cotonou, Benin.
J AIDS HIV Res 2011; 3:114–120.
38. Kouanfack C, Montavon C, Laurent C, Aghokeng A, Kenfack A, Bourgeois A, et al
Low levels of antiretroviral-resistant HIV infection in a routine clinic in Cameroon that uses the World Health Organization (WHO) public health approach to monitor antiretroviral treatment and adequacy with the WHO recommendation for second-line treatment.
Clin Infect Dis 2009; 48:1318–1322.
39. Aghokeng AF, Kouanfack C, Eymard-Duvernay S, Butel C, Edoul GE, Laurent C, et al
Virological outcome and patterns of HIV-1 drug resistance in patients with 36 months’ antiretroviral therapy experience in Cameroon.
J Int AIDS Soc 2013; 16:18004.
40. Taieb F, Aghokeng AF, Eymard-Duvernay S, Chia JE, Einterz E, Mpoudi-Ngole E, et al
Challenges of antiretroviral treatment monitoring in rural and remote-access regions in Africa.
AIDS Res Hum Retroviruses 2014; 30:623–625.
41. Dagnra AY, Vidal N, Mensah A, Patassi A, Aho K, Salou M, et al
High prevalence of HIV-1 drug resistance among patients on first-line antiretroviral treatment in Lome, Togo.
J Int AIDS Soc 2011; 14:30.
42. Aghokeng AF, Monleau M, Eymard-Duvernay S, Dagnra A, Kania D, Ngo-Giang-Huong N, et al
Extraordinary heterogeneity of virological outcomes in patients receiving highly antiretroviral therapy and monitored with the World Health Organization public health approach in sub-saharan Africa and southeast Asia.
Clin Infect Dis 2014; 58:99–109.
43. Boillot F, Serrano L, Muwonga J, Kabuayi JP, Kambale A, Mutaka F, et al
Implementation and operational research: programmatic feasibility of dried blood spots for the virological follow-up of patients on antiretroviral treatment in Nord Kivu, Democratic Republic of the Congo.
J Acquir Immune Defic Syndr 2016; 71:e9–e15.
44. 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.
45. Konou AA, Salou M, Vidal N, Kodah P, Kombate D, Kpanla P, et al
Virological outcome among HIV-1 infected patients on first-line antiretroviral treatment in semi-rural HIV clinics in Togo.
AIDS Res Ther 2015; 12:38.
46. WHO.
Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection. 2013; Geneva, Switzerland: World Health Organization, 272.
47. Edgar RC.
MUSCLE: a multiple sequence alignment method with reduced time and space complexity.
BMC Bioinformatics 2004; 5:113.
48. Edgar RC.
MUSCLE: multiple sequence alignment with high accuracy and high throughput.
Nucleic Acids Res 2004; 32:1792–1797.
49. Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel O.
New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0.
Syst Biol 2010; 59:307–321.
50. Lole KS, Bollinger RC, Paranjape RS, Gadkari D, Kulkarni SS, Novak NG, et al
Full-length human immunodeficiency virus type 1 genomes from subtype C-infected seroconverters in India, with evidence of intersubtype recombination.
J Virol 1999; 73:152–160.
51. Wensing AM, Calvez V, Gunthard HF, Johnson VA, Paredes R, Pillay D, et al
2015 update of the drug resistance mutations in HIV-1.
Top Antivir Med 2015; 23:132–141.
52. Shahriar R, Rhee SY, Liu TF, Fessel WJ, Scarsella A, Towner W, et al
Nonpolymorphic human immunodeficiency virus type 1 protease and reverse transcriptase treatment-selected mutations.
Antimicrob Agents Chemother 2009; 53:4869–4878.
53. Benjamini Y, Yekutieli D.
The control of the false discovery rate in multiple testing under dependency.
Ann Stat 2001; 4:1165–1188.
54. Machekano RN, Hubbard AE.
Empirical Bayes approach to controlling familywise error: an application to HIV resistance data.
UCB Biostatistics 2006; 205:1–21.
55. Cane PA, Green H, Fearnhill E, Dunn D. Resistance UKcgoHD.
Identification of accessory mutations associated with high-level resistance in HIV-1 reverse transcriptase.
AIDS 2007; 21:447–455.
56. Shafer RW, Schapiro JM.
HIV-1 drug resistance mutations: an updated framework for the second decade of HAART.
