Efavirenz is one of the most widely used nonnucleoside reverse transcriptase inhibitors (NNRTIs) in first-line antiretroviral therapy (ART) and is recommended as a preferred option in adult treatment guidelines [1–3]. However, data from primate studies  and some human case reports [4,5] have raised concern regarding a association of first-trimester efavirenz exposure with central nervous system congenital anomalies. These data resulted in a recommendation by the United States Food and Drug Administration (FDA) in 2005 and the European Medicines Agency (EMEA) to avoid using efavirenz-based regimens in the first trimester of pregnancy (http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_Product_Information/human/000249/WC500058311.pdf) .
Over nearly two decades, an accumulation of data from pregnant women exposed to antiretroviral drugs during the first trimester of pregnancy has allowed for an assessment of the potential association between efavirenz and congenital anomalies. Systematic reviews published in 2010 and 2011 summarized available data at the time and found no evidence of increased risk of congenital anomalies associated with first-trimester exposure to efavirenz [7,8]. On the basis of this evidence, in mid-2012, the WHO released a technical update that highlighted the significant clinical and programmatic benefits of efavirenz use in pregnancy and the potential risks, and recommended that efavirenz can be included as part of preferred first-line therapy in pregnant women and women of childbearing age because the benefits outweighed potential risks .
In order to further validate this technical update in the context of the 2013 revised WHO global guidelines for ART in low and middle-income countries, we conducted an updated review to assess the evidence for the safety of efavirenz in pregnancy.
Materials and methods
This systematic review was conducted according to a study protocol following the requirements of the PRISMA Statement . A preliminary version of this analysis was prepared for the WHO guidelines development group in December 2012. This article presents the final, updated analysis, based on updated literature searches up to 10 January 2014.
Search strategy and study selection
We used a compound search strategy to update a previous systematic review of the safety of efavirenz in pregnancy, published in 2011 . The following databases were searched from 01 July 2011 (the date of the last search) up to 30 June 2013: MEDLINE via PubMed, EMBASE, Cochrane CENTRAL, LILACS and Web of Science; this search was updated in MEDLINE via PubMed up to 10 January 2014. We also searched the websites of two major HIV conferences: all International AIDS Society (IAS) conferences (up to Kuala Lumpur, June 2013) and all Conferences on Retroviruses and Opportunistic Infections (CROI, up to Atlanta, March 2013). We also retrieved the latest report of the Antiretroviral Pregnancy Registry (http://www.apregistry.com) (up to January, 2014). Finally, bibliographies of all relevant articles were screened to check for additional publications. No date, language or other restriction was applied. Searches were done in duplicate (N.F., Z.S.).
After preliminary screening of all the titles obtained from our searches, all abstracts were then assessed for eligibility by two reviewers according to the inclusion criteria defined by the protocol. Our primary outcome of interest was any congenital anomaly (birth defect) (http://www.who.int/mediacentre/factsheets/fs370/en [Accessed 26 January 2014]); secondary outcomes included spontaneous abortions, terminations of pregnancy, stillbirths, preterm deliveries and adverse drug reactions. Studies that only reported data on secondary outcomes were included (i.e. they did not have to also report the primary outcome). We sought to compare the risk of congenital anomalies overall, and specifically neural tube defects, among infants born to women receiving efavirenz during the first trimester of pregnancy to the risk among infants born to mothers exposed to other antiretrovirals in the first trimester of pregnancy, and to background reference prevalence. Nonsystematic observations (case series or case reports that did not include denominators to allow for assessment of risk) and data from animal studies were excluded from all analyses.
Data extraction was done independently, in duplicate, using a standardized form. For each study, we gathered information on the study setting, the study population, the sample size, the timing and duration of efavirenz exposure and birth outcomes. Data on patient and study characteristics, and predefined indicators of potential risk of bias were also extracted. The GRADE system was used to assess the overall quality of the evidence .
