Atripla or efavirenz plus Truvada failed one in 16 people in first year

Although analyses of the failure rate of ART have been conducted before, a new analysis is the first to compute these rates for a single-drug regimen – namely efavirenz, tenofovir and emtricitabine, the one used most frequently in high-income countries as the first-line regimen between about 2006 and 2014. In 2007, the drugs were combined as Atripla, the first one-pill, once-a-day HIV treatment. The study found that in their first year on this regimen, just over 6% of people experienced a rebound of their viral load after having been virally supressed.

This combination of drugs, as generics, is still a first-line regimen preferred by the World Health Organization for lower-income settings.

Estimates of how often people on antiretroviral therapy (ART) do not remain virally undetectable are important both in order to gauge the potency of ART, and to answer the question of how often people with HIV maintain undetectable viral loads and therefore cannot transmit HIV.

Virological failure

The study was published by COHERE, a large pan-European ‘cohort of cohorts’ of patients spread across most countries in Europe. The current paper analysed virological failure, treatment interruption and switching to another regimen in 19,527 people in 20 cohorts in 12 countries, plus one international cohort.

The researchers collected data on patients whose first ART regimen consisted of efavirenz, tenofovir and emtricitabine, whether as Atripla, as efavirenz (Sustiva) plus Truvada, or as the separate drugs.

They found that 6.3% of people who had taken the regimen as their first one experienced virological failure (defined as a viral load over 200 copies/ml after initial suppression) in the first year on treatment, but that the rate declined to 3.5% in the second year and was down to about 1.7% by the seventh year.

These rates include people who switched to other regimens. When considering people who stayed on the Atripla drugs, the failure rate was 5.75% (one in 17) in the first year, 3% in the second year, and 1.5% by the sixth year.

The rates compare favourably to a previous analysis of people on first-line ART (any combination regimen) in a single cohort, UK-CHIC, which was published in 2017. That found a virological failure rate of 8.1% in the first year and 5.8% in the second year, which then settled down to a steady rate of 1.5% after eight years on the regimen. By year ten 30% of people had experienced viral failure cumulatively. This compares with a 35% cumulative virological failure rate after ten years in the COHERE analysis.

Discontinuation and switching

The COHERE analysis did not just look at virological failure, but also at rates of complete ART discontinuation and of switching to another regimen with or without viral load rebound. The complete discontinuation rate was 3% in the first year in all people who had started on the Atripla drugs and was down to no more than 1.2% by year seven. Discontinuation rates in people who had only taken the initial regimen were 2.5% in the first year and were down to below 0.8% by the seventh.

The annual rates of switching to another regimen, with or without virological failure, were 13.6% in the first year (one in seven people) and 8.5% in the second year, declining to about 5.5% per year by year seven. These rates are low considering that both efavirenz and tenofovir are associated with well-publicised side-effects.

Who did better?

Most patients in the COHERE group were either gay and bisexual men (about 60%) or heterosexual (30%): only 2.7% were injecting drug users. Ethnicity was not recorded in 42% of people, was white in 41% and black in 12%. Their average age was 39 when they started ART, with an average baseline CD4 count of 218 cells/mm³ and a viral load of 70,500 copies/ml.

Different types of people had different rates of viral failure and discontinuation. Heterosexual women, heterosexual men and male injecting drug users were 19%, 32% and 53% more likely, respectively, to experience viral load rebound than gay and bisexual men. People of black ethnicity were 33% more likely than white people to experience treatment failure.

High baseline viral load was associated with more failure after initial suppression: compared with people with a baseline viral load of 20,000 to 100,000 copies/ml, people with baseline viral loads over 100,000 copies/ml were 28% more likely to rebound and those with baseline viral loads over 500,000 copies/ml were 53% more likely to rebound.

People who started in later years were more likely to maintain viral suppression. Compared with people who started in 2009-2010, people starting in 2002-2004 were 64% more likely to experience viral failure, and people who started in 2013-2014 were 14% less so.

A few groups had notably higher rates of complete ART discontinuation. Injecting drug users of either sex were over three times more likely to discontinue than others; people of black ethnicity 54% more likely than white; and young people under 25 years 90% more likely than people aged 35-45.

Conclusions

As the authors note, efavirenz-based regimens are no longer the preferred first-line regimen, with the British HIV Association (BHIVA) taking them off their ‘preferred’ list in 2015. The European AIDS Clinical Society (EACS) recommended in 2017 the third-line drug as being either an integrase inhibitor, the protease inhibitor boosted darunavir, or the non-nucleoside reverse trancriptase inhibitor drug rilpivirine; tenofovir and emtricitabine are still, however, recommended as the second and third drugs.

Tenofovir and emtricitabine were licensed by the EU in 2003, their combination as Truvada in 2005, and Atripla in 2007. In the UK, prescriptions of Atripla or its component drugs peaked in 2010 and have declined since then. So the analysis may be regarded as a ‘historical snapshot’ of the success of one particular regimen.

The good news is that, because efavirenz-based regimens were tailed off due both to side-effects and because HIV found it easy to become resistant to it, the newer regimens are likely, when their turn comes for analysis, to had had lower failure rates.