iconstar paper   HIV Articles  
Back grey arrow rt.gif
 
 
Risk of Suicidal Behavior With Use of Efavirenz: Results from the Strategic Timing of Antiretroviral Treatment Trial
 
 
  Download the PDF here
 
originally presented at IAC 2016 - Increased risk of suicidal behaviour with use of efavirenz: Results from the START Trial - (07/26/16)
 
ACTG STUDY -Association Between Efavirenz as Initial Therapy for HIV-1 Infection and Increased Risk for Suicidal Ideation or Attempted or Completed Suicide: An Analysis of Trial Data - Conclusion:
Initial treatment with an efavirenz-containing antiretroviral regimen was associated with a 2-fold increased hazard of suicidality compared with a regimen without efavirenz. http://www.natap.org/2014/HIV/071414_10.htm
 
Efavirenz & Suicidality....New analysis among Medicaid & Commercial databases finds no association, reflects current EFV use patterns & safer/proper use of EFV Risk of Suicidal Behavior With Use of Efavirenz: Results from the Strategic Timing of Antiretroviral Treatment Trial, from BMS http://www.natap.org/2014/IDSA/IDSA_45.htm
 
-------------------------------------------
 
Clinical Infectious Diseases 12 March 2018
 
Alejandro Arenas-Pinto,1 Birgit Grund,2 Shweta Sharma,3 Esteban Martinez,4 Nathan Cummins,5 Julie Fox,6 Karin L. Klingman,7 Dalibor Sedlacek,8 Simon Collins,9 Patricia M. Flynn,10 William M. Chasanov,11 Eynat Kedem,12 Christine Katlama,13 Juan Sierra-Madero,14 Claudia Afonso,15 Pim Brouwers,7 and David A. Cooper16; for the INSIGHT START study group 1Medical Research Council Clinical Trials Unit, University College London, United Kingdom; 2School of Statistics and 3Division of Biostatistics, University of Minnesota, Minneapolis; 4Hospital Clinic, Barcelona, Spain; 5Mayo Clinic, Rochester, Minnesota; 6Guy's and St. Thomas' NHS Foundation Trust, London, United Kingdom; 7Division of AIDS, National Institute of Allergy & Infectious Diseases, National Institutes of Health, Bethesda, Maryland; 8University Hospital Plzen, Czech Republic; 9HIV i-Base, London, United Kingdom; 10St. Jude Children's Research Hospital, Memphis, Tennessee; 11Cooper University Hospital, Camden, New Jersey; 12Clinical Immunology Unit, Rambam Health Care Center, Haifa, Israel; 13Hospitalier Pitié-Salpétrière, Paris, France; 14Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Tlalpan, México; 15Hospital de Santa Maria, Lisbon, Portugal; and 16The Kirby Institute, University of New South Wales, Sydney, Australia
 
"In conclusion, in the START study, ART-naive participants who used EFV in the immediate ART group had an increased risk of suicidal behavior compared with ART-naive controls. Preexisting psychiatric conditions were the strongest predictor of suicidal behavior. Therefore, screening for major psychiatric conditions before EFV initiation would be advisable.....In the general population, patients with severe depression and other serious psychiatric conditions are at higher risk for suicidality [14-16]. In our study, the strongest predictor of suicidal behavior in the immediate ART group was a preexisting psychiatric diagnosis, both among those using EFV and those with other ART. Given the higher absolute risk of suicidal behavior among those with psychiatric conditions, the association between EFV exposure and suicidal behavior supports the recommendation in national and regional guidelines to avoid prescribing EFV for patients with past or current psychiatric conditions [8, 17, 18]. Recreational drug use, including alcohol, was also independently associated with suicidal behavior, which is also consistent with findings in the general population and those with other chronic conditions, particularly at a younger age [15, 19, 20]....Screening for psychiatric conditions (mainly depression) before prescribing EFV could reduce the risk of suicidality. However, screening may not be feasible or affordable, particularly in low- to middle-income countries where EFV continues to be frequently used in first-line ART."
 
Abstract
 
Background
Randomized trials have shown increased risk of suicidality associated with efavirenz (EFV). The START (Strategic Timing of Antiretroviral Treatment) trial randomized treatment-naive human immunodeficiency virus (HIV)-positive adults with high CD4 cell counts to immediate vs deferred antiretroviral therapy (ART).
 
Methods
The initial ART regimen was selected prior to randomization (prespecified). We compared the incidence of suicidal and self-injurious behaviours (suicidal behavior) between the immediate vs deferred ART groups using proportional hazards models, separately for those with EFV and other prespecified regimens, by intention to treat, and after censoring participants in the deferred arm at ART initiation.
 
