University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Adverse events associated with nevirapine and efavirenz-based first-line antiretroviral therapy: a systematic review and meta-analysis
Global Health Sciences Literature Digest
Published February 25, 2013
Journal Article

Shubber Z, Calmy A, Andrieux-Meyer I, Vitoria M, Renaud-Thery F, Shaffer N, Hargreaves S, Mills EJ, Ford N. Adverse events associated with nevirapine and efavirenz-based first-line antiretroviral therapy: a systematic review and meta-analysis. AIDS. 2013 Jan 22. [Epub ahead of print]


Systematic review of adverse events associated with nevirapine (NVP)- compared to efavirenz (EFV)-based first-line antiretroviral therapy (ART) among HIV-infected adults and children.


Randomized controlled trials (RCT) and prospective cohort studies.


Treatment-naïve HIV-infected patients.


NVP- or EFV-based first-line ART.

Main Outcome Measures

Primary: discontinuation of regimen as a result of any adverse event due to NVP or EFV. Secondary: hepatotoxicity, skin rash, hypersensitivity reaction, central nervous system (CNS) toxicity, other neurological toxicities, lipid changes, and toxicity-related mortality.

Search Methods

A range of relevant keywords and medical subject heading (MeSH) terms were used to search the Cochrane Central Register of Controlled Trials, EMBASE, LILACS, PubMed and Web of Science. It is not clear whether there were limits to language. An online archive of conference abstracts from International AIDS Society meetings was searched. The date range for the searches of the peer-reviewed literature and conference abstracts was from inception to July 2012; a follow-up search of PubMed was done in October 2012. The Current Controlled Trials database was also searched for ongoing trials. The reviewers checked reference lists of included trials and of other relevant articles.

Searches, Screening and Data Extraction

A total of 2,139 records was retrieved. It is not clear whether this includes duplicate records that were subsequently excluded. One reviewer made a first screening and excluded 1,809 records. Two reviewers working independently then screened the remaining 330 records and excluded 212 records. One hundred eighteen records were identified for full-text review. An additional 19 records were identified in bibliographies of other studies. Two reviewers working independently extracted data from included studies.


Thirty-four studies met the review's inclusion criteria. Eight were RCTs, and 26 were prospective cohort studies. All eight RCTs and 22 cohort studies reported data on adults (total n=26,446); four cohort studies also reported data on children aged &15 years (total n=3,975). The sample size of the included studies ranged from 54 patients to 3,481 patients. Nineteen of the 30 adult studies were conducted in low and low- and middle-income countries (LMIC). All four studies reporting outcomes in children were cohort studies carried out in LMIC (India, South Africa, Uganda, and Uzbekistan). The type and frequency of adverse drug event monitoring was inconsistently reported and differed among the studies. Most studies, however, reported doing routine laboratory assessments of both renal and liver functions. Study duration ranged from six months to nine years.

Overall risk of bias in the included studies was deemed low to moderate. The methods used for reporting adverse events were considered to be at low risk of bias for most studies, and for 23 of 25 studies where loss-to-follow up could be ascertained it was less than 20%.

Outcomes: Adults: Drug discontinuations due to adverse events were reported by 18 studies (17,512 patients). For patients on NVP, the proportion of patients discontinuing treatment due to any adverse event ranged from 1.3% (95% confidence interval [CI] 0.9% to 1.7%) to 30.5% (95% CI 21.3% to 40.5%), with an overall pooled proportion of 8.7% (95% CI 6.7% to 10.6%). For patients receiving EFV, this proportion ranged from 0.18% (95% CI 0.16% to 1.5%) to 22.6% (95% CI 16.7% to 29.2%), with a pooled proportion of 5.8% (95% CI 4.2% to 7.3%). Patients on NVP had more than two times higher odds of discontinuing treatment due to any adverse event compared to patients on EFV (odds ratio [OR] 2.2, 95% CI 1.9 to 2.6), though heterogeneity among the studies was high (I2=70.5%). There was a tendency towards a higher frequency of discontinuations among patients on NVP with a higher CD4 cell count (16.9% at CD4 =250 cells/µL versus 7.2% at CD4 <250 cells/µL, p=0.08); this association was not seen for EFV. A higher frequency of treatment discontinuation was also reported by studies with a follow-up time of greater than one year compared to studies with a follow-up time of less than one year (p<0.001), and for studies carried out in high-income countries compared to lower-income countries (p<0.02). All other subgroup findings were not statistically significant. No differences were found in the frequencies of treatment discontinuations reported by cohort studies and RCTs respectively.

Patients receiving NVP had higher odds of experiencing any grade of hepatotoxicity (OR 1.5, 95% CI 1.3 to 1.8) or severe hepatotoxicity (OR 3.3, 95% CI 2.5 to 4.2) compared to patients on EFV. They also had higher odds of experiencing any grade of skin toxicity (1.8, 95% CI 1.5 to 2.2), severe skin toxicity (OR 3.9, 95% CI 2.5 to 5.4), and severe hypersensitivity reactions (OR 2.4, 95% CI 1.9 to 2.9) compared to patients receiving EFV.