Point estimates and 95% confidence intervals (95% CI) were calculated for the proportion of congenital anomalies reported among live births for each study. We excluded spontaneous and induced abortions as well as stillbirths from the numerator and denominator for the estimate of congenital anomalies, consistent with current reporting conventions . For cohorts that reported and compared birth outcomes of infants born to mothers exposed to efavirenz during the first trimester of pregnancy to outcomes of infants born to mothers exposed to other antiretroviral drugs, relative risks (RRs) and 95% CIs were calculated. In the case of zero outcome events in one arm, the Haldane method was applied, adding 0.5 to each arm. The variance of the raw proportions was stabilised using a Freeman–Tukey type arcsine square-root transformation  and estimates were pooled using a DerSimonian–Laird random effects model . Pooled estimates were subsequently back-transformed to the original scale . Prevalence and 95% CIs were calculated for all secondary outcomes. We did not pool data on our secondary outcomes because the background prevalence of the secondary outcomes is known to vary considerably between study settings. We calculated the τ2 statistic using DerSimonian and Laird's method of moments estimator  to assess between-study heterogeneity . Subgroup analyses were conducted to assess the potential effects of study design, geographical location (low/lower-middle income country versus middle/high-income country, as defined by the World Bank), duration of efavirenz exposure and status of publication (full text article versus conference abstract or unpublished data) on the pooled estimates. Sensitivity analyses were conducted comparing the RR estimates if using fixed-effects rather than random-effect methods, the overall pooled prevalence of congenital anomalies with that reported by the antiretroviral pregnancy registry and by running a ‘leave-one-out’ analysis to assess the extent to which the overall result might be influenced by any single study. A P value less than 0.05 was considered significant. Publication bias was assessed using the eggers test for small study effects . All analyses were conducted using STATA (version 12, http://www.stata.com) and GRADE Pro (http://www.gradeworkinggroup.org).
Our updated search yielded 397 additional titles for screening, in addition to the previous review published in 2011 , bringing the total number of study titles screened to 2080 (Fig. 1). Two articles published since the previous review was completed were excluded because they reported outcomes that had previously been reported to the Antiretroviral Pregnancy Registry [19,20]. In total, two updated reports [13,21] from previously published cohorts [22,23], data from one additional published article  and one additional conference abstract  were included in this updated review. Overall, 23 studies were included in this review, comprising 20 articles and conference abstracts [21,24–42], one unpublished study , the Antiretroviral Pregnancy Registry  and one unpublished cohort (the MTCT Plus cohort: E. Abrams, personal communication; data unchanged and carried forward from previous systematic reviews ). Additional data on secondary outcomes for one study  were provided by a conference abstract .
Study characteristics are summarized in Table 1.
Assessment of study quality
Our assessment of the overall risks of bias found that most studies were at a moderate risk of bias. Very few studies considered potential confounders or baseline imbalances between patients receiving efavirenz compared with nonefavirenz regimens (Supplementary Table S1, http://links.lww.com/QAD/A490). Our GRADE review assessed the 12 studies that compared birth outcomes of women exposed to both efavirenz and non-efavirenz containing regimens during the first trimester of pregnancy. We rated the overall evidence base to be of low quality, mainly due to the observational nature of the studies, and due to the limited number of exposures and varying methodologies (Supplementary Table S2, http://links.lww.com/QAD/A490). There was no statistical evidence of publication bias (P = 0.14 using Egger's test for funnel plot asymmetry).
Of the 23 studies included in this review, 22 reported the birth outcomes of 2026 live births among women exposed to efavirenz during the first trimester of pregnancy. Details are summarized in Table 2. Forty-four congenital anomalies were reported, yielding a pooled proportion of 1.63% (95% CI 0.78–2.48; τ2 = 0.01). Of these, one was a neural tube defect. We found no additional cases of neural tube defects compared with previous reviews, thus keeping the overall prevalence of neural tube defects low (0.05%, 95% CI <0.01–0.28).
Results of the overall analysis were not statistically different to the raw proportion of congenital anomalies reported by the Antiretroviral Pregnancy Registry (2.3%, 95% CI 1.3–3.7), and the overall proportion of congenital anomalies in our analysis did not exceed 2% in the ‘leave-one-out’ sensitivity analysis (Supplementary Table S3, http://links.lww.com/QAD/A490).