Results
Of 4684 participants, 271 (5.8%) had a prior psychiatric diagnosis. EFV was prespecified for 3515 participants (75%), less often in those with psychiatric diagnoses (40%) than without (77%). While the overall intention-to-treat comparison showed no difference in suicidal behavior between arms (hazard ratio [HR], 1.07, P = .81), subgroup analyses suggest that initiation of EFV, but not other ART, is associated with increased risk of suicidal behavior. When censoring follow-up at ART initiation in the deferred group, the immediate vs deferred HR among those who were prespecified EFV was 3.31 (P = .03) and 1.04 (P = .93) among those with other prespecified ART; (P = .07 for interaction). In the immediate group, the risk was higher among those with prior psychiatric diagnoses, regardless of prespecified treatment group.
 
Conclusions
Participants who used EFV in the immediate ART group had increased risk of suicidal behavior compared with ART-naive controls. Those with prior psychiatric diagnoses were at higher risk.
 
Predictors of Suicidal Behavior
 
A prior psychiatric diagnosis was the strongest predictor of suicidal behavior among participants in the immediate group who started ART, both among those who were prespecified EFV (HR, 12.5; 95% CI, 4.7-33.6; P < .001; Table 4) and those prespecified other ART (HR, 9.3; 95% CI, 2.4-36.4; P = .001; Supplementary Table 5); there was no evidence for a difference by ART type (P = .79 for interaction). Among those who were prespecified EFV, heavy alcohol use and recreational drug use were independently associated with increased risk (HR of 4.6 and 2.6, respectively), while the risk decreased with age (HR, 0.51 per 10 years older). The time-updated indicator for EFV use was not associated with risk of suicidal behavior (HR, 1.4; P = .74); however, power for this variable was low because almost all participants started EFV shortly after randomization (Table 4). Median time from EFV start to suicidal behavior was 10.2 months.
 
BACKGROUND - Efavirenz (EFV) has been a preferred option for first-line antiretroviral therapy (ART) and is recommended in the World Health Organization guidelines for all human immunodeficiency virus (HIV)-positive adults and adolescents [1], preferably as part of a fixed-dose combination [2]. EFV is usually well tolerated, but neuropsychiatric adverse effects can result in treatment switching [3] or serious clinical complications. A metaanalysis of 4 trials in ART-naive patients showed an increased risk of suicidality in those randomized to EFV-based ART compared with those randomized to other regimens [4]. In contrast, several observational studies found no associations of EFV with suicidality [5-7], possibly because, consistent with some prescribing guidelines [8], providers avoided EFV in patients who were at increased risk of suicidality, resulting in a higher proportion of high-risk participants in non-EFV control groups, which might bias observational studies.
 
The Strategic Timing of Antiretroviral Treatment (START) trial demonstrated clear clinical benefit of immediate ART by decreasing the risk of serious AIDS and serious non-AIDS conditions by 57% in HIV-positive adults with near-normal CD4 cell counts [9]. In START, "suicidal behavior" was the second most frequent type of serious event reported, and 75% of participants randomized to immediate ART initiation used EFV-based ART. We investigated whether initiating EFV increased risk of suicidal behavior more than initiating other ART. This was done by using START's randomized design to find appropriate control groups.
 
DISCUSSION
 
In the START study, there was no overall difference between the immediate vs deferred ART groups in the incidence of suicidal behavior events. However, among participants whose prespecified ART regimen included EFV, those in the immediate arm who started EFV had a higher risk of suicidal behavior (HR, 3.31; 95% CI, 1.19.9; P = .03) compared with those who were randomized to the deferred arm and were not yet using any ART. Conversely, among participants who were prespecified other ART, there was no excess risk of suicidal behavior in the immediate ART group (HR, 1.04). This is consistent with a post hoc metaanalysis that combined data from 4 ART-naive trials [4]. In contrast, cohort analyses and data extracted from regulatory agency databases have failed to observe any excess risk of suicidality associated with EFV [5-7]. Observational studies that compared EFV use to other ART are unreliable. However, because EFV is often avoided for patients with elevated risk of suicidal behavior, resulting in higher a priori risk of suicidal behavior in the other ART group. This higher a priori risk was evident in the START study (Figure 1). We minimized bias by using the randomized design of the study to identify control groups for EFV (or other ART) prior to randomization.
 
In the general population, patients with severe depression and other serious psychiatric conditions are at higher risk for suicidality [14-16]. In our study, the strongest predictor of suicidal behavior in the immediate ART group was a preexisting psychiatric diagnosis, both among those using EFV and those with other ART. Given the higher absolute risk of suicidal behavior among those with psychiatric conditions, the association between EFV exposure and suicidal behavior supports the recommendation in national and regional guidelines to avoid prescribing EFV for patients with past or current psychiatric conditions [8, 17, 18]. Recreational drug use, including alcohol, was also independently associated with suicidal behavior, which is also consistent with findings in the general population and those with other chronic conditions, particularly at a younger age [15, 19, 20].
 
