Swiss study finds moderate drug resistance; highlights difficulty of interpreting trends in changing treatment populations

A study published in the April ^st edition of Clinical Infectious Diseases reports moderate drug resistance rates in a large Swiss cohort, while also cautioning that resistance data may be misleading if the treatment history of the population of interest is not taken into account. The cohort’s resistance level in 2007 was estimated to be 37% to 45%, much lower than that reported in some other studies.

The difference may be attributable in part to the study design, which distinguished between cohort members who were thought to be at higher and lower risk for resistance based on their treatment histories and viral load levels. Researchers assessed the prevalence of resistance among high-risk and low-risk cohort members who had undergone resistance testing, then factored the different resistance rates into their estimates for the full cohort.

HIV mutates as it reproduces in the body, and resistance occurs when the mutations render the virus immune to the antiretroviral drugs. Different strains of HIV emerge to overcome different types of drugs and, when resistant virus is transmitted, the newly infected person acquires the resistance as well.

The highly effective antiretroviral combinations introduced in the late 1990s are better at preventing resistance than earlier regimens were. Nonetheless, resistance remains a major ongoing challenge in the management of HIV infection, and is associated with higher rates of treatment failure and death. Missed doses of drugs, poor drug absorption, and drug interactions can all facilitate resistance.

The Swiss HIV Cohort Study was the source of data for the recently reported findings. Researchers analysed data from 8064 antiretroviral-experienced men and women who had been examined at least once between 1999 and 2007. Cohort members were divided into three subgroups: those thought to be at high risk for resistance mutations because they had either taken early, weak antiretroviral regimens or had experienced treatment failure; those thought to be at low risk because they were not undergoing treatment or had maintained viral suppression while on treatment; and those whose status was unknown. Most people in the high-risk group had initially been treated with one or two nucleoside reverse transcriptase inhibitors (NRTIs).

About 40% of the study population had undergone HIV genotypic resistance testing. Tests were categorised according to whether people had been in the high-risk subgroup, low-risk subgroup, or unknown subgroup when the sampling occurred. Researchers calculated the proportion of tests in each subgroup that indicated the presence of at least one major resistance mutation. Resistance to fusion inhibitors was disregarded, as was resistance to some etravirine (Intelence) mutations that may take multiple forms.

The known resistance levels provided the basis for estimates of upper and lower bounds of resistance for the full cohort. For the upper-bound estimate, resistance levels in tested members of each subgroup were taken to be representative of the entire subgroup. The estimate was derived by averaging the three subgroup-specific rates of resistance, with weighting in accordance with subgroup sizes. The lower-bound estimate reflected the proportion of cohort members who either belonged to the high-risk group or belonged to other groups and had test results indicating resistance.

The cohort size steadily increased between 1999 and 2007, while the high-risk population decreased as a result of deaths and losses to follow-up. Therefore a growing proportion of cohort members were likely to have initiated antiretroviral therapy with highly effective regimens, which would greatly reduce their risk of developing drug resistance.

This shift in the composition of the study population appeared to influence long-term resistance trends. In 1999, between 50 and 57% of people were estimated to have at least one resistance mutation, while in 2007, the proportion had dropped to between 37 and 45%. Since resistance mutations generally persist, the change may seem paradoxical. But the researchers believe the most likely explanation to be the simultaneous growth of the low-risk subgroup and shrinking of the high-risk subgroup over time.

To distinguish between changes in the study population and changes in resistance levels, the researchers performed the same set of analyses on a cohort of individuals seen every year from 2002 through to 2007. In other words, this 'closed' cohort had no deaths or losses to follow-up, and all cases of resistance were carried forward. Resistance levels were estimated at 45 to 52% in 2002, and 49 to 55% in 2007. Importantly, increases in prevalence became smaller over the years.

Resistance change rates, calculated using lower-bound estimates for the open and closed cohort, were stratified according to the potency of cohort members’ first antiretroviral regimens. In the open cohort, resistance increased by 1.5% per year among people who had initiated therapy with a highly effective regimen that included an unboosted protease inhibitor (PI), and by 0.6% per year among those who had started on newer regimens (combinations including either a non-nucleoside reverse transcriptase inhibitor (NNRTI) or a ritonavir-boosted PI). In the closed cohort, resistance increased by 1.2% per year for people who had started on unboosted PIs and by 1.6% per year for people who had started on newer regimens.

The researchers report that other studies based on analyses of antiretroviral resistance databases have found resistance rates of 76 to 90%. “However, patients who had undergone drug resistance tests are not representative of the entire [antiretroviral]-exposed population,” they note. “Studies of such patients tend to be biased toward patients who had experienced problems while undergoing therapy.”

The Swiss HIV Cohort Study appears to be very representative of diagnosed HIV-positive people in Switzerland, and the researchers suggest that findings in this study population may also apply to HIV-positive people in other countries with highly developed health care systems.

The researchers speculate that the prevalence of antiretroviral resistance will continue to decrease at the population level in developed countries “because of the observed dilution effects, but also, and more importantly, because durable therapy success can more often be achieved with newer drugs and thus the emergence of new drug resistance is considerably slowed.”