Share this article on:

Varied Patterns of HIV-1 Drug Resistance on Failing First-Line Antiretroviral Therapy in South Africa

Wallis, Carole L MSc (Med)*; Mellors, John W MD; Venter, Willem D F MD; Sanne, Ian MD§; Stevens, Wendy MD

JAIDS Journal of Acquired Immune Deficiency Syndromes: April 1st, 2010 - Volume 53 - Issue 4 - p 480-484
doi: 10.1097/QAI.0b013e3181bc478b
Clinical Science

Background: The South African national antiretroviral therapy roll-out program is entering its sixth year, with over 570,000 adults accessing treatment. HIV-1 drug resistance is a potential consequence of therapy. This study determined the pattern of HIV-1 drug resistance mutations after failure of first-line treatment regimens in South Africa.

Methods: Two hundred and twenty-six patients virologically failing first-line regimens were studied to determine resistance patterns.

Results: The most common reverse transcriptase mutation was M184V/I (72%; n = 163); 11% of patients (n = 25) had only nonnucleoside reverse transcriptase inhibitor (NNRTI) mutations and 17% (n = 38) had no known resistance mutations. The K65R mutation was detected in 4%. The frequency of thymidine analog mutations was significantly higher with azidothymidine-containing (31 of 57) than stavudine-containing regimens (39 of 169; P < 0.001). The Y181C mutation was more frequent with failure of nevirapine (NVP)-containing (26%) than efavirenz (EFV)-containing therapy (3%; P < 0.001). The V106M mutation was more frequent with EFV (30%) than NVP (4%; P = 0.012).

Conclusions: HIV-1 drug resistance patterns varied broadly after failure of first-line therapy, ranging from no known resistance mutations (17%) to multinucleoside reverse transcriptase inhibitor and NNRTI resistance (23%). NNRTI mutation profiles differed for patients on EFV- compared with NVP-containing regimens. Overall, these findings suggest that HIV-1 drug resistance testing would be useful in identifying most appropriate second-line regimens.

From the *Department of Molecular Medicine and Haematology, University of the Witwatersrand, Johannesburg, South Africa; †Department of Infectious Disease, University of Pittsburgh, USA; ‡RHRU, University of the Witwatersrand, Johannesburg, South Africa; §CHRU, University of the Witwatersrand, Johannesburg, South Africa; and ¶National Health Laboratory Services (NHLS), Johannesburg, South Africa.

Received for publication May 27, 2009; accepted August 10, 2009.

This work was presented as a poster at Montreal, Canada, at the 16th Conference of Retroviruses and Opportunistic Infections and the XVIII International HIV Drug Resistance Workshop.

Supported by Willem D. F. Venter acknowledges supported from the President's Emergency Plan for AIDS Relief (PEPFAR) and an National Institute of Health (NIH) International Clinical, Operational, and Health Services Research and Training Award (ICOHRTA). John W. Mellors acknowledges support from the AIDS Clinical Trials Group (NIAID U01AI38858) and the National Cancer Institute (SAIC contract 20XS190A).

Correspondence: Carole L. Wallis, MSc (Med), 3B20, 3rd Floor, Wits Medical School, 7 York Rd, Parktown, 2193, Johannesburg, South Africa (e-mail:

Back to Top | Article Outline


Global access to antiretroviral therapy (ART) has increased rapidly in recent years in developing countries, where 3 million of the 9.7 million infected patients are now receiving ART.1 ART can fail as a result of toxicity, transmitted drug resistance, inadequate medication adherence, or incomplete suppression of viral replication resulting in the emergence of drug-resistant HIV-1. There is currently only a partial understanding of patterns of HIV-1 drug resistance. Moreover, the subtypes circulating in developing countries are generally not subtype B, whereas most drug resistance data have been generated for HIV-1 subtype B in North America and Europe.

