Resistance mutations associated with AZT or d4T treatment accumulate at a relatively slow rate in people who remain on failing antiretroviral treatment for long periods, an analysis of the EuroSIDA cohort has shown.
The study, published in the September 1st edition of the Journal of Infectious Diseases, show that once initial resistance to AZT or d4T develops (thymidine analogue resistance), the acquisition of further resistance mutations tends to be a slow process, with an average of one thymidine analogue mutation developing during 4.3 years of follow-up.
The findings are likely to provide ammunition for both sides in the debate about laboratory monitoring in developing countries, following the results of the recently reported DART study, which was conducted in Zimbabwe and Uganda.
DART showed that clinical outcomes in people who received only clinical monitoring during antiretroviral treatment, without CD4 counts, were only slightly poorer over five years when compared to patients who received regular CD4 cell monitoring.
Critics of the DART approach say that the lack of laboratory monitoring leads to long periods of uncontrolled viral replication during which high-level resistance can develop. This puts patients at risk of a poor response to second-line treatment, and also places them at risk of immunological deterioration that can lead to new AIDS-related illness or death.
Those who believe laboratory monitoring may not always be essential point out that little evidence exists about responses to second-line treatment after prolonged treatment failure, and question the extent to which viral load testing would result in better outcomes, especially if patients are switched to regimens with a low risk of cross-resistance to first-line treatment.
(A successor to the DART study, called EARNEST, is planning to test the efficacy of lopinavir/ritonavir (Kaletra) monotherapy in second-line treatment, and also the use of Kaletra and raltegravir, in comparison to the WHO-recommended second-line regimen of a boosted protease inhibitor plus two nucleoside analogues. Second-line treatment and studies of resistance in resource-limited settings are reviewed in a recent edition of HIV & AIDS Treatment in Practice.)
One of the critical questions is the extent to which protracted treatment with a failing regimen leads to resistance that will compromise future treatment options. Investigators with the EuroSIDA cohort study set out to analyse the emergence of the key form of cross-resistance, the development of thymidine analogue mutations in people who remain on failing treatment.
Thymidine analogue mutations (TAMs) arise when antiretroviral treatment containing d4T or AZT is not fully suppressing viral load. Three or more TAMs can significantly compromise the subsequent response to nucleoside or nucleotide analogue drugs recommended for second-line use (tenofovir, abacavir and ddI) if they accumulate in a pattern called the TAM-1 pathway, more frequently seen in people failing d4T-containing treatment.
TAMs that cause a modest amount of cross-resistance emerge fairly quickly after treatment failure, but the speed of accumulation of further mutations is unclear. Cross-sectional studies have identified TAMs in anywhere from 5% to 45% of patients on failing therapy in resource-limited settings.
The EuroSIDA investigators identified 339 patients in the cohort who had been taking a regimen containing a thymidine analogue, who had a viral load above 500 copies/ml, at least two genotypic resistance tests while their viral load was detectable, and a detectable viral load throughout the period between resistance tests.
Individuals eligible for inclusion tended to be highly treatment-experienced; 56% had experienced virological failure of at least four drugs prior to the first resistance test eligible for analysis in this study, and 72% had experienced the failure of a protease inhibitor.
By the time the first resistance test took place, patients had already been taking AZT for a median of 9 months (29% of the sample) and d4T for a median of 15 months (the remainder of the sample), and the median viral load was 4.11 log (around 12,500 copies/ml). Seventy per cent were taking 3TC alongside the thymidine analogue and 52% were taking either nevirapine or efavirenz.
During the interval between the first and second viral load tests, an average of six months, viral loads remained very stable, rising by just 0.03 log per month. During this observation period 126 new TAMs were identified during 548 person-years of follow-up, a rate of 0.23 new mutations per year. At this rate, it would take 4.3 years to acquire one new thymidine analogue mutation, the investigators say.
At the time of the first resistance tests 90% of patients had at least one TAM; this had increased to 93% by the time of the second test. Quite a high proportion of patients with a TAM-1 resistance profile, associated with d4T, already had mutations conferring a high level of cross-resistance to other NRTIs (68% had the 215Y mutation, for example, and 65% the 41L mutation). Patients with a TAM-2 resistance profile, more commonly associated with AZT treatment, had a somewhat lower frequency of resistance mutations at baseline (62% had at least one mutation).
Acquisition of new TAMs was faster in patients with three or fewer TAMs at baseline and in patients with higher sensitivity to all drugs in the failing regimen at baseline. Acquisition of resistance was much faster when thymidine analogues accumulated in the TAM 2 resistance pathway, usually associated with AZT.
This faster rate of resistance accumulation in the TAM-2 group in those with fewer mutations might be explained by lower rates of drug resistance at the first test in this group. Patients in the TAM-1 group, who already had a high frequency of mutations at the first resistance test, may simply have reached the point where new resistance mutations were less likely to emerge, owing to the accumulating cost of further viral mutations to the virus’s ability to replicate.
Patients taking tenofovir had the slowest emergence of new TAMs, while patients taking NNRTIs had a higher rate of new TAMs than patients taking protease inhibitors.
The authors say that they can’t determine why patients were kept on failing regimens for a long time, but rule out slow viral load increases by showing that there was no difference in the rate of accumulation between those who had one pair of resistance tests and those who had more than one pair of tests.
A weakness of this analysis is that it cannot provide information on subsequent response, but using two algorithms for predicting subsequent drug sensitivity based on resistance mutations (the Rega Interpretation System and the ANRS system), the researchers report that given the mutations which emerged between the two resistance tests, the nucleoside pair that would suffer the least loss of activity as a result of the resistance mutations arising would be tenofovir and 3TC, one of the two nucleoside backbones recommended by the World Health Organization for use in second-line treatment in resource-limited settings.
Implications
The findings are not likely to settle the disagreements about treatment monitoring strategies, but do indicate that the risks to second-line treatment response of delayed switching vary substantially between patients. More information is needed from analyses of studies such as DART to determine how resistance profile and duration on failing treatment affect subsequent responses to second-line and third-line treatment.
Cozzi-Lepri A et al. Rate of accumulation of thymidine analogue mutations in patients continuing to receive virologically failing regimens containing zidovudine or stavudine: implications for antiretroviral therapy programs in resource-limited settings. J Infect Dis 200: 687-97, 2009.