AIDS Rev 2008; 10:67–84.
57. Shafer RW, Rhee SY, Bennett DE.
Consensus drug resistance mutations for epidemiological surveillance: basic principles and potential controversies.
Antivir Ther 2008; 13 (suppl 2):59–68.
58. Bennett DE, Camacho RJ, Otelea D, Kuritzkes DR, Fleury H, Kiuchi M, et al
Drug resistance mutations for surveillance of transmitted HIV-1 drug-resistance: 2009 update.
PLoS One 2009; 4:e4724.
59. Kantor R, Katzenstein DA, Efron B, Carvalho AP, Wynhoven B, Cane P, et al
Impact of HIV-1 subtype and antiretroviral therapy on protease and reverse transcriptase genotype: results of a global collaboration.
PLoS Med 2005; 2:e112.
60. Rhee SY, Sankaran K, Varghese V, Winters M, Hurt CB, Eron JJ, et al
HIV-1 protease, reverse transcriptase, and integrase variation.
J Virol 2016; 90:6058–6070.
61. Parkin NT, Gupta S, Chappey C, Petropoulos CJ.
The K101P and K103R/V179D mutations in human immunodeficiency virus type 1 reverse transcriptase confer resistance to nonnucleoside reverse transcriptase inhibitors.
Antimicrob Agents Chemother 2006; 50:351–354.
62. Scherrer AU, von Wyl V, Yang WL, Kouyos R, Böni J, Yerly S, et al
Emergence of acquired HIV-1 drug resistance has almost been stopped in Switzerland – a 15 year prospective cohort analysis.
Clin Infect Dis 2016; 62:1310–1317.
63. Stevens WS.
Slow accumulation of HIV resistance mutations: implications for resource-limited settings?.
J Infect Dis 2009; 200:670–672.
64. Cozzi-Lepri A, Phillips AN, Martinez-Picado J, Monforte A, Katlama C, Eg Hansen AB, et al
Rate of accumulation of thymidine analogue mutations in patients continuing to receive virologically failing regimens containing zidovudine or stavudine: implications for antiretroviral therapy programs in resource-limited settings.
J Infect Dis 2009; 200:687–697.
65. Hill A, McBride A, Sawyer AW, Clumeck N, Gupta RK.
Resistance at virological failure using boosted protease inhibitors versus nonnucleoside reverse transcriptase inhibitors as first-line antiretroviral therapy – implications for sustained efficacy of ART in resource-limited settings.
J Infect Dis 2013; 207 (suppl 2):S78–84.
66. Hosseinipour MC, Gupta RK, Van Zyl G, Eron JJ, Nachega JB.
Emergence of HIV drug resistance during first- and second-line antiretroviral therapy in resource-limited settings.
J Infect Dis 2013; 207 (suppl 2):S49–S56.
67. Gupta R, Hill A, Sawyer AW, Pillay D.
Emergence of drug resistance in HIV type 1-infected patients after receipt of first-line highly active antiretroviral therapy: a systematic review of clinical trials.
Clin Infect Dis 2008; 47:712–722.
68. Walmsley SL, Antela A, Clumeck N, Duiculescu D, Eberhard A, Gutierrez F, et al
Dolutegravir plus abacavir-lamivudine for the treatment of HIV-1 infection.
N Engl J Med 2013; 369:1807–1818.
69. Mesplede T, Wainberg MA.
Is resistance to dolutegravir possible when this drug is used in first-line therapy?.
Viruses 2014; 6:3377–3385.
70. WHO.
What's new in HIV treatment.
Fact sheet: HIV treatment and care 2015; Geneva, Switzerland: World Health Organization, 2.
71. Haas AD, Keiser O, Balestre E, Brown S, Bissagnene E, Chimbetete C, et al
Monitoring and switching of first-line antiretroviral therapy in adult treatment cohorts in sub-Saharan Africa: collaborative analysis.
The Lancet HIV 2015; 2:e271–e278.
72. Graham SM, Chohan V, Ronen K, Deya RW, Masese LN, Mandaliya KN, et al
Genital shedding of resistant human immunodeficiency virus-1 among women diagnosed with treatment failure by clinical and immunologic monitoring.
Open Forum Infect Dis 2016; 3:ofw019.