There is conflicting evidence as to whether resistance patterns differ among HIV-1 subtypes after first-line ART failure. An initial study from Zimbabwe suggested that resistance mutations developing in HIV-1 subtype C were similar to those described for subtype B.2 However, more recent studies by Brenner3 and Morris4 identified subtype C-specific resistance mutations. For example, the V106M mutation in HIV-1 reverse transcriptase (RT) is frequent with failure of efavirenz (EFV)- or nevirapine (NVP)-containing therapy, in contrast to the V106A mutation that infrequently emerges with the same drugs in HIV-1 subtype B. The K103N mutation develops at a greater frequency in women infected with subtypes C and D as compared with subtype A.5 In addition, the K65R mutation in RT is selected by nucleoside reverse transcriptase inhibitor (NRTI) more frequently in HIV-1 subtype C than subtype B, both in vitro and in patients on failing therapy.6-8 Finally, a number of other studies have confirmed the presence of different baseline protease (PR) polymorphisms in HIV-1 subtype C vs. subtype B.9-13

A recent report of patients with failure of first-line ART [stravudine (d4T)/lamivudine (3TC)/EFV] in Malawi showed that 93% of patients had a nonnucleoside reverse transcriptase inhibitors (NNRTI) mutation and 81% the M184V mutation.14 The M184V/I mutation did not occur in isolation; 56% had additional thymidine analog mutations (TAMs) and 44% of this group also had 3 of more TAMs. Seventeen percent demonstrated resistance to all NRTI currently available. Similarly, data from the Nigerian ACTION program revealed that 53% of patients on a failing ART regimen had no active NRTI available to them.15

In South Africa, the national roll-out program is now entering its sixth year with over 570,000 individuals receiving ART.16 The program, which relies on a d4T/3TC backbone with an NNRTI, makes provision for both biannual CD4 and plasma HIV-1 RNA monitoring, but not for routine HIV-1 drug resistance genotyping. Switches for toxicity are permitted, and usually involve changing from d4T to azidothymidine (AZT).17 A switch to lopinavir/ritonavir-based second-line therapy with an AZT/didanosine backbone is recommended when plasma HIV-1 RNA is elevated on 2 consecutive measurements, with tenofovir available for those with toxicity to any NRTI.

An ongoing concern about ART roll-out is the selection and potential spread of HIV-1 drug resistance. We therefore describe HIV-1 drug resistance patterns in RT and PR among patients with virologic failure on a variety of first-line regimens in South Africa.

Back to Top | Article Outline


Patient Samples

Plasma samples from patients on failing ART were sent for HIV-1 drug resistance testing from the adult HIV-1 clinics at 2 large state hospitals in Johannesburg, South Africa. Virologic failure varied between the hospitals, and was defined as having confirmed (2 consecutive measurements) plasma HIV-1 RNA greater than 5000 copies/mL (Charlotte Maxele Academic Hospital) or 1000 HIV-1 RNA copies/mL (Helen Joseph Hospital), despite adherence counseling. Data on length of time on the failing regimen were not available.

Back to Top | Article Outline

Population Genotype Analysis

Population-based genotyping was performed using an in-house drug resistance assay (Wallis CL, Papathanasopoulos M.A., Rinke de Wit T.F., Stevens W. Subtype C specific In-house HIV-1 Drug Resistance Assay, 5th European HIV Drug Resistance Workshop, 28-30 March 2007, Cascais, Portugal). Briefly, viral RNA was extracted from 200 μL of plasma samples using the automated Roche MagNa Pure LC analyzer and the MagNA Pure LC Total Nucleic Acid Isolation Kit (Roche, Germany). A 1.7-kb amplicon was generated by RT-initiated polymerase chain reaction encompassing the entire PR and partial RT coding regions using primers designed from the consensus HIV-1 subtype C sequence available on the Los Alamos Database ( The amplicon was sequenced using 5 primers that ensure bidirectional coverage from codons 1 to 99 of PR and codons 1 to 230 of RT. Sequencing was performed with an ABI Prism 3100 Avant Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).

