October 5, 2013
Suicides among people with HIV on antiretroviral therapy are uncommon, but they appear to occur with significantly greater frequency among those who take efavirenz (Sustiva, Stocrin), according to an analysis presented at IDWeek 2013.
The effects of efavirenz on the central nervous system have been well documented, but much of the discussion around these effects has tended to focus on the drug's potential to cause sleep loss, peculiar dreams, anxiety and even depression. Less-well elucidated has been the potential effect of efavirenz on suicide in particular.
To tease out any potential relationship between efavirenz (a component of the popular, fixed-dose-combination medication efavirenz/tenofovir/emtricitabine [Atripla]) and suicide risk, Katie Mollan and colleagues at the University of North Carolina at Chapel Hill and Harvard School of Public Health, along with a nationwide team of collaborators, closely examined data from four pivotal AIDS Clinical Trials Group (ACTG) studies involving efavirenz from the 2001-to-2007 time period. (The study was funded by the U.S. National Institute of Allergy and Infectious Diseases.)
These four clinical trials -- ACTG 5095, 5142, 5175 and 5202 -- played a role in cementing efavirenz's place within our HIV treatment pantheon. Due to the high total number of patients involved and the long-term nature of the studies, they also cumulatively offer additional opportunities for exploring issues such as suicide, which can be a challenge to examine due to the small number of suicides that occur from trial to trial.
Each of the four studies randomly assigned treatment-naive patients to either an efavirenz-containing regimen or a non-efavirenz-containing regimen and followed them for upwards of three years. A total of 5,332 patients were included in Mollan et al's analysis, and the patient population was relatively well-representative of the U.S. HIV-infected population as a whole: approximately three-quarters of enrolled patients were male, 39% were white, roughly 35% were black and 22% were Latino. Median age at baseline fell roughly in the mid-30s.
The actual regimens used varied by study, but most of the efavirenz-containing regimens also included two NRTIs -- usually lamivudine (3TC, Epivir), emtricitabine (FTC, Emtriva) or zidovudine (AZT, Retrovir). Efavirenz-free regimens also varied, but usually involved the use of a boosted protease inhibitor (typically atazanavir [Reyataz]) and two NRTIs.
In all cases, documentation of patients' psychiatric history utilized the Medical Dictionary for Regulatory Activities. Mollan et al focused in particular on suicidality -- that is, the likelihood of a person 1) expressing a desire to commit suicide, 2) actually attempting suicide and failing, or 3) attempting suicide and succeeding.
Cumulatively, across all four ACTG studies (in an intent-to-treat analysis), 62 suicidality events occurred; the rate of events among patients on efavirenz was 8.08 per 1,000 patient-years, more than double the rate of 3.66 events per 1,000 patient-years among patients who did not take efavirenz (HR = 2.28; CI = 1.27 to 4.10; P = .006). The greater suicidality rate among patients on efavirenz began immediately and persisted throughout the study period.
Of course, it's one thing to express suicidal ideation; it's another thing entirely to actually attempt suicide. On this front, the results fell just shy of statistical significance, but the trend was clear: 17 events occurred among those on efavirenz compared to 5 events among those not on efavirenz, a hazard ratio of 2.58 (CI = 0.94 to 7.06, P = .06). Trend lines indicated that the greater rates of attempted suicide did not begin to emerge until after about one year on treatment, and then increasingly diverged over time.
Mollan noted several factors across the four studies that were associated with suicidality in a multivariable analysis: efavirenz use, of course, with a roughly two-fold higher risk (HR = 2.15, P = .01); but also younger age (people below the age of 30 had a nearly tripled risk relative to people 45 or older), a history of injection drug use (which appeared to increase suicidality risk just as much as efavirenz use) and a history of psychiatric problems (which nearly quadrupled the risk of suicidality). Gender, CD4+ cell count at baseline and occurrence of AIDS-related events were not significantly associated with suicidality risk.
As with any study of this nature, the research here doesn't claim to be perfect: Although the four trials used in this analysis were similar in many ways, they were not identical; they often involved the use of antiretroviral regimens that are no longer commonly prescribed; and none of them were actually designed to in any way closely examine suicidality.
In addition, the actual clinical implications here are not terribly clear. Because the overall number of suicidality events was ultimately so small, it's a challenge (to say the least) to devise from these results some sort of "suicidality risk formula" into which a patient can be plugged prior to taking efavirenz.
Similarly, there's no explicit suggestion from these data that a person with HIV should avoid an efavirenz-containing regimen entirely if he or she has a history of psychiatric disorder, particularly if the drug combination is likely to be the best choice for that individual (in terms of convenience, tolerability, resistance profile, risk for other adverse events, etc.).
On the flip side, it also appears clear from these data that, all other non-psychiatric variables being equal, other first-line treatment options may be worth considering in such patients.
As with almost any issue in HIV care, the best takeaway message from a study like this is probably: Be careful. Be attentive, be cautious, and be aware of the patient's mental condition. If it feels like danger is likely and there are safer alternatives, consider prescribing them. If the patient is on efavirenz and exhibits signs of suicidality, take that into account alongside other pros and cons of a regimen switch. And keep up the psych monitoring: Don't assume that, because the patient seems OK at four months, he or she will still be OK at 14 months, or 24 months, or 34 months. Refer the patient for psychiatric care and counseling as needed, and follow up to ensure he or she receives it.
Most of these items are no doubt already on most providers' checklists. This study simply casts one particular section of those checklists into more stark relief.
Myles Helfand is the editorial director of TheBody.com and TheBodyPRO.com.
Follow Myles on Twitter: @MylesatTheBody.