Resistance already emerging in first months of treatment despite viral load decline

The emergence of drug resistance is already detectable in the first months of HIV treatment even if viral load is falling, and is especially pronounced in people who experience slower viral load declines, a small, intensive German study has found.

As selection of drug-resistant virus correlated significantly with the length of time until viral load became undetectable, the researchers recommend therapy intensification during the early treatment phase to be considered as a strategy to control replication.

In this study, published in the March 30th edition of AIDS, 15 mainly chronically infected patients were followed just prior to and during the early months of antiretroviral therapy.

Plasma sampling was performed before and during treatment, as was HIV-1 quantification and genotypic analysis. Three key resistance mutations were tracked, the L90M (protease) and K103N and M184V (reverse transcriptase), using a PCR assay designed to detect the three specific mutations. The specificity of the assay allowed the detection of mutant virus populations if they comprised no more than 0.01-0.2% of the total virus population, compared to a normal resistance assay threshold of around 10-15%.

The drug regimens varied, but all contained 3TC (lamivudine, Epivir) combined with another nucleoside reverse transcriptase inhibitor (NRTI) plus either a protease inhibitor (PI) or a nonnucleoside reverse transcriptase inhibitor (NNRTI).

Fourteen of the patients were chronically infected and eleven were first diagnosed with HIV showing AIDS-defining symptoms. Two of the patients were women, the mean age was 39 (median 32), CD4 cells ranged from 6 to 912 cells/mm³, and viral loads ranged from 7,500 to 360,000 HIV-1 RNA copies/mL.

Four patients had one or two drug-resistant mutations at baseline and selection of drug-resistant variants occurred following treatment. Two other patients did not have detectable drug-resistant mutations at baseline, but even as viral load was decreasing, there was a substantial and simultaneous quantitative increment of drug-resistant viruses that developed soon after ARV initiation. One patient was lost to follow-up after week seven of treatment and it took up to twelve months for the other six patients to reach undetectability. Slower viral load decline in some patients was related to poor compliance to therapy.

It is not entirely clear if slower viral load decline led to a higher risk of selecting drug-resistant viruses or if it was the presence of drug-resistant viruses that extended the period needed to reach undetectable viral load. The authors suspect that it is the former, but compliance may also have a part in this puzzle.

Six patients who did not have detectable drug-resistant HIV-1 at baseline, along with three who did, showed no selection of drug-resistant variants (as measured by the three key resistance mutations) and viral load was undetectable within three to four months. It is interesting that the three patients with drug-resistant viruses at baseline did not have an apparent increased risk for the development of drug-resistant mutations.

This study is one of the first to show that drug-resistant viruses in some patients are selected within the first months of treatment and that, despite viral load reduction, emergence of drug resistance can occur through residual replication of HIV-1 variants. Because of the risk of incomplete suppression, the authors suggest that intensifying antiretroviral treatment until replication is controlled and viral load is undetectable would be a reasonable treatment strategy. At that point, treatment could be scaled back to avoid toxicities and reduce side effects.

Further investigation of the long-term impact of early selection of drug-resistant viruses and strategies to avoid that occurrence are recommended.