Highlights of HIV drug resistance studies presented at the 3rd IAS Conference

Ian Frank MD for

Issues concerning HIV drug resistance are a growing concern in the treatment and management of HIV infection. In the review that follows, Dr. Ian Frank summarizes selected studies on resistance presented at the 3rd IAS meeting in Rio de Janeiro, Brazil.

Unless otherwise stated, all references in this review article are to the Program and Abstracts of the 3rd International AIDS Society Conference on HIV Pathogenesis and Treatment. July 24-27, 2005. Rio de Janeiro, Brazil.

Epidemiology: contacts of the New York City “superinfection” case identified

Early in 2005 a New York City man was identified as an individual with apparent recent HIV infection who had multiple class resistant virus, dual-tropic virus, and experienced rapid disease progression. [1] Two possible contacts, who were sexual partners of this NYC case, were identified independently by Quest Diagnostics and LabCorp after searching their genotyping databases for matches to the unusual mutational pattern of the virus from the NYC case, and traced to an HIV practice in Connecticut. [2]

These two contacts had a history of unprotected insertive intercourse among themselves (phylogenetic analyses of isolates suggests that one individual may have acquired the particular resistance profile of the other by superinfection), and admitted to unprotected intercourse with the New York City case at a sex club.

Although neither of these individuals experienced rapid disease progression and both had R5, rather than mixtures of R5 and X4 phenotypes or dual tropic virus, the epidemiologic evidence for transmission of this uniquely resistant virus is strong.

The authors take-home messages were that: the community must continue to be educated that safer sex practices are important among infected individuals; that the HIV uninfected community should not misconstrue the benefits of today’s antiretroviral therapy to believe that becoming HIV infected means just taking a few pills a day, and that your genotype may be able to identify you as easily as your social security number.

Infection with resistant virus is not associated with more rapid disease progression

Motivated by the New York City case, a group of European investigators evaluated whether infection with resistant virus was associated with more rapid disease progression. [3]

A case control study was performed using a prospective, multicenter cohort of 1415 individuals diagnosed in 2003. Seventy-eight individuals were identified at presentation with at least one primary resistance associated mutation, based upon the IAS-USA resistance algorithm. These cases were compared to 77 randomly selected controls matched by baseline viral load and CD4+ count with sensitive virus.

Median baseline viral loads and CD4+ counts were 4.8 log copies/mL and 359 cells/mm3 in the cases (resistant virus) and 4.7 log copies/mL and 365 cells/mm3 in the controls. Subjects were followed for a median of 16 months for the development of one of three endpoints; fall in CD4+ count to <200 cells/mm3, new AIDS-defining clinical illness, or initiation of therapy.

During follow-up, 17 subjects with resistant virus started therapy, two experienced CD4+ cell count declines to <200 and one developed an AIDS-defining event, compared to 28 subjects with sensitive virus who started therapy, one had a CD4+ count decline to <200, and two developed an AIDS-defining event. There are no statistically significant differences in immunologic or clinical disease progression between the two groups.

These data suggest that individuals who get infected with resistant virus have similar disease progression as those who get infected with sensitive virus.

New mechanisms for NRTI resistance: mutations in the RNase H encoding region

As reverse transcriptase (RT) constructs a DNA strand, the RNase portion of the enzyme is simultaneously degrading the RNA template. New data suggests that mutations in the portion of the RT genome that encodes the RNase H domain influence susceptibility to the thymidine analogs zidovudine (ZDV; Retrovir) and stavudine (d4T; Zerit). [4]

To date the number of RNase H sequences available in the GenBank database are relatively few, and the commercially available genotyping resistance assays do not amplify the RNase H domain (amino acids 441 – 560). For this reason, the influence of RNase H on susceptibility to NRTIs has not been carefully evaluated.

In vitro, site directed mutants in RNase H increased the IC50 value of wild type virus to ZDV and d4T by 10- to 100-fold. These mutations conferred synergistic levels of resistance when combined with thymidine analog mutations (TAMs).

To evaluate whether mutations in RNase H contribute to RT resistance in patients, RNase H sequences from NRTI-experienced and naive subjects were cloned and sequenced, and then inserted into a wild RT genome, replacing the WT RNase, which could be used to test the phenotypic susceptibility of recombinant isolates.

Among patients with RT sequences that were wild type for RT mutations in the pol gene associated with resistance, treatment experienced patients had RNase sequences that were associated with 2.4- to 5.7-fold resistance to ZDV and 1.0- to 1.8-fold resistance to d4T, in some cases, clinically significant levels of resistant.

