Efavirenz and the brain: are we nearer to solving a mysterious side-effect?

One of the most potent HIV drugs, efavirenz, unfortunately also causes mysterious and sometimes chronic disruptions of mood, thought and sleep. Researchers may have found the key to these side-effects,¹ and evidence that the drug may contribute to the subtle deficits in brain function usually attributed to HIV.

As Sustiva or Stocrin, or in the once-a-day pill Atripla, efavirenz is a staple part of many people’s antiretroviral therapy (ART): in London last year, 53% of all people on ART took it. Yet ever since its introduction in 1999, the drug has been linked to vivid dreams, poor sleep, dizziness, lack of concentration and, in some cases, mood swings, anxiety and depression.²

It’s accepted that these effects are common in the first month or so on the drug. What’s disputed is whether these persist and, if so, whether they have real impact. While some studies found no evidence for long-term side-effects,³ others did^4,5 – and some found that, even in people who had tolerated efavirenz for years, changing to other drugs, within⁶ or outside⁷ the NNRTI class, resulted in improvements in mood.

Early studies found that only 6% of people stopped efavirenz in the first year⁸ but a more recent study⁹ found that as many as 20% of people taking Atripla (efavirenz/tenofovir/FTC) stop taking it within a year, largely due to psychological side-effects.

There have been very few randomised, controlled studies to assess the extent to which these side-effects are peculiar to efavirenz, or which can provide any information about whether some of these effects might be due to HIV infection itself. One of these few¹⁰ found that people on efavirenz had slower reaction times and poorer decision-making skills than people on protease inhibitors. But most research has been in studies where participants knew they were on efavirenz – and the reputation of the drug is such that they may have blamed mental distress on efavirenz when it in fact had another cause.

The picture gets more complicated if you factor in HIV-associated neurocognitive disorder (HAND). This is the complex of deficits in mental performance that has been associated with HIV infection; it too tends to include deterioration of memory, co-ordination and concentration.

Marked brain impairment in people with HIV not taking antiretroviral therapy (ART) usually gets better if they start ART, but the mild version of HAND often persists. Some studies have found that people taking efavirenz have poorer HAND scores¹¹ and others have found that HAND may improve less in people taking efavirenz.¹²

Although mental disturbance, including psychosis and mania, has been linked to exceptionally high efavirenz levels,^{13 ,14} other studies have found no link with levels of efavirenz in blood.¹⁵ But higher drug levels and lower HIV viral load in the cerebrospinal fluid (CSF) that surrounds the brain were associated in one study with worse HAND symptoms,¹⁶ in another with nerve cell loss,¹⁷ and, in a third, HAND symptoms unexpectedly improved when patients stopped ARVs;¹⁸ these however were not efavirenz-specific effects.

The new research found that it might not be efavirenz itself that damages nerve cells but a metabolite, a chemical produced when the body acts on the efavirenz molecule.

In the body, efavirenz turns into two molecules called 7- and 8-hydroxy-efavirenz. The researchers incubated nerve cells taken from rats with efavirenz and these two compounds and found that, while all three were toxic to nerve cells, 8-hydroxy-efavirenz (8HE) was ten times more so. This additional toxicity appeared to have an independent mechanism; 8HE jammed open portals into nerve cells called voltage-operated calcium channels (VOCCs). When these gateways are breached, calcium floods into nerve cells, unleashing a burst of brain signals, causing unpredictable effects.

With high drug levels this led to nerve cell death, but with lower levels the cells lost their ‘dendritic spines’: these are a nerve cell’s ‘input wires’ and mean that the cell is less responsive to the signals reaching it. The typical level of 8HE seen in daily dosing of efavirenz is about three times the level sufficient to denude cells of dendritic spines.

This could explain why blood levels of efavirenz are not always related to side-effects: people who efficiently metabolise efavirenz may have lower blood efavirenz levels but higher peak levels of the 8HE compound.¹⁹

This research comes at a crucial time for efavirenz. The drug will be one of the first widely used ARVs to come off patent next year (see The generic generation). Generic efavirenz will be available cheaply and there will be cost pressures for people to stay on it if they can, rather than switch to costlier drugs.

The fact that a drug harms cells in a lab dish doesn’t necessarily mean it will harm them in the body. Clearly, this line of research needs to be extended. We need to do the same experiments on neurones with metabolites of other drugs; we need to relate brain function in people to levels of 8HE; and there should be longitudinal research to look at how brain impairment improves or deteriorates in people with HIV on and off treatment, and to try to tease apart the symptomatic profiles of different kinds of neural damage.

Efavirenz has long been blamed, rightly or wrongly, for a subtle but pervasive deterioration in the quality of life of some people taking it and it would be good to find out whether the drug’s benefits continue to outweigh this. The question is, with less money available as ARVs become generic, who will fund the research?