Twelve studies reported birth outcomes of women exposed to both efavirenz and nonefavirenz-containing regimens during the first trimester of pregnancy. This analysis found no differences in risk of congenital anomalies between these two groups (RR 0.78, 95% CI 0.56–1.08) (Fig. 2). Heterogeneity between the studies was low (τ2 = 0), and none of the subgroup analyses assessing the impact of study design, geographical location, duration of efavirenz exposure and publication status were significant (all P > 0.1).
The reporting of secondary outcomes varied between the studies. Nine studies reported spontaneous abortions, with prevalence ranging from 0 to 16.1% (95% CI 7.6–28.3). The prevalence of stillbirths, reported by seven studies, ranged from 0 to 7% (95% CI 4.0–11.3). Five studies reported on the prevalence of preterm delivery, with prevalence ranging from 9.1 (95% CI 5.3–15.5) to 18.2% (95% CI 7.0–35.5). These data were not pooled because of differences in background prevalence across populations. The large variation in point estimates and wide CIs resulting from small sample sizes of individual studies make these findings difficult to interpret with respect to any indication of increased risk for these outcomes.
The proportion of medical terminations of pregnancy, reported by 10 studies, ranged from 0 to 34%. Three studies reported data on terminations of pregnancy (terminations not associated with prenatal screening) for women exposed to efavirenz and nonefavirenz-based regimens; pooling these data gave an RR of 2.81 (95% CI 0.94–8.36) for pregnancy terminations compared with women exposed to nonefavirenz-based therapy. A fourth study, from South Africa, reported that 19 of 56 women (34%) who conceived while on efavirenz terminated their pregnancy; data from this study were not included in the pooled analysis, as comparative information for non-EFV regimens was not reported . None of these terminations was due to congenital anomalies detected in utero (Sheree Schwartz, personal communication). Data on secondary outcomes are summarized in Supplementary Table S4, http://links.lww.com/QAD/A490.
Adverse drug reactions
Only two studies reported on adverse drug reactions among mothers receiving efavirenz during pregnancy. The first study, reporting data on 25 first-trimester exposures, found no adverse drug reactions resulting from efavirenz treatment . The second study, reporting data on 56 first-trimester exposures, reported one adverse drug reaction resulting from efavirenz therapy (vomiting) .
This review provides an updated, systematic review of evidence on the safety of efavirenz use during the first trimester of pregnancy and includes information from an additional 589 live births compared with the last review published in 2011 , including published data up to January 2014. Nevertheless, the use of efavirenz in pregnancy remains a controversial topic, and this is reflected by varying recommendations in national guidelines: guidelines from the British HIV Association recommend using efavirenz in pregnancy  while those issued by the European AIDS Clinical Society  recommend avoiding efavirenz during the first 8 weeks of pregnancy. A recent report from France suggested an increased risk of neurologic defects (none of which were neural tube defects) among infants born to women receiving efavirenz during the first trimester of pregnancy . However, when these data were considered together with other available data in our review, we still found no evidence of an increased risk of congenital anomalies associated with first-trimester exposure to efavirenz compared with exposure to other antiretroviral drugs. In the available dataset of published reports, there is only one neural tube defect, giving a prevalence of 0.05%, which is in line with the prevalence of 0.1% reported in the general population . The prevalence of overall congenital anomalies is also in line with that reported in the general population . Although this finding is based on an evidence base that is rated as low quality according to the GRADE approach, randomized trials are unlikely to ever be conducted to address this question.
