Emergence of resistance mutations: a comparison between drug classes

One of the largest studies yet of resistance mutation development has confirmed that, in people on 3TC (lamivudine, Epivir)-containing antiretroviral combinations with detectable viral loads, major protease inhibitor mutations developed more quickly than thymidine analogue mutations (TAMs – mutations conferring resistance to AZT and d4T). In the absence of 3TC, protease mutations and TAMS developed at similar rates.

The ideal of antiretroviral therapy is to keep viral load fully suppressed, i.e., undetectable. However, some people remain on antiretroviral regimens that have failed, so that their viral loads continue at detectable levels. It is well understood that, in these situations, HIV will continue to build up drug resistance mutations. Researchers have been trying to understand this process in more detail by finding how quickly resistance mutations develop and which are most common.

In this retrospective study by Goetz and associates, published in the December 15^th edition of the Journal of Acquired Immune Deficiency Syndromes, the team built on earlier work by other researchers by investigating the rates of resistance mutation development in a group of people who remained adherent to stable antiretroviral therapy, after virologic failure had occurred.

Several studies have already shown that, in ongoing “failed therapy” situations, major protease inhibitor resistance mutations develop more quickly than resistance mutations to the nucleosides known as “thymidine analogues” – zidovudine (AZT, Retrovir) and stavudine (d4T, Zerit). (Note that these studies all concern genotypic measurements of resistance mutations, not with high-level resistance as would be measured by phenotypic tests.) One other recent study showed seemingly the reverse – that nucleoside mutations accumulated more quickly than protease mutations. However, this study (by Napravnik et al.) looked at all nucleoside analogue drugs rather than thymidine analogues only.

The new Goetz study looked at medical records of patients in the Veterans Affairs Greater Los Angeles healthcare system, who had a prolonged period of detectable viral load while remaining adherent (as evaluated by case notes and refill records) to regimens of at least three antiretrovirals, and for whom ongoing genotypic resistance test results were available (from between August 1996 and August 2004). The researchers only considered people who initially showed minimal resistance – no multiple drug resistance, and two or fewer TAMs (thymidine analogue mutations). People with two or more major protease inhibitor (PI) mutations were included in the study, but excluded from the PI part of the analysis. This was done to “avoid a ceiling effect on the potential emergence of mutations”, but the researchers concede that it may have biased the results by excluding people who developed resistance more quickly. (Mutations were specified as per the IAS-USA guidelines, Nov 2004.)

A total of 46 episodes of failed therapy, in 41 patients, were analysed. (Five patients had two separate “treatment failures”.) The “episodes” were, on average, 110 weeks long: the researchers believe that their study is the first to follow “as large a cohort of failing patients with low rates of baseline resistance… for as long a period.”

The study did not report actual numeric rates at which mutations emerged. It did, however, note that:

• Major PI mutations emerged more rapidly than TAMs.
• This difference was only seen when 3TC (lamivudine) was part of the combination.
• When 3TC was included, it took an average of 300 days more for a new TAM to emerge than a new major PI mutation.
• TAMs emerged significantly faster in people who were not taking 3TC – “consistent with and extend[ing] previous observations” by other investigators.

The researchers note that the study is limited by “the retrospective and potentially biased case selection.” Also, most of the PIs used were nelfinavir (Viracept), indinavir (Crixivan), and saquinavir (Invirase). Only one person had used amprenavir (Agenerase), none had used any of the more recent PIs, and ritonavir (Norvir) boosting was also rare – a situation that does not reflect current standards of care, and possibly making the results “relevant only for the specific regimens received by our subjects.”

The researchers concluded that “major protease inhibitor mutations arose more frequently and rapidly than did TAMs in patients with sustained virologic failure who received lamivudine [3TC].” However, they acknowledged that the analysis may have been biased by “eliminating patients in whom resistance emerges particularly quickly”, and “strongly caution that these results should not be used to make clinical management decisions… especially for those persons who have access to remaining effective antiretroviral agents.”