In a recent systematic review and metaanalysis, a higher incidence of severe neuropsychiatric adverse events but not suicidality was observed in ART-naive adults randomized to EFV-based ART compared with other regimens. However, there were no completed suicides in the metaanalysis studies and the overall rate of suicidal ideation was extremely low (0.6%), affecting their power to investigate differences between groups [21].
 
Screening for psychiatric conditions (mainly depression) before prescribing EFV could reduce the risk of suicidality. However, screening may not be feasible or affordable, particularly in low- to middle-income countries where EFV continues to be frequently used in first-line ART. The Encore1 trial reported a lower rate of EFV-related adverse events in patients treated with reduced compared with standard doses [22], suggesting that lower doses of EFV may provide a better safety profile.
 
The main strengths of our study are the use of START's randomized design and the standardized reporting of serious suicidal behavior events. Our study had several limitations. First, it was a post hoc analysis of a trial in which EFV exposure was not the randomized intervention. However, by comparing the immediate and deferred treatment groups within subgroups by prespecified ART type, we estimated the effect of EFV (and other ART) vs randomized control groups, thus minimizing selection bias. Second, not all participants in the immediate group started their prespecified regimen, and some discontinued prespecified EFV. However, adherence overall was high (Figure 3). Third, participants in the deferred group gradually started ART; while we censored follow-up at ART start in the third analysis, this censoring compromises the protection by randomization. Fourth, nucleoside reverse transcriptase inhibitors differed between the EFV and other ART groups. Finally, the number of events was low, which limited the power; therefore, results need to be interpreted with caution.
 
In conclusion, in the START study, ART-naive participants who used EFV in the immediate ART group had an increased risk of suicidal behavior compared with ART-naive controls. Preexisting psychiatric conditions were the strongest predictor of suicidal behavior. Therefore, screening for major psychiatric conditions before EFV initiation would be advisable.
 
RESULTS
 
Baseline Characteristics

 
The START study enrolled 4684 participants at 215 sites in 35 countries. Baseline characteristics have been described previously [9, 13]. The prespecified ART regimen included EFV in 3515 (75%) participants. Table 1 shows the baseline characteristics by type of prespecified ART. Participants with prespecified EFV less frequently lived in high-income countries (34% vs 81%), were less likely to be current smokers (29% vs 42%), and less likely to have used recreational drugs (24% vs 38%). Within the 2 subgroups, the immediate and deferred treatment groups were well balanced for baseline characteristics (Supplementary Appendix, Supplementary Table 2).
 

AASDL1

Preexisting psychiatric diagnoses were less common in those with prespecified EFV compared with other ART (3.1% vs 13.9%), as was use of psychotropic medication (5.2% vs 16.8%; Table 1). The prevalence of psychiatric conditions at baseline was higher in high-income regions (United States, Europe, and Australia) compared with low- to middle-income regions (Latin America, Africa, and Asia; 11% and 1.4%, respectively).
 
ART Use Through Follow-up
 
Figure 3A shows the use of EFV and other ART over time in the immediate and deferred ART groups, separately for participants with prespecified EFV and other ART. Among those prespecified EFV in the immediate ART group, 94% were using ART by month 4 and 82% were on EFV (Figure 3A, upper-left panel). In the deferred arm, median time to ART initiation was 3.2 years, 46% ever initiated ART, and 31% initiated EFV (upper-right panel). Among those with other ART prespecified, ART use in both treatment groups was slightly higher, and a few participants also used EFV (Figure 3A, lower panels).
 
Figure 3B shows cumulative ART use as a proportion of follow-up time accrued. Among those with EFV prespecified, ART was used for 94% of follow-up time in the immediate group vs 26% in the deferred group, and EFV was used for 76% and 15%, respectively. Among those with other ART prespecified, ART was used for 95% and 35% of time in the immediate and deferred groups, respectively. EFV was used for 6% and 4% of time, respectively.
 
Suicidal Behavior
 
Suicidal behavior was reported for 28 participants in the immediate ART group and 25 in the deferred group over a mean follow-up of 3.2 years, rates of 0.39 and 0.34/100 person-years (PY), respectively. The estimated HR (immediate vs deferred group) was 1.07 (95% confidence interval [CI], 0.6-1.8); there was no evidence for a treatment difference (P = .81). Among those with EFV prespecified, 19 participants in the immediate ART group reported suicidal behavior vs 12 in the deferred group, rates of 0.35 and 0.22/100 PY, respectively (HR, 1.38; 95% CI, 0.7-2.9; Table 2). Among those with other ART prespecified, 9 (rate 0.50/100 PY) and 13 participants (rate 0.69/100 PY) had suicidal behavior, respectively (HR, 0.74; 95% CI, 0.3-1.7). The ratio between the HRs, comparing the prespecified EFV subgroup vs the other ART subgroup for "excess risk" of suicidal behavior in the immediate group, was not significant (HR, 1.86; P = .24).
 

