There appears to be a wide variation in the transmissibility of drug-resistant HIV according to the specific drug-resistance mutation it carries, Spanish researchers report in the October 15th edition of Clinical Infectious Diseases.
Using data on the prevalence of resistance mutations in 89 recent seroconverters and 520 patients with detectable viral load after at least six months of antiretroviral therapy, the researchers estimated a fivefold difference in transmissibility between the most easily transmitted single mutation (the T215Y thymidine analogue mutation that conveys high-level resistance to AZT [zidovudine, Retrovir] and d4T [stavudine, Zerit]) and the least easily transmitted (the K103N mutation that conveys cross-resistance to all three currently-used non-nucleosides (NNRTIs).
The Spanish study had limitations; it compared two different populations with drug resistance, but did not study actual transmission events. For this reason it could not estimate the absolute frequency of transmission of specific HIV resistance mutations, only how easily they were transmitted relative to each other.
The sample size was too small to estimate the transmissibility of serious but less frequently transmitted mutations such as the 69SSS mutation that conveys cross-resistance to all nucleosides (NRTIs).
However it does provide an additional explanation for why NRTI resistance appears more frequently than NNRTI or protease inhibitor (PI) resistance in recent seroconverters
Dr Carmen de Mendoza and colleagues compared two populations of patients with HIV who attended two clinics in Madrid between January 1997 and June 2003.
The first was a group of 89 people whose date of seroconversion could be established to within a year.
The second was a group of 520 patients who were virologically failing HIV therapy, that is who had detectable HIV viral loads despite being on highly active antiretroviral therapy (HAART), and who could therefore transmit drug resistant virus. The two groups were dubbed ‘seroconverters’ and ‘potential transmitters’. The patients were all resident in the Madrid area.
These two groups of patients were ‘paired off’ against each other in the following way: the six patients failing therapy who were ‘nearest’ to one of the seroconverters in terms of time of virologic failure and transmission route (in this case heterosexual or gay sex) were paired with one of the seroconverters.
The researchers then compared the relative frequency of specific drug resistance mutations and mutations to each drug class between the two groups as a whole and within the ‘pairs’ created.
They then compared the ratio of the frequency of each mutation in seroconverters to the frequency in potential transmitters.
This produced a rough estimate of the relative frequency of transmission of each mutation within the local population.
On average, resistance testing was carried out 8.7 months after seroconversion in the recent seroconverter group.
The frequency of HIV resistance among the ‘potential transmitters’ was 80%, in other words 20% were failing without drug resistance mutations, presumably due to poor adherence. Seventy-four percent had resistance to NRTIs, 36% to NNRTIs and 39% to PIs.
The prevalence of NRTI resistance rose from 66% of subjects in 1997-98 to 78% in 2000-03, and of NNRTI resistance from 15% to 47%, which reflects the adoption of NNRTIs in Spain.
The authors caution that the genotyping method used may not have been able to detect resistance mutations that were present in less than 30% of the virus sequenced from a plasma sample.
The most frequent individual mutations in the ‘potential transmitters’ were the T215Y/F thymidine analogue mutation that conveys high-level resistance to AZT and d4T, and some resistance to abacavir, tenofovir and ddI; the M184V 3TC (lamivudine, Epivir) / FTC (emtricitabine, Emtriva) mutation; and the M41L thymidine analogue mutation. These three were each carried by 35-45% of the group.
The most common protease inhibitor mutation was the L90M nelfinavir (Viracept) / saquinavir (Invirase / Fortovase) mutation, with the V82A indinavir (Crixivan) / ritonavir (Norvir) mutation and M46L, which conveys high-level resistance to every PI except saquinavir (and tipranavir) nearly as common.
The two major NNRTI mutations (K103N and Y181C) were about as common as each other, present in 27% and 16% of the possible transmitter group.
In the seroconverter group, only 15 of 89 patients (17%) had drug-resistant virus.
All but two of these (15%) had NRTI resistance, four (3%) had NNRTI resistance and four had PI resistance.
The most common NRTI mutations were different variations of mutations at position 215 of the reverse transcriptase gene. Eight patients had these. However five were classed as ‘revertants’, which means there was evidence that they had acquired NRTI-resistant HIV, but that it was reverting to a variety which though different from ‘wild-type’ virus, does not actually convey significant NRTI resistance. This suggests that the T215Y NRTI resistance mutation usually transmitted is poor at replicating and will eventually tend to revert to wild type in the absence of drug pressure.
Six patients had the M41L thymidine analogue mutation and two the M184V 3TC mutation.
The numbers of patients with NNRTI and PI mutations were so low that their relative frequency was not significant. However the PI mutations transmitted were the three most frequent in the ‘potential transmitter’ group.
One person had PI resistance alone; one had NNRTI resistance alone; one had NRTI and NNRTI resistance; and two had NRTI and PI resistance. No-one acquired three-class resistant virus.
In terms of the transmissibility of different mutations, the 215 and M41L mutations appeared to be the most transmissible, appearing about 20% as often in the seroconverters as in the potential transmitters.
In contrast the M184V 3TC and K103N NNRTI mutations only appeared 4-6% as often in seroconverters as potential transmitters. The Y181C NNRTI mutation’s transmissibility was intermediate at about 14%. The most transmissible PI mutation appeared to be M46L, comparable with Y181C, and the other two PI mutations were about as transmissible as M184V.
There was thus a nearly fivefold difference between the transmissibility of the T215Y mutation and the K103N mutation.
The investigators also measured the viral loads of patients with specific mutations compared with those without, since a mutation that produced a lower viral load would be less transmissible. The only significant association was M184V, which produced about a 0.25 log10 (nearly twofold) reduction in viral load.
De Mendoza C et al. Evidence for differences in the sexual transmission of HIV strains with distinct drug resistance genotypes. Clin Infect Dis 39 (October 15 edition, now available online), 2004.