73. Yang WL, Kouyos R, Scherrer AU, Boni J, Shah C, Yerly S, et al
Assessing the paradox between transmitted and acquired HIV type 1 drug resistance mutations in the Swiss HIV cohort study from 1998 to 2012.
J Infect Dis 2015; 212:28–38.
74. Mourad R, Chevennet F, Dunn DT, Fearnhill E, Delpech V, Asboe D, et al
A phylotype-based analysis highlights the role of drug-naive HIV-positive individuals in the transmission of antiretroviral resistance in the UK.
AIDS 2015; 29:1917–1925.
75. Kantor R, Smeaton L, Vardhanabhuti S, Hudelson SE, Wallis CL, Tripathy S, et al
Pretreatment HIV drug resistance and HIV-1 subtype C are independently associated with virologic failure: results from the multinational PEARLS (ACTG A5175) clinical trial.
Clin Infect Dis 2015; 60:1541–1549.
76. 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.
77. Yang WL, Kouyos RD, Boni J, Yerly S, Klimkait T, Aubert V, et al
Persistence of transmitted HIV-1 drug resistance mutations associated with fitness costs and viral genetic backgrounds.
PLoS Pathog 2015; 11:e1004722.
78. Sluis-Cremer N, Jordan MR, Huber K, Wallis CL, Bertagnolio S, Mellors JW, et al
E138A in HIV-1 reverse transcriptase is more common in subtype C than B: implications for rilpivirine use in resource-limited settings.
Antiviral Res 2014; 107:31–34.
79. Häggblom A, Svedhem V, Singh K, Sönnerborg A, Neogi U.
Virological failure in patients with HIV-1 subtype C receiving antiretroviral therapy: an analysis of a prospective national cohort in Sweden.
Lancet HIV 2016; 3:e166–e174.
80. Santoro MM, Perno CF.
HIV-1 genetic variability and clinical implications.
ISRN Microbiol 2013; 2013:481314.
81. Wainberg MA, Brenner BG.
The impact of HIV genetic polymorphisms and subtype differences on the occurrence of resistance to antiretroviral drugs.
Mol Biol Int 2012; 2012:256982.
82. Theys K, Vercauteren J, Snoeck J, Zazzi M, Camacho RJ, Torti C, et al
HIV-1 subtype is an independent predictor of reverse transcriptase mutation K65R in HIV-1 patients treated with combination antiretroviral therapy including tenofovir.
Antimicrob Agents Chemother 2013; 57:1053–1056.
83. Neogi U, Haggblom A, Santacatterina M, Bratt G, Gisslen M, Albert J, et al
Temporal trends in the Swedish HIV-1 epidemic: increase in non-B subtypes and recombinant forms over three decades.
PLoS One 2014; 9:e99390.
84. Jacobs GB, Wilkinson E, Isaacs S, Spies G, de Oliveira T, Seedat S, et al
HIV-1 subtypes B and C unique recombinant forms (URFs) and transmitted drug resistance identified in the Western Cape Province, South Africa.
PLoS One 2014; 9:e90845.
85. Hemelaar J, Gouws E, Ghys PD, Osmanov S. WHO-UNAIDS Network for HIV Isolation and Characterisation.
Global trends in molecular epidemiology of HIV-1 during 2000–2007.
AIDS 2011; 25:679–689.
86. Zhang M, Foley B, Schultz AK, Macke JP, Bulla I, Stanke M, et al
The role of recombination in the emergence of a complex and dynamic HIV epidemic.
Retrovirology 2010; 7:25.
87. Wainberg MA.
Early HIV treatment to forestall drug resistance.
Lancet Infect Dis 2016; 16:512–513.
88. TenoRes Study Group.
Global epidemiology of drug resistance after failure of WHO recommended first-line regimens for adult HIV-1 infection: a multicentre retrospective cohort study.
Lancet Infect Dis 2016; [Epub ahead of print].
89. Skhosana L, Steegen K, Bronze M, Lukhwareni A, Letsoalo E, Papathanasopoulos MA, et al
High prevalence of the K65R mutation in HIV-1 subtype C infected patients failing tenofovir-based first-line regimens in South Africa.
PLoS One 2015; 10:e0118145.
90. Sunpath H, Wu B, Gordon M, Hampton J, Johnson B, Moosa MY, et al
High rate of K65R for antiretroviral therapy-naive patients with subtype C HIV infection failing a tenofovir-containing first-line regimen.
AIDS 2012; 26:1679–1684.