In contrast, the RNase sequences from treatment naive patients had no impact on phenotypic susceptibility, suggesting that treatment may have selected for virus with RNase sequences that impart some phenotypic resistance independent of that associated with mutations in other areas of the pol gene.

The level of phenotypic resistance was amplified in treatment-experienced patients with TAMs. The RNase H sequence from one subject with TAMs conferred 1839-fold resistance, compared to 11-fold resistance with a wild type RNase sequence.

The mechanism by which RNase H mutations cause RT resistance is not known, but is hypothesised to occur by delaying RT processivity, thereby increasing the time that excision of the terminal nucleotide could occur. More studies need to be done to investigate whether RNase mutations may be associated with resistance to non-thymidine analogs and what mutations in RNase H may be clinically significant, to ultimately determine whether genotyping and phenotyping assays need to be designed to include an evaluation of RNase H.

NRTI resistance: absence of K65R in subjects failing tenofovir in GS934

GS934 was an open-label comparison of tenofovir (Viread), emtricitabine (Emtriva), and efavirenz (Sustiva) versus fixed dose zidovudine/lamivudine and efavirenz in antiretroviral naive subjects. Subjects who received tenofovir + emtricitabine had better virologic outcomes in the intent to treat analysis. In Rio, data on the mutational analysis in subjects with virologic failure were presented and are summarised in the accompanying table. [5]

Table 1: Resistance in GS934*

TDF+FTC+EFV (n=244) ZDV/3TC+EFV (n=243)
Resistance analysis 12 (5%) 23 (10%)
Resistant virus n
Any mutations 9 17
EFV-R 9 16
M184V/I 2 7
Any TAM 0 1
K65R 0 0
WT or as baseline 3 5

*Excluding subjects with resistant virus at baseline

Mutations were not identified in every subject, some of whom rebounded with wild type virus. As expected, the majority of subjects with resistant virus had NNRTI resistance. A greater proportion of subjects developed I84V/I on 3TC than FTC, though those numbers are not statistically significant. In contrast to GS903, a comparison of TDF, 3TC (Epivir), and EFV versus d4T, 3TC, and EFV, in which 8 subjects of 47 subjects (17%) with virologic failure in the TDF arm had virus with a K65R mutation (one subject in the d4T arm developed K65R) [6] no subjects in GS934 failing on tenofovir developed K65R. Whether the long half-life of FTC can “protect” against the selection of K65R will be determined with data from larger studies currently underway.

PI resistance: establishing the clinical cutoffs for atazanavir and tipranavir

Rick Pesano from ViroLogic provided data describing the establishment of the clinical cutoffs for the protease inhibitors (PI) atazanavir (Reyataz) and tipranavir (Aptivus) relevant to their Phenosense assay. [7]

The clinical cutoff for unboosted atazanavir was derived from specimens available from the BMS 043 trial, a comparison of unboosted atazanavir versus fixed dose lopinavir/ritonavir in patients with previous virologic failure on a PI-containing regimen. The clinical cutoff for boosted atazanavir was derived from specimens available from BMS 045, a comparison of boosted atazanavir versus fixed-dose lopinavir/ritonavir.

The clinical cutoff for unboosted atazanavir is 2.2 (corresponding to approximately 3 mutations), and for boosted atazanavir, it is 5.2 (corresponding to approximately 5 mutations). The clinical cutoff for boosted tipranavir is 4.0, corresponding to approximately five mutations.



  1. Markowitz et al. Infection with multidrug resistant, dual-tropic HIV-1 and rapid progression to AIDS: a case report. Lancet 2005;365:1031-38.
  2. Blick G et al. “Patient Zero”: The Connecticut source of the multi-drug resistant, dual-tropic, rapidly progressing HIV-1 strain found in NYC. Abstract MoOa0101.
  3. Wensing AMJ et al. Comparative disease progression observed in newly diagnosed patients infected with drug resistant and susceptible HI-1: no signs for increased virulence. Abstract WeOaLB0101.
  4. Pathak et al. RNase H domains from treatment-experienced patients increase AZT resistance. MoFo0303.
  5. Pozniak et al. Superior outcome for tenofovir DF (TDF), emtricitabine (FTC) and efavirenz (EFV) compared to fixed dose zidovudine/lamivudine (CBV) and EFV in antiretroviral naive patients. Abstract WeOa0202.
  6. Gallant et al. Efficacy and safety of tenofovir DF vs stavudine in combination therapy in antiretroviral-naive patients. JAMA 292; 2004; 191-201.
  7. Pesano R. Mapping drug resistant phenotypes of HIV. Abstract MoFo0301.

Links to other websites are current at date of posting but not maintained.