Prevalence of antiretroviral resistance in a South London cohort of treatment-naive HIV-1-infected patients

Abstract

Data from North America, Europe and Australia suggest that the transmission of drug-resistant HIV is an emerging public health problem [1]. Conservative estimates indicate that the prevalence of resistance to at least one antiretroviral drug may be as high as 14% in treatment-naive patients with early infection (< 12 months’ duration) and 12% in those with established infection [1]. The transmission of multidrug-resistant HIV has also been reported at a prevalence of 1–4%[1]. As baseline resistance profiles predict responses to antiretroviral therapy [2], the consensus is emerging that resistance testing should be considered for all patients before the initiation of therapy [3,4]. However, variations in prevalence rates between different geographical areas and risk groups are to be expected, and epidemiological surveys are needed to inform local and national guidelines for testing. Data from the UK remain limited [5]. The benefit of testing patients with established infection also needs to be defined, as resistant mutants may revert to wild type soon after transmission in the absence of drug pressure. Our aim was to determine the prevalence of antiretroviral resistance in treatment-naive patients with defined seroconversion dates between January 1996 and October 2000 attending two HIV clinics in south London. These centres care for a heterogeneous population, including a large proportion of patients from sub-saharan Africa and the Caribbean basin. Of the 60 eligible patients, 14 had acute seroconversion illness, 25 had early infection (< 12 months interval between first positive and last negative antibody test), and 21 had established infection (> 12 months interval between first positive and last negative antibody test). Forty-four (73%) were Caucasian, 10 (17%) afro-Caribbean, and six (10%) sub-saharan Africans. Risk factors for infection were male homosexual contact (45, 75%), heterosexual contact (14, 23%), and intravenous drug use (one). Genotypic resistance was determined by direct sequencing of RNA from stored plasma samples (TruGene assay, Visible Genetics, Evry, France). Viral subtypes were determined from consensus reverse transcriptase (RT) and protease sequences (NCBI HIV-1 Genotyping Tool). Fifty-one (85%) patients were infected with B and nine (15%) with non-B subtype virus (Table 1). Three patients (subtype B) had key mutations conferring resistance to nucleoside RT inhibitors. Two patients had M41L with either T215D or T215C. M41L arises during treatment with zidovudine and to a lesser extent stavudine, and confers high-level resistance to zidovudine when present with other mutations, including T215F/Y. T215D/C are unusual mutations, which may represent incomplete reversion from zidovudine-resistant to wild-type virus [6]. T215C has also been associated with resistance to zalcitabine. The two patients had early and established infection, respectively, suggesting persistence and good replication competence of the transmitted mutants. The third patient had the M184V mutation, which confers high-level resistance to lamivudine, and may also affect susceptibility to didanosine, zalcitabine and abacavir.Table 1: Antiretroviral resistance mutations in 60 treatment–naive HIV–infected patients. Four patients had mutations at RT codons involved in resistance to the non-nucleoside RT inhibitors. However, only one (subtype G) had a key mutation, K103T, which confers resistance to delavirdine [7], and may also have resistance effects for nevirapine and efavirenz. Other mutations were K103R (subtype B), K103K/R/I (subtype D), and A98S (subtype C), which are likely to represent natural polymorphisms. The key protease mutation M46I was detected in one patient (subtype G). In combination with other mutations, M46I reduces susceptibility to indinavir and other protease inhibitors (PI). Forty-eight (80%) patients had accessory protease mutations (codons 10, 20, 36, 63, 71 and 77), which are likely to reflect natural polymorphisms. Although individually these mutations have limited effects on drug susceptibility, they may influence the rate at which PI-resistant viruses are selected during therapy [8]. Whereas most patients had only one or two mutations, four patients had three, with potential cumulative effects on PI resistance. The predominant mutations in B subtypes were those at codons 63 (22, 43%) and 77 (16, 31%), whereas M36I was found in all non-B subtypes. In conclusion, in south London there remains a relatively low prevalence (4/60, 7%) of key resistance mutations in treatment-naive patients with defined seroconversion dates between 1996 and 2000. The prevalence is lower than that reported in other studies of patients with similar risk factors [1,9], suggesting that as yet undefined demographic factors influence the likelihood of acquiring drug-resistant HIV. As patients with defined seroconversion dates are a highly selected subgroup of treatment-naive patients, a study of a south London cohort that is representative of all risk groups is ongoing. Consistent with previous findings [10], a high degree of polymorphism was detected in the protease gene of both B and non-B subtypes. The clinical significance of this observation and the implications on the choice of antiretroviral therapy remain to be established. Anna Maria Gerettiab Melvyn Smithb Natasha Osnerb Siobhan O'Sheac Ian Chrystiec Philippa Easterbrooka Mark Zuckermanb