Efavirenz (EFV) was reclassified by the Food and Drug Administration (FDA) from a category C (‘risk cannot be ruled out’) to a category D (‘evidence of human fetal risk’) drug following case reports of birth defects in infants with first trimester EFV exposure . These defects were similar to those observed in primate studies, in which anencephaly and anophthalmia, micropthalmia, and cleft palate occurred in three of 20 cynomolgus monkeys born after maternal exposure to EFV from day 20 to 150 at plasma concentrations comparable to those in humans. Thus far, three cases of myelomeningocele and two of Dandy–Walker malformation (DWM) have been reported in humans [2–5]. However, without a control group or knowledge of the denominator, the presence or magnitude of the teratogenic risk of EFV is unknown.
Although in these cases, the birth defects noted mirrored some of those found in animal studies, a recent case report of Dandy–Walker variant associated with antenatal nevirapine (NVP) exposure  highlights the poor predictive value of animal toxicology studies translated to humans. No teratogenic effects from NVP were observed during reproductive studies with rats and rabbits .
Pregnant women are generally excluded from experimental drug studies. This makes assessment of drug safety in pregnancy difficult. Clinicians make decisions on a drug's safety informed by prior extensive experience with its use in pregnancy or from prospective registries, the Acyclovir Registry being an example .
Data from the international Antiretroviral Pregnancy Registry (APR) have not demonstrated any significant increase in the prevalence of birth defects following first trimester EFV use in pregnancy. Birth defects were observed in 14 out 477 live births (2.9%), of whom one infant had a myelomeningocele and another had facial cleft and anopthalmia . The number of women exposed to efavirenz-based antiretroviral therapy (EFV-based ART) in the European Collaborative Study, Women and infants’ Transmission Study, and from a study in Botswana were relatively small for analysis of birth defects [10–12]. A comprehensive study of congenital abnormalities in over 8000 infants provided some reassuring data on 205 infants with early pregnancy EFV exposure . Birth defects occurred in five infants (2.4%), with no cases of DWM or myelomeningocele identified.
EFV use in pregnancy has declined considerably internationally, consequent to wider availability and access to different ART regimens. In resource-poor countries, however, ART options and classes of different agents are limited. In South Africa, EFV is prescribed more often than NVP for the treatment of HIV-1-infected adults, because it is safer, easier to monitor, superior on time to viral load suppression, and may be associated with improved survival rates compared with NVP .
With increasing numbers of reproductive-aged women accessing ART services, it is anticipated that an increasing number of them will conceive on EFV-based ART.
Against this background, we sought to determine the prevalence and type of birth defects in women taking EFV-based ART during pregnancy.
Frere Hospital, situated in East London, is a poorly staffed regional hospital that serves as a referral centre for a large geographical area within the Eastern Cape province of South Africa. HIV prevalence among pregnant women in this province is 29% . Approximately 7000 babies are born annually at this institution (unpublished data).
A local Efavirenz in Pregnancy Registry was established in January 2006 to enable prospective follow-up of pregnant women within the region taking EFV-based ART.
For pregnant women with indications for lifelong ART, EFV-based ART was commenced with the following provisos – the pregnancy had progressed beyond 14 weeks gestation, a written undertaking to use contraception postpartum, a detailed ultrasound scan of fetal anatomy was performed before commencing ART, and major psychiatric illness was excluded.
Requirements for lifelong ART included CD4 cell count less than 250 cells/μl or World Health Organization (WHO) stage 4 disease.
The nucleoside reverse transcriptase inhibitors most commonly prescribed with EFV were stavudine (d4T) and lamivudine (3TC). Our local HIV directorate provided authority for the use of zidovudine (ZDV) only when toxicity to d4T occurred, and lopinavir/ritonavir (LPV/r) use was restricted to the first trimester, as a substitute for EFV in women with CD4 cell counts more than 250 cells/μl.