Strengths of this review include a broad search strategy that identified a number of studies not yet published in the literature and the inclusion of updated data for several cohorts. Results appeared to be consistent across studies, as demonstrated by low statistical heterogeneity and the robustness of the main findings to sensitivity and subgroup analyses. There was no statistical evidence of publication bias, but these tests are less reliable when the number of studies is small  and we cannot rule out the possibility of publication bias, but consider that publication bias is likely to favour the reporting of congenital anomalies among women exposed to efavirenz considering prior concerns; such publication bias would be expected to lead to an overestimation of the risk of efavirenz compared with other antiretroviral drugs. Few studies reported on risk of bias or attempted to control for potential confounders, in particular women on efavirenz may differ from those not on efavirenz in ways that were not reported by the studies. For example, the latter group may include more women who planned their pregnancies and so were more likely to be exposed to protective factors (such as folate supplementation) and reduced risk factors (such as smoking and poor nutrition). Consideration of confounding is all the more important given that it would not be ethically acceptable to conduct a randomized trial to assess risk. Nevertheless, such differences are unlikely to affect our results to an important degree and would be expected to result in an overestimation of the risk of congenital anomalies in the efavirenz group. Finally, future studies should be encouraged to provide full descriptions of the types of birth defects that occur.
The variation in the reporting and difficulty in interpreting available evidence on secondary outcomes including spontaneous abortions, terminations of pregnancy, stillbirths, and preterm deliveries, as well as the still limited prospective data on congenital anomalies compared with larger number of therapeutic exposures confirms the need for better monitoring of these birth outcomes at sentinel sites . The review highlights that medical termination of pregnancy for women exposed to efavirenz has been in the past more frequent than for women not exposed to efavirenz; this is expected to change due to the reassurances provided by recent technical updates and guidelines. The limited reporting of adverse drug reactions means that very little information is available to draw any conclusions regarding potential increased risk of adverse drug reactions associated with efavirenz use in pregnancy, in particular the potential for depression that has been associated with efavirenz use in some , but not all , studies. Future studies should be encouraged to report data on efavirenz safety and tolerability in pregnancy for both the mother and the child.
In conclusion, this updated review does not find any evidence of an increased risk of congenital anomalies in general, or increased risk of neural tube defects, associated with efavirenz exposure during the first trimester of pregnancy and provides the supporting evidence for WHO's 2013 recommendation that efavirenz should be part of the recommended first-line ART regimen, including for women of child-bearing potential, and can be used during first trimester and throughout pregnancy . As with all drugs used in pregnancy, programmes should be encouraged and supported to collect and report on birth outcome data to further assess any potential risk of adverse outcomes [7–9]. Surveillance planning efforts have recently been established in several countries  and such efforts need to be sustained and supported as an increasing number of countries adopt the recommended first-line ART with tenofovir/lamivudine/efavirenz and implement recommendations for provision of ART to all pregnant and breastfeeding women for both prevention of mother-to-child HIV transmission as well as maternal health. The data generated from these efforts to improve data collection and reporting will inform future guidelines on the safety of efavirenz and other antiretrovirals in pregnancy.
We would like to thank Martina Penazatto for comments on an earlier draft, and Heather Watts for assisting with access to data from the Antiretroviral Pregnancy Registry. The review was supported by funds from the Bill and Melinda Gates Foundation. This manuscript represents the views of the authors and does not necessarily represent the view of the National Institutes of Health or Department of Health and Human Services.
NF and LM were the lead authors. NF and ZS scrutinized identified studies for eligibility, extracted data and assessed the methodological quality of included studies. NF performed the statistical analysis. All authors critically reviewed the manuscript before submission.
Conflicts of interest
2. Williams I, Churchill D, Anderson J, Boffito M, Bower M, Cairns G, et al. British HIV Association guidelines for the treatment of HIV-1-positive adults with antiretroviral therapy
. HIV Med
2012; 13 (Suppl 2):1–85.
4. Fundarò C, Genovese O, Rendeli C, Tamburrini E, Salvaggio E. Myelomeningocele in a child with intrauterine exposure to efavirenz
5. De Santis M, Carducci B, De Santis L, Cavaliere AF, Straface G. Periconceptional exposure to efavirenz and neural tube defects
. Arch Intern Med
7. Ford N, Calmy A, Mofenson L. Safety of efavirenz in the first trimester of pregnancy: an updated systematic review and meta-analysis
8. Ford N, Mofenson L, Kranzer K, Medu L, Frigati L, Mills EJ, Calmy A. Safety of efavirenz in first-trimester of pregnancy: a systematic review and meta-analysis of outcomes from observational cohorts
10. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement
. PLoS Med
11. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, Schunemann HJ. Grade: an emerging consensus on rating quality of evidence and strength of recommendations
12. Antiretroviral Pregnancy Registry Steering Committee. Antiretroviral Pregnancy Registry International Interim Report for 1 January 1989 through 31 July 2013
. Wilmington, NC.