Back to Top | Article Outline

Data Analysis

Sequences were assembled, manually edited using Sequencher version 4.5 software (Genecodes, Ann Arbor, MI), and submitted to the ViroScore database, which uses the IAS-USA mutation list18 to identify known HIV-1 drug resistance mutations associated with decreased activity of the PR and RT inhibitors. The frequency distribution for each of these mutations was analyzed by ART regimen. The REGA HIV-1 subtyping tool ( on the Stanford database was used to determine HIV-1 subtype of each patient sample. A nonparametric chi-square test was used to determine significance. A P value of <0.05 was considered significant.

Back to Top | Article Outline


A total of 226 patients on failing first-line therapy in the South African national roll-out program were studied. The baseline characteristics and failing regimens are reported in Table 1. At time of resistance testing, the median plasma HIV-1 RNA level was 22,000 copies/mL and the median CD4 cell count was 165 cells/μL. The majority of the patients were receiving d4T, 3TC, EFV (65%; 147 of 226), or AZT, 3TC, EFV (23%; 52 of 226); 22 were on d4T, 3TC, NVP, and 5 on AZT, 3TC, NVP. All drugs are provided as separate formulations. The most common mutation found was the M184V/I (72%; n = 163) but 11% of patients (n = 25) had only NNRTI mutations and 17% (n = 38) had no known drug-resistance mutations (Fig. 1). The K65R mutation (4%) was observed without TAMs and in combination with the Q151M complex in 3 of 7 cases (Fig. 1). Only 12% of patients had more than 3 TAMs, with the D67N pathway predominating over the M41L pathway.





The frequency of TAMs was significantly higher with AZT- (31 of 55; 56%) than d4T-containing (39 of 169; 23%) regimens (P < 0.001; Fig. 2). The most common TAMs was D67N, occurring in 35% (19 of 55) of patients on AZT, but in only 14% (24 of 169) of patients on the d4T regimen (P = 0.015). The K70R mutation was the second most common mutation in the with AZT- and d4T-containing regimens (Fig. 2).



Of the 23% (n = 51) with complex resistance profiles; 40 had 2 or more TAMs resulting in variable resistance to Food and Drug Administration-approved NRTIs (Table 1). The remaining 11 patients had the K65R (n = 6; 3%), Q151M (n = 2; 1%), and/or K65R and Q151M (n = 3; 1%). Virus with K65R would be expected to have reduced susceptibility to all approved NRTIs except AZT. Similarly, Q151M would be expected to confer reduced susceptibility to all approved NRTI except tenofovir. K65R with Q151M would be expected to confer resistance all approved NRTI. A generalized nonlinear model of multiple variable analysis was performed for CD4, viral load, age, and sex and none were associated with a greater risk of complex mutation patterns (Table 1).

The NNRTI mutation profiles differed for patients failing EFV- vs. NVP-containing regimens. The Y181C mutation was more frequent with failure of NVP- (26%) than EFV- (3%) containing therapy (P < 0.001; Fig. 3). As expected, the V106M mutation was more frequent (P = 0.012) with EFV (n = 59; 30%) than NVP (n = 1; 4%). The K103N mutation occurred at a higher frequency on EFV (n = 78; 40%) than NVP (n = 10; 37%), but this difference was not significant (P = 0.37). EFV seemed to select for a wider range of mutations in the RT compared with NVP (Fig. 3). Of the 19 NNRTI resistance mutations detected, 8 were associated with both EFV and NVP therapy (Fig. 3). The other 9 mutations only occurred with EFV therapy (L100I; K101P; V106I; V179D; Y181I; Y188C/H; G190S; and M230L).



In the 226 patients failing first-line therapy, no major PR mutations were observed; however, naturally occurring polymorphisms in HIV-1 subtype C [K20R (30%); M36I (85%); H69K (100%); and I93L (99%)] were detected.