As per protocol, women who conceived on EFV-based ART were given the choice of termination of pregnancy (TOP) up to 20 weeks gestation. For the women who elected to continue with their pregnancies, EFV was substituted in the first trimester for NVP if their CD4 cell count was less than 250 cells/μl or for LPV/r if their CD4 cell count was more than 250 cells/μl. From January 2007, all women in the first trimester who had conceived on EFV were switched to NVP regardless of the CD4 cell count, following data that showed a lower risk of NVP toxicity in ART-experienced women .
Women who presented at or beyond 14 weeks remained on EFV-based ART, unless clinical reasons necessitated a change in treatment. Women who conceived on EFV-based ART were additionally registered with the APR since early 2008 .
Baseline demographic details (age, parity, CD4 cell count, plasma HIV-1 RNA viral load, WHO stage, and gestational age) were obtained from all women. For women who conceived on EFV-based ART, additional information was obtained: duration of EFV use prior to conception, previous miscarriage, stillbirth or neonatal death, plasma viral load at presentation and delivery, concurrent co-trimoxazole (CTX) use, and any history of birth defects. Women who had taken EFV-based ART during the entire first trimester were classified as ‘complete first trimester EFV exposure’. Women in whom EFV was substituted for another drug were classified as ‘partial first trimester EFV exposure’. The first trimester ended at 13 weeks and 6 days.
A detailed fetal anatomy ultrasound scan was done at 18–23 weeks. An amniocentesis/cordocentesis was performed where necessary to further investigate abnormalities found on ultrasound examination.
Following delivery, the neonate was examined for birth defects by a paediatrician or a senior obstetrician. Mothers and infants were followed up until 6 weeks postpartum.
Outcomes measured were the prevalence and type of birth defects. As the follow-up in this cohort was censored at 6 weeks postpartum, case ascertainment was limited to structural defects observed within this period. The prevalence of birth defects was calculated as the ratio of the number of birth defects observed in all pregnancies beyond 24 weeks gestation and the total number of live births. A cohort of at least 200 live births exposed to EFV-based ART was deemed sufficient to detect a two-fold increase in the overall risk of birth defects compared with those observed internationally .
Miscarriage, defined as pregnancy loss prior to 24 weeks , was excluded from this analysis. However, pregnancies that were terminated at any gestation due to identification of fetal abnormality on ultrasound examination were included. Birth defects in twins were counted separately. More than one defect in a singleton was counted as one event.
Data were analysed using Epi-info software, version 3.2.2 (Centre for Disease Control and Prevention, Atlanta, Georgia, USA, 2005). Continuous variables are presented as medians with interquartile ranges (IQRs) and dichotomous data as proportions with 95% confidence intervals (CIs). Approval for this study was obtained from the hospital's research and ethics committee.
Between 1 January 2006 and 31 December 2008, 744 pregnant women initiated EFV-based ART from the second trimester onward. Eighty-nine women were still pregnant and 32 women (4.9%) were lost to follow-up. Over the same period, 220 women conceived on EFV-based ART and another 42 women conceived on NVP-based ART.
Pregnancies were ongoing in 17 women (EFV-based ART) and seven women (NVP-based ART). Eight women (3.9%) and two women (5.7%), respectively, in the two therapies were lost to follow-up.
Data on 851 women with pregnancy outcomes are presented. Table 1 summarizes maternal characteristics, infant outcomes, and the prevalence of birth defects within each group.
Women commenced on efavirenz-based antiretroviral therapy during pregnancy (n = 623)
All women were ART-naive. Their median age was 28 years and the median gestation at ART initiation was 30 weeks. Two hundred and twenty-eight women (37%) were primigravidae. Median CD4 cell count and median plasma viral load were 171 cells/μl and 4.42 log10 copies/ml, respectively. Twenty-six women (4%) had WHO stage 4 disease and 41 women (7%) were concurrently on treatment for tuberculosis (TB).
Median gestation at delivery was 39 weeks. Plasma viral load results were available for 459 women (74%) at delivery, 320 of whom (70%) had viral load less than 50 copies/ml.