13. Freeman M, Tukey J. Transformations related to the angular and the square root
. Ann Inst Stat Mathematics
14. Fleiss JL. The statistical basis of meta-analysis
. Stat Methods Med Res
15. Miller J. The inverse of the freeman-tukey double arcsine transformation
. Am Stat
16. DerSimonian R, Laird N. Meta-analysis in clinical trials
. Control Clin Trials
17. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis
. Stat Med
18. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test
19. Knapp KM, Brogly SB, Muenz DG, Spiegel HM, Conway DH, Scott GB, et al. Prevalence of congenital anomalies in infants with in utero exposure to antiretrovirals
. Pediatr Infect Dis J
20. Shanske AL. Bilateral oblique facial clefts and extremity anomaly in an infant after intrauterine efavirenz exposure and review of its teratogenic risk
21. Floridia M, Mastroiacovo P, Tamburrini E, Tibaldi C, Todros T, Crepaldi A, et al. Birth defects in a national cohort of pregnant women with HIV infection in Italy, 2001–2011
22. Floridia M, Tamburrini E, Ravizza M, Anzidei G, Tibaldi C, Bucceri A, et al. Antiretroviral therapy at conception in pregnant women with HIV in Italy: wide range of variability and frequent exposure to contraindicated drugs
. Antivir Ther
23. Antiretroviral Pregnancy Registry Steering Committee. Antiretroviral Pregnancy Registry International Interim Report for 1 January 1989 through 31 January 2011
. Wilmington, NC.
24. Schwartz S, Taha TE, Venter WD, Mehta S, Rees H, Black V. Efavirenz conceptions and regimen management in a prospective cohort of women on antiretroviral therapy
. Infect Dis Obstet Gynecol
25. Sibiude JEA. Birth defects and ART in the French perinatal cohort, a prospective exhaustive study among 13,124 live births from 1994 to 2010 [
Abstract 81]. 20th Conference on Retroviruses and Opportunistic Infections; 2012; Atlanta, GA.
26. Areechokchai D, Bowonwatanuwong C, Phonrat B, Pitisuttithum P, Maek ANW. Pregnancy outcomes among HIV-infected women undergoing antiretroviral therapy
. Open AIDS J
27. Batallan AMG, Levine M, Longuet P, Bodard M, Legac S, et al. In utero exposure to efavirenz: evaluation in children born alive [
Abstract 1100]. 2nd IAS Conference on HIV Pathogenesis and Treatment; 13–16 July 2003; Paris, France.
28. Bera E, McCausland K, Nonkwelo R, Mgudlwa B, Chacko S, Majeke B. Birth defects following exposure to efavirenz-based antiretroviral therapy during pregnancy: a study at a regional south african hospital
29. Blood E, Beckwith C, Bazerman L, Cu-Uvin S, Mitty J. Pregnancy among HIV-infected refugees in rhode island
. AIDS Care
30. Bussmann H, Wester CW, Wester CN, Lekoko B, Okezie O, Thomas AM, et al. Pregnancy rates and birth outcomes among women on efavirenz-containing highly active antiretroviral therapy in botswana
. J Acquir Immune Defic Syndr
31. Cressey TR, Stek A, Capparelli E, Bowonwatanuwong C, Prommas S, Sirivatanapa P, et al. Efavirenz pharmacokinetics during the third trimester of pregnancy and postpartum
. J Acquir Immune Defic Syndr
32. Ekouevi DK, Coffie PA, Ouattara E, Moh R, Amani-Bosse C, Messou E, et al. Pregnancy outcomes in women exposed to efavirenz and nevirapine: an appraisal of the IeDEA West Africa and ANRS databases, Abidjan, Cote d’Ivoire
. J Acquir Immune Defic Syndr
33. Gonzalez-Tome MF-IM, Ramos Amador J, MunozGalligo E, Rojo C, Nieto O, et al. The Spanish cohort of HIV-infected mother -infant pairs. Efavirenz in pregnancy:, maternal characteristics of women who become pregnant on EFV-containing regimen and neonatal effects [
Abstract CDB0207]. XVII International AIDS Conference, 3–8 August 2008; Mexico City, Mexico.
