Efavirenz is not Depressing

Abstract & Commentary

By Dean L. Winslow , MD, FACP, Chief, Division of AIDS Medicine, Santa Clara Valley Medical Center, Clinical Professor of Medicine, Stanford University School of Medicine, Section Editor, HIV, is Associate Editor for Infectious Disease Alert.

Dr. Winslow is a consultant for Bayer Diagnostics and Pfizer/Agouron, and is on the speaker's bureau for Pfizer/Agouron.

Synopsis: In a randomized trial of efavirenz (EFV) vs protease inhibitor (PI)-containing highly active antiretroviral therapy depressive disorder was encountered no more frequently in the EFV group than in the PI group.

Source: Journot V, et al. Use of Efavirenz is not Associated with a Higher Risk of Depressive Disorders: A Substudy of the Randomized Clinical Trial ALIZE-ANRS 099. Clin Infect Dis. 2006;42:1790-1799.

Alize-anrs was a 48 week randomized controlled trial in which 355 virologically-suppressed, HIV-infected patients receiving PIs were randomized to a continuation regimen containing EFV vs PI. The trial's adverse event database and a self-administered depression scale questionnaire were used to assess depressive disorder. Depressive disorder or suicide attempt occurred in 7% of the PI-treated patients and 8% of the EFV-treated patients (P = .56).


The non-nucleoside reverse transcriptase inhibitor (nnRTI) efavirenz has been an important component of HAART since its regulatory approval in the late 1990s. I was very fortunate to be working at the DuPont Company in the Viral Diseases Drug Discovery Group during the late 80s through mid-90s when our company in-licensed from Merck its most promising L-drug, nnRTI, which we designated as DMP 266 during its development (later given the generic name efavirenz). Our group in Discovery Research verified its amazing in vitro potency,1 while other colleagues in preclinical development demonstrated its favorable pharmacokinetic characteristics, great safety and tolerability in animals. We did learn from these studies that CNS effects were seen in non-human primates at high doses and were seen in the range of blood levels we would likely see in humans dosed at 600 mg daily, so CNS side effects in humans were anticipated before we started our clinical trials. It was an exciting time to be engaged in antiretroviral research. Due to the sudden resignation of the Director of Oncology in Clinical Research and the impending initiation of clinical trials of DMP 266, I gave up most of my laboratory research and took over the job of running the clinical trials of 2 oncology drugs and designing the clinical development plan for DMP 266. We initiated single, followed by multiple, dose pharmacokinetic studies in humans in late 1995, Phase II/III studies (including 003 and 006) and drug interaction studies in early 1996, worked with the ACTG to get DMP 266 in several additional clinical trials, and the rest is history. After I left DuPont in early 1996, Nancy Ruiz and her team did a wonderful job completing those trials and filing the NDA.

Due to the autoinduction of metabolism (by the cytochrome P450 system) by EFV, which occurs in the first 2 weeks of therapy, CNS side effects are generally most prominent in the first few weeks of treatment. In patients who experience these side effects, the most common symptoms include insomnia, often bizarre and vivid dreams, and sometimes jitteriness during the day. Most clinicians find that, if forewarned of these potential side effects, most patients find that these effects become much less severe if they hang in there for a couple of weeks. However, most of us have had a few patients where these CNS side effects are so severe that the EFV must be discontinued. The relationship between EFV side effects and plasma EFV levels is fairly impressive, and the association of reduced metabolism of EFV in certain individuals including those with certain cytochrome P450 2B6 polymorphisms has recently been demonstrated.2

Due to the high prevalence of mood disorders in most HIV-infected populations, concern was raised early on that EFV could exacerbate depression. The registrational trials did not support this concern, but anecdotal experience after regulatory approval and widespread use raised this concern in the minds of many practitioners. A handful of publications in mainly non-peer reviewed publications were purported to show an association between EFV use and depression, and many practitioners started thinking it was inappropriate to prescribe efavirenz in patients with a history of psychiatric disease. The majority of post-registrational, controlled trials did not show such an association.3

This well-controlled trial by ANRS (essentially the French ACTG) dispels the concern about a significant causal association between EFV and new onset of depression or exacerbation in patients with pre-existing depression. While this should reassure providers that EFV can be safely used in patients with stable depression, I still take somewhat of a pragmatic approach and tend to use EFV with caution, or chose a PI in the rare patient who has poorly controlled underlying psychiatric illness (either mood disorder or thought disorder/psychosis). This is because supporting a patient who is already struggling through the initial 2-3 weeks of therapy when CNS side effects are most severe often requires a tremendous expenditure of effort and resources, which is multiplied in the presence of pre-existing poorly controlled psychiatric disease.

It may be of interest to some of the readers that efavirenz was almost not developed. The entire class of Merck L-drug nnRTIs was shelved by Merck management after rapid selection of high-level resistance was observed in vivo as a result of single amino acid substitutions in RT when these drugs were studied as monotherapy in the late 1980s. When the DuPont Merck joint venture was created in 1991, a few believers transferred from the Merck to DuPont organization, including Paul Friedman who came on board at DuPont Merck as President of R&D. Paul knew that antiretrovirals with a low genetic barrier to development of resistance (like the nnRTIs) needed to be studied as part of combination therapy—a concept which was still controversial in the early 1990s; however, this seemed obvious to me, hence the design of our registrational studies. It may also be of interest to some readers that DuPont (the company which discovered the angiotensin receptor blocker, losartan) had probably the premier HIV protease inhibitor discovery program in the world in the late 1980s and early 1990s, and culminated in the discovery using computer-assisted drug design technology of the small molecule cyclic urea compounds.4 Unfortunately after the DuPont Merck joint venture was formed, DuPont's protease inhibitor program was disbanded since it was viewed by Merck management as competing with indinavir, which was entering into Phase II/III trials at that time.


  1. Winslow DL, et al. Selection Conditions Affect the Evolution of Specific Mutations in the Reverse Transcriptase Gene Associated with Resistance to DMP 266. AIDS. 1996;10:1205-1209.
  2. Haas D, et al. A Common CYP2B6 Variant is Associated with Efavirenz Pharmacokinetics and Central Nervous System Side Effects. AACTG Study NWCS 214. 11th Conference on Retroviruses and Opportunistic Infections; Feb 8-11, 2004; San Francisco. Abstract 133.
  3. Clifford DB, et al. Impact of Efavirenz on Neuropsychological Performance and Symptoms in HIV-Infected Individuals. Ann Intern Med. 2005;143:714-721.
  4. Lam PY, et al. Cyclic HIV Protease Inhibitors: Synthesis, Conformational Analysis, P2/P2' Structure-Activity Relationship, and Molecular Recognition of Cyclic Ureas. J Med Chem. 1996;39:3514-3525.