Twenty-eight women (5%) had tubal ligation performed at delivery upon request. Sixteen women (2.6%) had twin pregnancies and one had a set of triplets. There were six miscarriages (including a set of triplets) and 10 stillbirths. Of the 623 live births, 319 (51.2%) were male.
Birth defects occurred in 16 of 623 live births, a corresponding prevalence of 2.6% (95% CI 1.5–4.2). The birth defects noted were arachnoid cyst, pulmonary stenosis (unconfirmed), postaxial polydactyly (seven cases), facial asymmetry and overlapping fingers, bilateral clubfeet, congenital naevus, and umbilical hernia (four cases).
The infant with arachnoid cyst benefited from ultrasound diagnosis at the time of ART initiation. A posterior fossa cyst was identified in the fetus at 26 weeks gestation. Following delivery, a computed tomography (CT) scan confirmed the presence of an arachnoid cyst. The newborn with pulmonary stenosis had severe intrapartum asphyxia at birth, was ventilated, and developed congestive cardiac failure. Pulmonary stenosis was strongly suspected, but the baby demised prior to confirmation of the defect. Due to a lack of perinatal postmortem services at our institution, we were unable to confirm or refute the diagnosis of pulmonary stenosis in this newborn. All the infants with postaxial polydactyly had fleshy digits involving their hands, tied off in four infants by means of a suture at the base of the digit.
Women who conceived on efavirenz-based antiretroviral therapy (n = 195)
All pregnancies were unintended. One hundred and eighty-one women (93%) had been on EFV-based ART for longer than 1 month before conception.
Their median age was 30 years and 48 women (25%) were primigravidae. Median CD4 cell count at presentation was 275 cells/μl. Fifty women (26%) were on treatment for TB at the time of initial ART commencement and 37 (19%) had ART initiated as a result of WHO stage 4 disease prior to conception. Sixty-five women (33%) were taking CTX prophylaxis at confirmation of pregnancy. Two women had a family history of birth defects.
Median gestational age at presentation was 19 weeks. Only nine women (5%) presented prior to 6 weeks, whereas 46 women (24%) presented between 6 and 13 weeks and 140 (72%) at/after 14 weeks' gestation.
EFV was substituted with NVP in 55 women (28%) and with LPV/r in eight women. Of 140 women who presented after 14 weeks, 134 had complete first trimester EFV exposure. Six women had already been switched from EFV to NVP at their local ART centres prior to referral.
All but three women with pregnancies less than 20 weeks declined TOP. Two women had ectopic pregnancies, necessitating salpingectomy.
Five women (2.6%) had twin pregnancies. There were six miscarriages (including a set of twins) and four stillbirths (including an in-utero fetal death in a twin pregnancy). Of the 184 live births, 90 (48.9%) were male.
Data on viral load at delivery were available for 131 (72%) of 181 women with live-born infants (including three sets of twins), 103 of those women (79%) had viral load less than 50 copies/ml. Significantly more women who continued EFV throughout pregnancy (75/90, 83%; 95% CI 74–90) had undetectable viral load at delivery compared with women who were switched to NVP or LPV/r (28/41, 68%; 95% CI 52–82) (P = 0.031).
One of the four stillbirths was an intrapartum fetal death at 32 weeks' gestation following induction of labour in a 43-year-old woman. Trisomy 18 was diagnosed in this fetus by ultrasound examination, confirmed on amniotic fluid analysis.
A rare but serious birth defect, arthrogryposis multiplex congenita (AMC), was the attributable cause of an early neonatal death in this group. This neonate was born with severe joint contractures, webbed limbs, pulmonary hypoplasia, absent sacrum, and a unilateral cleft lip and palate. Fetal akinesia and polyhydramnios were the pre-eminent findings on ultrasound examination. Chromosomal studies done on amniotic fluid obtained antenatally were normal.