34. Jeantils V, Khuong MA, Delassus JL, Honore P, Taverne B, Uzan M, Tassi S. [Efavirenz (sustiva) in pregnancy: a study about 12 HIV patients]
. Gynecol Obstet Fertil
35. Joao ECG, Cunha C, Menezes J, Martins E, Medeiros A, et al. Pregnancy outcome in women exposed to efavirenz
[Abstract CDB0698]. XVI International AIDS Conference; 13–18 August 2006; Toronto, ON.
36. Laher FFJ, Mohapi L, Gray G. Efavirenz conceptions in Soweto, South Africa
[Abstract TUPEC047]. 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention; 19–22 July 2009; Cape Town, South Africa.
37. Machado ES, Hofer CB, Costa TT, Nogueira SA, Oliveira RH, Abreu TF, et al. Pregnancy outcome in women infected with HIV-1 receiving combination antiretroviral therapy before versus after conception
. Sex Transm Infect
38. Patel D, Thorne C, Fiore S, Newell ML. Does highly active antiretroviral therapy increase the risk of congenital abnormalities in HIV-infected women?
. J Acquir Immune Defic Syndr
39. Phanuphak NTS, Chinmahun A, Ooprasithiwong J, Mathajittiphun P, Jittarach K, Pima W, et al. Efficacy and safety of maternal triple-drug ARV regimens: Thai Red Cross PMTCT Program, 2004 to 2010
[Abstract 742]. 18th Conference on Retroviruses and Opportunistic Infections; 27 February–2 March 2011; Boston, MA.
40. Rossouw T. Quantifying antiretroviral risk in pregnancy
. S Afr Med J
41. Townsend CL, Willey BA, Cortina-Borja M, Peckham CS, Tookey PA. Antiretroviral therapy and congenital abnormalities in infants born to HIV-infected women in the UK and Ireland, 1990–2007
42. Westreich DRD, MacDonald P, Majuba P, Maskew M, Nagar S, Jaffray I, et al. Pregnancy, efavirenz, and birth outcomes in Johannesburg
[Abstract 922]. 17th Conference on Retroviruses and Opportunistic Infections; 16–19 February 2010; San Francisco, CA.
43. Oliveira M. Pharmacovigilance of antiretrovirals drugs in carrying pregnant women of HIV and children displayed to the HIV during the gestation in use of prophylaxis of the vertical transmission: study incidence pilot
[Masters thesis]. University of Rio de Janeiro, Rio de Janeiro; 2007. 123
44. Joao ECG, Sidi L, Cruz L, Cardoso C, Cunha C, et al. Increased incidence of spontaneous abortion during first trimester exposure to efavirenz
[Abstract TUPEB113]. 4th IAS Conference on HIV, Pathogenesis, Treatment and Prevention; 22–25 July 2007; Sydney, Australia.
45. Taylor GT, Clayden P, Dhar J, Gandhi K, Gilleece Y, Harding K, et al. BHIVA guidelines for the management of HIV infection in pregnant women 2012
. HIV Med
2012; 13 (Suppl 2):87–157.
48. Sterne JA, Sutton AJ, Ioannidis JP, Terrin N, Jones DR, Lau J, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials
49. Watts DH, Mofenson LM. Antiretrovirals in pregnancy: a note of caution
. J Infect Dis
50. Clifford DB, Evans S, Yang Y, Acosta EP, Ribaudo H, Gulick RM. Long-term impact of efavirenz on neuropsychological performance and symptoms in HIV-infected individuals (ACTG 5097S)
. HIV Clin Trials
51. Journot V, Chene G, De Castro N, Rancinan C, Cassuto JP, Allard C, 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