Seven studies reported 64 cases of Stevens-Johnson syndrome among 7,391 patients exposed to NVP, giving a pooled proportion of 0.7% (95% CI 0.5% to 0.9%). In contrast, patients receiving EFV had higher odds of experiencing any CNS-related adverse event (OR 2.1, 95% CI 1.9 to 2.4) and severe CNS-related adverse events (OR 3.4, 95% CI 2.1 to 5.4). There were no differences between the two drugs in the occurrence of other neurological events (OR 0.9, 95% CI 0.6 to 1.2) or lipid-associated adverse events (OR 0.9, 95% CI 0.6 to 1.2). Deaths attributed to toxicity were rare (<1%) for both NVP (14 deaths among 5,835 patients) and EFV (three deaths among 1,380 patients).

Outcomes: Children: Four prospective cohort studies reported adverse drug outcomes in children. In a study from India, the frequency of rash was found to be higher for NVP (20%, 95% CI 13.1% to 30.0%) compared to EFV (14%, 95% CI 5.8% to 26.7%); the differences in hepatotoxicity were not statistically significant. In a study from Uganda, 11 patients on NVP (6.4%, 95% CI 3.3% to 11.2%) developed lipodystrophy resulting in antiretroviral drug substitution; no patients on EFV developed this adverse event. The occurrence of CNS-associated adverse events was lower among patients on NVP (7.6%, 95% CI 4.1% to 12.6%) compared to EFV (14.1%, 95% CI 9.4% to 20.1%); one patient (0.6%, 95% CI 0.1 to 3.1%) on EFV, changed treatment for this reason. A conference abstract from Uzbekistan reported a greater frequency of rash among children receiving NVP (7.5%) compared to EFV (3.8%). A study from South Africa found that children had seven times higher odds of discontinuing NVP-based first-line ART compared to EFV-based therapy (OR 7.1, 95% CI 3.3 to 15.4).


The reviewers conclude that their analysis supports prior findings of individual studies reporting a greater frequency of both liver and skin toxicities associated with NVP compared to EFV and a greater frequency of CNS toxicity associated with EFV compared to NVP. They also point out the relative paucity of pediatric evidence.

Quality of the Evidence

The reviewers used the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology(1) to assess evidence quality. GRADE ranks the quality of evidence on four levels: "high," "moderate," "low" and "very low." Evidence from randomised controlled trials starts at "high," but can be downgraded based on study limitations, inconsistency of results, indirectness of evidence, imprecision or for reporting bias. Evidence from observational studies starts at "low," but can be upgraded if the magnitude of treatment effect is very large, if there is a significant dose-response relation, or if all possible confounders would decrease the magnitude of an apparent treatment effect. Evidence from observational studies can also be downgraded. The reviewers found that the evidence base for the critically important outcomes of discontinuations and severe adverse events was of moderate quality, except for severe skin toxicity, for which the evidence base was determined to be of low quality. For children, the evidence base was determined to be of very low quality. This was mainly due to the limited number of studies, each of which reported different outcomes, resulting in imprecision of the estimated frequency for specific adverse drug outcomes.

Quality of the Review

This was a very high quality systematic review. It meets every relevant criterion of the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines.(2)

In Context

NVP will likely continue to be an important drug for the management of HIV-infected individuals, and CNS-associated side effects associated with EFV remain a concern, particularly in specific patient groups such as post-natal women and patients with psychiatric comorbidities. Nevertheless, the overall benefit may be greater in EFV in terms of reduced toxicity-driven drug substitution and better compatibility with tuberculosis drugs. Recent evidence also suggests superior efficacy of EFV compared to NVP,(3, 4) and no greater risk of teratogenicity.(5, 6)

Programmatic Implications

Clinicians should continue to follow current World Health Organization (WHO) ART guidelines for adults and adolescents(7) and for children.(8) WHO is expected to release new ART guidelines in July 2013.


  1. Guyatt GH, Oxman AD, Schünemann HJ, Tugwell P, Knottnerus A. GRADE guidelines: a new series of articles in the Journal of Clinical Epidemiology. J Clin Epidemiol. 2011 Apr;64(4):380-2. doi: 10.1016/j.jclinepi.2010.09.011. Epub 2010 Dec 24.
  2. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009) Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097. doi:10.1371/journal.pmed.1000097
  3. Tang MW, Kanki PJ, Shafer RW. A review of the virological efficacy of the 4 World Health Organization-recommended tenofovir-containing regimens for initial HIV therapy. Clin Infect Dis 2012; 54 (6):862-875.
  4. Amoroso A, Etienne-Mesubi M, Edozien A, Ojoo S, Sheneberger R, Obiefune M, et al. Treatment outcomes of recommended first-line antiretroviral regimens in resource-limited clinics. JAIDS 2012; 60 (3):314-320.
  5. Ford N, Mofenson L, Kranzer K, Medu L, Frigati L, Mills EJ, et al. Safety of efavirenz in first-trimester of pregnancy: a systematic review and meta-analysis of outcomes from observational cohorts. AIDS 2010; 24 (10):1461-1470.
  6. Ford N, Calmy A, Mofenson L. Safety of efavirenz in the first trimester of pregnancy: an updated systematic review and meta-analysis. AIDS 2011; 25 (18):2301-2304.
  7. World Health Organization. Antiretroviral therapy for HIV infection in adults and adolescents: Recommendations for a public health approach (2010).[accessed 10 February 2013]
  8. World Health Organization. Antiretroviral therapy for HIV infection in infants and children: Towards universal access. Recommendations for a public health approach (2010). [accessed 10 February 2013]