Back to Top | Article Outline


This study describes the HIV-1 drug resistance patterns associated with first-line regimen failures in the South African national roll-out program. Eighty-four percent of patients had virologic failure with known NRTI and/or NNRTI mutations. The mutation profiles were similar to that observed in subtype B, with the exception of a higher frequency of V106M and K65R. The M184V/I mutation was observed at highest frequency, followed by several NNRTI mutations (K103N, V106M, and G190A), which is similar to recently published data from Cape Town and Durban, South Africa.19,20 The majority of patients in the current study failed as a result of M184V and an NNRTI mutation, preserving most second-line options. However, in 23% of patients, complex resistance patterns were observed (defined as the presence of ≥2 TAMs, K65R, and Q151M) indicating cross-resistance to most NRTI, which would likely impact second-line therapy options and supports the need for resistance testing.

Very few patients had isolated NNRTI mutations, suggesting that resistance to NNRTI is not the initial event in failure of regimens containing EFV or NVP. In addition, we found that 17% of patients had no resistance mutations at failure, which is similar to that reported for the Development of AntiRetroviral Therapy in Africa (DART) study (10%),21 but higher than that in Cape Town, South Africa (6.4%).20 The reason for this difference is not clear, but may be related to better medication adherence in the Cape Town clinic20 or mutations occurring outside the segment of RT that was sequenced, such as the N348I mutation in the connection domain of RT associated with initial failure of AZT, 3TC, and NVP22 and has been observed with and without TAMs.22 Full-length sequencing of RT from failure samples is planned.

Patients with first-line regimen failure did not have lopinavir/ritonavir resistance mutations, which is a key component of second-line therapy. However, 23% are likely to have reduced susceptibility to tenofovir, didanosine, or zidovudine, which are other components of second-line therapy. Performing resistance testing on first-line regimen failures will identify whether TAMs, K65R, or Q151M complex are present. Identifying each of these mutational patterns may be important in NRTI selection. The Q151M complex confers multi-NRTI resistance as can multiple TAMs, although the level of cross-resistance to NRTI with different TAMs patterns is variable. K65R confers resistance to tenofovir and didanosine, thus prescribing either of these drugs would likely be suboptimal to a zidovudine containing regimen.23

The most common TAMs detected was D67N, indicating that the majority of patients had the more favorable D67N pathway.24 Only 12% of patients had more than 3 TAMs, which is considerably less common than reported from Malawi, where 56% of patients had more than 3 TAMs. The higher reported frequency of multiple TAMs in Malawi may have been related to the duration of treatment failure before resistance testing or other factors. In our study, multiple TAMs were more frequent with AZT- than d4T-containing regimens, which is different than previously reported with longer duration of treatment failure in subtype B HIV-1 infection.25

The K65R mutation was observed in 4% of patients, of which 8 of the 9 were on d4T. Three of 5 patients with the Q151M complex also had K65R. The reason a subset of patients develops K65R with Q151M is not clear but deserves further study. All 3 of the patients with K65R and Q151M were on regimens containing d4T and 3TC. Only one patient had K65R with a TAMs (D67N and K70R), which is consistent with the antagonistic effect of these mutations.26,27 The reason d4T-containing regimens may select K65R alone or TAMs alone is also not known. The more frequent development of the K65R mutation may be a result of subtype C polymorphisms6,28 and/or a delay in treatment switch. Two published studies from South Africa show different frequencies of the K65R mutation. Orrell et al20 in Cape Town reported a frequency of K65R similar to our study, but Marconi et al19 in Durban reported a lower frequency of K65R mutation in patients sequenced soon after virologic failure.

The NNRTI profile observed varied depending on EFV or NVP use and could have an impact on the use of etravirine in third-line regimens. NVP selects for the Y181C and G190A mutations both of which result in reduced susceptibility to etravirine.29,30 By contrast, EFV most commonly selects for K103N which has no effect on response to etravirine. If etravirine becomes available for second- or third-line therapy, EFV use in first-line might be encouraged over NVP.

