HIV Articles  
Rapid accumulation of nonnucleoside reverse transcriptase inhibitor-associated resistance: evidence of transmitted resistance in rural South Africa
18 October 2008
Barth, Roos E; Wensing, Annemarie M; Tempelman, Hugo A; Moraba, Robert; Schuurman, Rob; Hoepelman, Andy I
aDepartment of Internal Medicine and Infectious Diseases, University Medical Centre Utrecht, The Netherlands
bDepartment of Virology, University Medical Centre Utrecht, Utrecht, The Netherlands
cNDLOVU Medical Centre, Elandsdoorn, Mpumalanga, South Africa.
In a large cohort in rural South Africa, 73% of subtype-C-infected patients initiating highly active antiretroviral therapy achieved viral suppression. In patients with subsequent virological failure, an unexpected, rapid accumulation of nonnucleoside reverse transcriptase inhibitor-associated mutations was observed, whereas no thymidine analogue-associated mutations emerged. It appeared that several patients had drug-associated mutations prior to starting antiretrovirals, suggesting that transmission of resistance may have contributed to the accumulation of nonnucleoside reverse transcriptase inhibitor-mutations. Importantly, monitoring of HIV-RNA and prompt switch of treatment may prevent development of thymidine analogue-associated mutations
Availability of highly active antiretroviral therapy (HAART) in developing countries has increased considerably over the past years. In general, the WHO's recommended first-line therapy consisting of two nucleoside/nucleotide reverse transcriptase inhibitors (NRTIs) and a nonnucleoside reverse transcriptase inhibitor (NNRTI) is adopted by these countries.
Extensive experience with NNRTI-based HAART regimens in subtype-B has shown that good clinical and virological results can be achieved [1]. However, subtype-C predominates the worldwide HIV epidemic and treatment-response data regarding this subtype are still limited, in particular regarding the emergence of antiretroviral drug resistance [2].
In a rural, South African clinic in Limpopo (, 2353 HIV-positive patients have been in care since 2003. Over 1400 patients have started HAART. Cohort characteristics have been described in more detail previously [3]. We evaluated longitudinal, virological data of the first 313 patients (22%) receiving a NNRTI and two NRTIs (lamivudine and a thymidine analogue) as first-line HAART. Patients treated for prevention of mother-to-child-transmission (PMTCT) were excluded.
Viral suppression (HIV-RNA level <50 copies/ml) was achieved by 73% (230/313) of patients during the first year of therapy. Longitudinal blood testing showed subsequent virological failure (HIV-RNA >1000 copies/ml) after initial suppression in 13% (31/230) of these patients. Blood samples at time of failure were available for 74% (23/31) of patients. Genotyping to determine HIV drug resistance was successful in 21 of 23 patients. All viruses were of HIV-1 subtype-C origin and were evaluated for the presence of reverse-transcriptase-resistance mutations. Drug-resistance mutations were classified according to the international AIDS society guidelines of 2007, supplemented with currently known NNRTI-associated mutations listed in a recent literature study [4,5]. Two patients did not show any drug-resistance mutations, making nonadherence a possible cause for virological failure.
The observed drug-associated mutations are summarized in Table 1. Interestingly, a high frequency of NNRTI-associated mutations was detected in the majority of individuals; 58% showed at least two and 26% at least three such mutations. One patient even harbored five NNRTI-resistance mutations at the time of failure. All but one individual harbored at least one 'major' NNRTI mutation (K103N, V106M, Y181C or V190A). The accumulation of multiple NNRTI-associated mutations was observed in a significant proportion of individuals, in spite of relatively short treatment durations (range 24-52 weeks) and limited time spans between viral suppression and virological failure (maximum 12 weeks). The V106M mutation, conferring cross-resistance to all NNRTIs and usually associated with efavirenz treatment [6], was observed in 37% (7/19) of patients, 57% of which were treated with nevirapine.
The most prevalent NRTI-associated mutation was 184V, conferring resistance to lamivudine. In contrast, thymidine-analogue-associated mutations (TAMs) did not emerge at all. Previous studies on HIV-1 subtype-B-infected individuals treated with similar regimens have shown frequent selection of the 184V mutation, NNRTI-associated-resistance mutations and, less frequently, TAMs shortly after virological failure [7]. For HIV-1, subtype-C, similar resistance patterns were found in most studies, but the prevalence of TAMs was higher [8,9]. Accumulation of several NNRTI-mutations in the absence of any TAM, as observed here, has, to our knowledge, not yet been described for subtype-C-infected patients. The complete absence of TAMs may be explained by relatively short time intervals (maximum 12 weeks) between treatment failure and resistance testing. Furthermore, the rate of selection of TAMs in patients receiving lami vudine-containing HAART has been reported to be slow [10]. However, the short duration of treatment failure does not explain the observed accumulation of NNRTI-associated mutations. Therefore, we wondered whether some NNRTI-mutations were already present at baseline.
Blood samples prior to initiation of HAART were available for 14 of 19 (74%) patients showing drug-resistance mutations at time of failure. To our surprise, NNRTI-associated mutations were observed at baseline in four of 19 (21%) patients. This is in contrast to the low prevalence of baseline resistance previously reported in sub-Sahara Africa [11,12]. In the current study, one patient even showed three NNRTI mutations at baseline. Presence of NNRTI-associated mutations at baseline may have contributed to the rapid accumulation of NNRTI mutations during treatment.
Prevalence rates of transmitted antiretroviral drug resistance (TDR) vary between different regions [13]. Baseline-resistance testing is standard of care in many countries with evidence of TDR. To date, the reported prevalence of resistance mutations is low in countries where access to antiretroviral therapy is limited. However, due to the scale up of access to therapy in many African countries, TDR is on the rise [2,14].
Our data suggest that the prevalence of TDR may also increase in rural South Africa. Our dataset is small, though, and may not reflect a representative sample of the HIV-infected population. Therefore, additional, specifically designed studies are needed for TDR surveillance in this population.
We retraced four patients with baseline NNRTI resistance to find that one of them had received two NRTIs (stavudine and didanosine) by a general practitioner prior to entering our treatment program. However, none received NNRTIs previously.
In a Thai study [15], baseline-resistance mutations were described in seven treatment-naive HIV-infected patients, all of whom had sexual partners with prior treatment failure. Two of our four patients with baseline resistance were known to be related to an HIV-infected person; one had an HIV-positive husband and the other was a child from an HIV-infected mother. To our knowledge, these relatives did not receive any ART prior to our patients, not even in a PMTCT-program. The fourth patient developed cryptococcal meningitis and died 3 months after starting HAART. We could not find additional information explaining her baseline-resistance pattern.
We observed the emergence of a specific drug-resistance pattern in a cohort of HIV-1 subtype-C infected patients in rural South Africa. The accumulation of NNRTI-associated mutations without the emergence of any TAMs was remarkable.
Drug-associated mutations were present prior to starting HAART in some patients, suggesting that TDR could play a role when patients experience failure on a WHO-recommended, first-line HAART regimen in resource-poor settings. The accumulation of several NNRTI-associated mutations in our cohort clearly limits chances of treatment success with a second-generation NNRTI. However, our data demonstrate that accumulation of TAMs may be prevented if continuous viral replication is timely controlled. Therefore, regular monitoring of viral load and switching treatment soon after virological failure are needed to preserve treatment options in resource-limited settings.
  iconpaperstack view older Articles   Back to Top