HIV Variants of Low Frequency May Have a Big Impact

June 2, 2007

Much has been written over the years about the prevalence of HIV drug resistance mutations and about the fitness and transmissibility of resistant strains of virus and their present and potential impact on the HIV pandemic.

Much has been written over the years about the prevalence of HIV drug resistance mutations and about the fitness and transmissibility of resistant strains of virus and their present and potential impact on the HIV pandemic. The prevalence of extended resistance to 3 antiretroviral drug classes ranges from 4% to 15% in resource-rich nations, and such a high degree of resistance is linked to decreased survival.1,2

Reports from the recent Conference on Retroviruses and Opportunistic Infections added to this concern. The reports involved examination of antiretroviral-exposed patients who had wild-type HIV-1 at baseline by standard bulk sequencing techniques but showed less than adequate responses to certain drugs compared with their drug-naive counterparts.3,4 More sensitive assays, including single genome sequencing and allele-specific polymerase chain reaction (PCR), permitted improved de­tection of drug-resistant HIV in these patients and could facilitate the design of more effective HIV treatment strategies.

The development of point-mutation assays based on PCR has enabled identification of HIV-1 variants persisting below the 20% frequency threshold of sequence detection by standard genotyping. These techniques selectively amplify mutant sequences in the context of the total number of viruses present, without the need to determine input virus copy number.5

In an initial proof-of-principle study, a real-time PCR assay was used to assess the K65R reverse transcriptase mutation, a key marker for tenofovir-selected resistance, in simian immunodeficiency virus (SIV)mac-infected monkeys.5 K65R alleles could be identified at frequencies 40-fold lower than those detectable by direct sequence analysis, with a mean detection limit for mutant virus of 0.4%, range 0.1% to 2%, in a wild-type background.

Real-time PCR assays with detection cutoffs as low as 0.3% were subsequently used to assess resistance mutations in HIV-positive drug-naive persons.4 Variants were found at a rate double the estimated prevalence of transmitted resistance. In transmitted mutant virus populations, use of real-time PCR identified 30% more isolates with more than 2–drug class resistance.4

A retrospective analysis of patients participating in either an antiretroviral therapy trial or in a study of single-dose nevirapine (NVP) given during delivery of a second child to women previously exposed or unexposed to NVP, was also conducted. In the antiretroviral therapy study, explanations were sought as to why most patients experienced a mean 2 log10 copies/mL decrease in HIV RNA level while some had less than a 1 log10 copies/mL decrease. Detection of low-frequency mutations correlated with virological failure.3,4

Real-time PCR identified drug resistance mutations in 29 (14%) of 211 treatment-naive persons. Four persons had 2–drug class resistance mutations, and an additional 2% to 9% of each of 6 identified mutations (L90M in protease and M41L, K70R, K103N, Y181C, and M184V in reverse transcriptase) were found in samples known to have other resistance mutations by standard techniques.4 This resulted in a cumulative 30% increase in the 6 mutations tested and a 40% increase in identification of greater than 2–drug class resistance.

In the maternal-fetal transmission study, 43% of women who were thought to have wild-type virus by population sequencing had an NVP resistance mutation, K103N, by real-time PCR. These variants were detectable for more than 36 weeks, compared with a mean of only 13 weeks when they were detected by standard genotyping.3

It had been previously shown that women who had received single-dose NVP experienced higher rates of virological failure when subsequent NVP-based antiretroviral therapy was instituted within 6 months than did with women without previous exposure to the drug (18.4% vs 5.0%; P = .002).6 It was then discovered that in single-dose NVP–treated women, the incidence of HIV transmission to their second child following another course of single-dose NVP was 2.6-fold higher (11.1% vs 4.2%; P = .02).3 Only 3 (15%) of 20 HIV-transmitting women had NVP mutations detectable by bulk sequencing, while more sensitive assays detected relevant mutations in 60%.

Of 259 participants in a trial of efavirenz/lamivudine plus abacavir or zidovudine, real-time PCR detected 16 patients with resistance mutations at baseline, compared with only 7 identified by bulk sequencing.4 The association between resistance mutations and virological failure increased significantly when these low-frequency variants were included. In one model, those with minority resistance mutations had 8.1 times the odds of treatment failure of those without such mutations, independent of HIV RNA level or CD4 counts.4

Apart from such trials, plasma collected between 2003 and 2005 from antiretroviral-naive persons from Los Angeles and Chicago in whom HIV infection was recently diagnosed was evaluated for resistance mutations.3 Sixty-five percent of the participants were men who have sex with men; 20%, injection drug users; and 68%, minorities. Standard genotyping revealed mutations in at least 1 class of drug in 20%. But of 205 samples classified as bearing only wild-type virus, 15% had at least 1 major resistance mutation and 4.5% had thymidine analogue mutations, particularly M41L, by real-time PCR.

It is unclear how these low-frequency variants are selected for in treatment-naive patients. It is also unclear what the threshold for establishment and persistence of such minor variants is among various HIV-positive populations. A standard for testing needs to set clear cutoff points: limits for clinically important drug variants above that which might occur normally and have no bearing on drug sensitivity in vivo.

Four steps are now required to determine whether these more sensitive assays should become the standard of care: expanded evaluation with treatment-naive and treatment-experienced persons in real-world settings; establishment of testing standards; validation of selected assays on divergent HIV-1 subclasses; and assessment of commercial feasibility.


References1. Zaccarelli M, Tozzi V, Lorenzini P, et al; Collaborative Group for Clinical Use of HIV Genotype Resistance Test (GRT) at National Institute for Infectious Diseases Lazzaro Spallanzari. Multiple drug class-wide resistance associated with poorer survival after treatment failure in a cohort of HIV-infected patients. AIDS. 2005;19:1081-1089.
2. Laurence J. Multidrug-resistant HIV. AIDS Reader. 2005;15:266-267.
3. Johnson J. Clinical implications of low-frequency HIV-1 variants. 14th Conference on Retroviruses and Opportunistic Infections; February 25-28, 2007; Los Angeles. Abstract 61.
4. Johnson J, Li JF, Wei X, et al. Low-frequency mutations substantially increase the prevalence of transmitted drug resistance and greatly strengthen the relationship between resistance mutations and virologic failure. 14th Conference on Retroviruses and Opportunistic Infections; February 25-28, 2007; Los Angeles. Abstract 639.
5. Johnson JA, Rompay KK, Delwart E, Heneine W. A rapid and sensitive real-time PCR assay for the K65R drug resistance mutation in SIV reverse transcriptase. AIDS Res Hum Retroviruses. 2006;22:912-916.
6. Lockman S, Shapiro RL, Smeaton LM, et al. Response to antiretroviral therapy after a single, peripartum dose of nevirapine. N Engl J Med. 2007;356:135-147.