In conclusion, HIV-1 drug resistance in the majority of patients failing the first-line regimen of the South African roll-out program were detected before multiple mutations had evolved, leaving good current and future second-line treatment options. However, in a subset of patients (23%), the occurrence of K65R, Q151M complex, or multiple TAMs is likely to negatively impact on future regimens containing tenofovir and other NRTIs. Overall, the varied drug resistance patterns observed after failure of first-line therapy suggests that resistance testing may be useful in identifying the most appropriate second-line therapy, especially if new classes of antiretrovirals or additional agents within existing classes become available. Blind regimen “switches” are not likely to provide optimal second-line treatment responses in all patients, although the overall cost-effectiveness of resistance testing remains uncertain. Sentinel surveillance for drug resistance should be part of national treatment programs to identify effective and low-cost treatment regimen sequencing after first-line failure.

Back to Top | Article Outline


We would like to thank the patients for participating in this study, the nurses and clinics at Charlotte Maxele Academic Hospital and Helen Joseph Hospital. All staff in the genotyping laboratory at the University of the Witwatersrand for sample processing.

Back to Top | Article Outline


1. UNAIDS. 2007 AIDS Epidemic Upda te. Available at: Joint United Nations Programme on HIV/AIDS (UNAIDS) and World Health Organization (WHO) 2007. Accessed 30 September 2009.
2. Kantor R, Zijenah LS, Shafer RW, et al. HIV-1 subtype C reverse transcriptase and protease genotypes in Zimbabwean patients failing antiretroviral therapy. AIDS Res Hum Retroviruses. 2002;18:1407-1413.
3. Brenner B, Turner D, Oliveira M, et al. A V106M mutation in HIV-1 clade C viruses exposed to efavirenz confers cross-resistance to non-nucleoside reverse transcriptase inhibitors. AIDS. 2003;17:F1-F5.
4. Morris L, Pillay C, Chezzi C, et al. Low frequency of the V106M mutation among HIV-1 subtype C-infected pregnant women exposed to nevirapine. AIDS. 2003;17:1698-1700.
5. Flys TS, Chen S, Jones DC, et al. Quantitative analysis of HIV-1 variants with the K103N resistance mutation after single-dose nevirapine in women with HIV-1 subtypes A, C, and D. J Acquir Immune Defic Syndr. 2006;42:610-613.
6. Brenner BG, Oliveira M, Doualla-Bell F, et al. HIV-1 subtype C viruses rapidly develop K65R resistance to tenofovir in cell culture. AIDS. 2006;20:F9-F13.
7. Sirivichayakul S, Phanuphak P, Pankam T, et al. HIV drug resistance transmission threshold survey in Bangkok, Thailand. Antivir Ther. 2008;13(Suppl 2):109-113.
8. Wallis CL, Bell C, Boulme R, et al. Emerging HIV-1 drug resistance patterns from two clinics on the SA ARV roll-out programme. CROI. 2007:661.
9. Cane PA, de Ruiter A, Rice P, et al. Resistance-associated mutations in the human immunodeficiency virus type 1 subtype c protease gene from treated and untreated patients in the United Kingdom. J Clin Microbiol. 2001;39:2652-2654.
10. Grossman Z, Vardinon N, Chemtob D, et al. Genotypic variation of HIV-1 reverse transcriptase and protease: comparative analysis of clade C and clade B. AIDS. 2001;15:1453-1460.
11. Kantor R, Katzenstein D. Polymorphism in HIV-1 non-subtype B protease and reverse transcriptase and its potential impact on drug susceptibility and drug resistance evolution. AIDS Rev. 2003;5:25-35.
12. Petch LA, Hoffman IF, Jere CS, et al. Genotypic analysis of the protease and reverse transcriptase of HIV type 1 subtype C isolates from antiretroviral drug-naive adults in Malawi. AIDS Res Hum Retroviruses. 2005;21:799-805.
13. Rhee SY, Kantor R, Katzenstein DA, et al. HIV-1 pol mutation frequency by subtype and treatment experience: extension of the HIVseq program to seven non-B subtypes. AIDS. 2006;20:643-651.