AASDL2

We repeated the analyses for the first year of follow-up only. Among those with EFV prespecified, the HR was 3.74 (95% CI, 0.8-17.5; P = .09) compared with an HR of 1.02 among those with other ART prespecified (Table 2). However, event numbers were small in this analysis, CIs were large, and there was insufficient evidence for heterogeneity between the subgroups (HR, 3.67; P = .15).
 
In the third analysis, we compared EFV (or other ART) use to no ART by starting follow-up in the immediate group at ART initiation and censoring follow-up in the deferred group at ART initiation. Among those with EFV prespecified, 18 participants (0.36/100 PY) experienced suicidal behavior in the immediate group (1 event occurred in a participant who was excluded in this analysis because the participant never used EFV) compared with 4 (0.10/100 PY) in the deferred group (HR, 3.31; 95% CI, 1.1-9.9; P = .03). Among those with other ART prespecified, rates in both treatment groups were higher (0.56 and 0.66/100 PY), but the HR was 1.04 (95% CI, 0.4-2.7; P = .93). The excess risk of suicidal behavior in the immediate ART group vs no ART was 3.18-fold higher among those with EFV prespecified compared with those in the other ART subgroup, but there was insufficient evidence for heterogeneity between subgroups (P = .07; Table 2). For the last comparison, the Kaplan-Meier estimates for the cumulative proportion of participants with suicidal behavior are shown in Figure 4.

AASDL3

Kaplan-Meier estimates of the cumulative percent of participants with suicidal or self-injurious behavior, within subgroups by pre-specified antiretroviral therapy (ART) type. In the deferred ART group, follow-up is censored at ART initiation. A, Kaplan-Meier estimates for participants who were pre-specified efavirenz (EFV)-containing ART regimens. In the immediate ART group, follow-up time is started at EFV initiation, excluding 116 (6.6%) participants who never initiated EFV. B, Kaplan-Meier estimates for participants whose pre-specified regimen did not contain EFV. In the immediate group, follow-up time is started at ART initiation, excluding 32 participants who never initiated ART or whose first regimen contained EFV; for the 31 participants who started EFV, follow-up is censored at EFV start. Abbreviations: ART, antiretroviral therapy; Def., deferred; EFV, efavirenz; Imm., immediate; NRTI, nucleoside reverse transcriptase inhibitor.
 
Of the 53 suicidal behavior events, 3 were completed suicide, all in the deferred ART group and after ART initiation. Two of these participants had preexisting depression and anxiety/depression and were treated for these conditions. Both were on Truvada plus darunavir/ritonavir. The third participant had a history of alcohol abuse and started EFV-based ART 18 months before the suicide. Table 3 shows the components of suicidal behavior by MedDRA preferred term for each of the comparisons in Table 2.

AASDL4

The incidence of suicidal behavior was substantially higher among the 371 participants with prior psychiatric diagnosis; among those prespecified other ART, 8 of 162 participants with prior psychiatric diagnoses experienced an event (1.7/100 PY) compared with 14 of 1007 without diagnoses (0.5/100 PY; P = .003). Supplementary Tables S3 and S4 show the event rates and HRs from Table 2 for participants with and without prior psychiatric diagnoses, respectively. Among the 109 participants with prior psychiatric diagnoses who were prespecified EFV, none experienced suicidal behavior in the deferred ART group compared with 6 (2.7 per 100 PY) in the immediate group after starting EFV (Supplementary Table 3).
 
Predictors of Suicidal Behavior
 
A prior psychiatric diagnosis was the strongest predictor of suicidal behavior among participants in the immediate group who started ART, both among those who were prespecified EFV (HR, 12.5; 95% CI, 4.7-33.6; P < .001; Table 4) and those prespecified other ART (HR, 9.3; 95% CI, 2.4-36.4; P = .001; Supplementary Table 5); there was no evidence for a difference by ART type (P = .79 for interaction). Among those who were prespecified EFV, heavy alcohol use and recreational drug use were independently associated with increased risk (HR of 4.6 and 2.6, respectively), while the risk decreased with age (HR, 0.51 per 10 years older). The time-updated indicator for EFV use was not associated with risk of suicidal behavior (HR, 1.4; P = .74); however, power for this variable was low because almost all participants started EFV shortly after randomization (Table 4). Median time from EFV start to suicidal behavior was 10.2 months.

AASDL5

 
 
 
 
  iconpaperstack View Older Articles   Back to Top   www.natap.org