14. Hosseinipour MC, van Oosterhout JJ, Weigel R, et al. The public health approach to identify antiretroviral therapy failure: high-level nucleoside reverse transcriptase inhibitor resistance among Malawians failing first-line antiretroviral therapy. AIDS. 2009;23:1127-1134.
15. Shepherd JCEM, Griffifth D, Charurat MEO, et al. High levels of class-wide NRTI resistance among HIV-positive patients failing first-line antiretroviral regimens in Nigeria. Presented at: HIV Implementers' Meeting; 2008; Kampala, Uganda. Abstract 796.
16. International Treatment Preparedness Coalition. Missing the Target #4: Time is Running Out to End AIDS - Treatment and Prevention for All!, Missing the Target Series, ITPC, 18 July 2007, 119 pp. Available at: Accessed 30 September 2009.
17. Boulle A, Orrel C, Kaplan R, et al. Substitutions due to antiretroviral toxicity or contraindication in the first 3 years of antiretroviral therapy in a large South African cohort. Antivir Ther. 2007;12:753-760.
18. Johnson VA, Brun-Vezinet F, Clotet B, et al. Update of the Drug Resistance Mutations in HIV-1: Spring 2008. Top HIV Med. 2008;16:62-68.
19. Marconi VC, Sunpath H, Lu Z, 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.
20. Orrell C, Walensky RP, Losina E, et al. HIV-1 clade C resistance genotypes after first virological failure in a large community ART programme. Antivir Ther. 2009;14:523-531.
21. Virological response to a triple nucleoside/nucleotide analogue regimen over 48 weeks in HIV-1-infected adults in Africa. AIDS. 2006;20:1391-1399.
22. Yap SH, Sheen CW, Fahey J, et al. N348I in the connection domain of HIV-1 reverse transcriptase confers zidovudine and nevirapine resistance. PLoS Med. 2007;4:e335.
23. Parikh UM, Koontz DL, Chu CK, et al. In vitro activity of structurally diverse nucleoside analogs against human immunodeficiency virus type 1 with the K65R mutation in reverse transcriptase. Antimicrob Agents Chemother. 2005;49:1139-1144.
24. Novitsky V, Wester CW, DeGruttola V, et al. The reverse transcriptase 67N 70R 215Y genotype is the predominant TAM pathway associated with virologic failure among HIV type 1C-infected adults treated with ZDV/ddI-containing HAART in southern Africa. AIDS Res Hum Retroviruses. 2007;23:868-878.
25. Kuritzkes DR, Bassett RL, Hazelwood JD, et al. Rate of thymidine analogue resistance mutation accumulation with zidovudine- or stavudine-based regimens. J Acquir Immune Defic Syndr. 2004;36:600-603.
26. Parikh UM, Bacheler L, Koontz D, et al. The K65R mutation in human immunodeficiency virus type 1 reverse transcriptase exhibits bidirectional phenotypic antagonism with thymidine analog mutations. J Virol. 2006;80:4971-4977.
27. Parikh UM, Barnas DC, Faruki H, et al. Antagonism between the HIV-1 reverse-transcriptase mutation K65R and thymidine-analogue mutations at the genomic level. J Infect Dis. 2006;194:651-660.
28. Doualla-Bell F, Avalos A, Brenner B, et al. High prevalence of the K65R mutation in human immunodeficiency virus type 1 subtype C isolates from infected patients in Botswana treated with didanosine-based regimens. Antimicrob Agents Chemother. 2006;50:4182-4185.
29. Lazzarin A, Campbell T, Clotet B, et al. Efficacy and safety of TMC125 (etravirine) in treatment-experienced HIV-1-infected patients in DUET-2: 24-week results from a randomised, double-blind, placebo-controlled trial. Lancet. 2007;370:39-48.
30. Madruga JV, Cahn P, Grinsztejn B, et al. Efficacy and safety of TMC125 (etravirine) in treatment-experienced HIV-1-infected patients in DUET-1: 24-week results from a randomised, double-blind, placebo-controlled trial. Lancet. 2007;370:29-38.

HIV-1 drug resistance; subtype C; first-line failure; South Africa

© 2010 Lippincott Williams & Wilkins, Inc.