Birth defects were observed in five of 184 live births and in one of four stillbirths, a prevalence of 3.3% (95% CI 1.2–7.0). The birth defects noted were Trisomy 18, AMC, oesophageal atresia with tracheo-oesophageal fistula (OA-TOF), postaxial polydactyly of fingers and preaxial polydactyly with syndactyly of toes, postaxial polydactyly, and lower central incisor tooth. Details of birth defects and outcomes are shown in Table 2.
No significant difference in the prevalence of birth defects was observed between first and second/third trimester EFV use (prevalence ratio 1.27; 95% CI 0.50–3.20, P = 0.301). The prevalence was similar with maternal CTX use (2/63; 3.2%) compared with no CTX exposure (4/121; 3.3%) (prevalence ratio 0.96; 95% CI 0.18–5.10, P = 0.664). The prevalence of birth defects also did not differ between complete (4/131; 3.1%) and partial (2/53; 3.8%) first trimester EFV exposure (prevalence ratio 0.81; 95% CI 0.15–4.29, P = 0.556).
Table 3 outlines a summary of birth defects by organ system of all pregnancies exposed to EFV-based ART, utilizing the APR system of classification of birth defects .
Women who conceived on nevirapine-based antiretroviral therapy (n = 33)
From this group, one woman had a set of twins and one a miscarriage. A birth defect (umbilical hernia) was observed in one of 33 infants, yielding a prevalence of 3.0% (95% CI 0.1–15.8). There was no significant difference in prevalence of birth defects following conception on EFV compared with conception on NVP (prevalence ratio 1.08; 95% CI 0.13–8.65, P = 0.690).
Table 1 shows a comparison of pregnancy outcomes among women who initiated EFV-based ART, who conceived on EFV-based ART, and who conceived on NVP-based ART.
This is the largest study to-date on EFV-based ART exposure from the second trimester of pregnancy onward. All women were prospectively enrolled and patient retention was high. Previously published cohorts on second/third trimester EFV use were not sufficiently large to detect a two-fold increase in the overall risk of birth defects [9,11,13]. We found that the prevalence in this cohort was 2.6%, which is consistent with between 1.4 and 3.5% from the published data on other classes of commonly used antiretroviral drugs internationally [9–11,13]. The observed birth defects in our study were of questionable relation to EFV, and in the case of the infant with arachnoid cyst, was unrelated to EFV exposure. The commonest defect observed was postaxial polydactyly, which is relatively common in South Africa, with a reported prevalence of 15.49 per 1000 live births . Prevalence estimates of birth defects from the South African health department range from 2.5 to 8% .
Several ART guidelines recommend the use of NVP ahead of EFV after the first trimester, even for women with TB or with CD4 cell counts more than 250 cells/μl [7,21,22].
This recommendation is based on the paucity of infant safety data following second/third trimester EFV exposure, and the potential for birth defects in a future pregnancy with continued EFV use after delivery. At our centre, EFV was preferred, owing to concern over significant drug interactions between NVP and rifampicin in women on TB treatment . At very low CD4 cell counts, NVP appears less effective than EFV . In addition, hepatotoxicity is more frequent, especially in women with CD4 cell counts higher than 250 cells/μl . The results from this cohort are reassuring, and our findings suggest a need to review current recommendations favouring NVP from the second trimester onward.
The prevalence of birth defects following first trimester EFV-based ART exposure was not significantly different from the prevalence after second/third trimester EFV exposure. There were no cases of anophthalmia, neural tube defects, or DWM following first trimester EFV use in this cohort. Although no specific pattern of birth defects was found, the number of exposed infants does not permit any definitive conclusions to be made on the teratogenicity or safety of EFV. Furthermore, the finding of cleft palate and sacral agenesis in the neonate with AMC remains a concern, as these defects may occur following teratogen exposure, which is consistent with those found in EFV-exposed cynomolgus monkeys.
Despite the size limitation of this study, the value of switching from EFV to another drug in the first trimester needs to be revisited. This practice did not meaningfully influence the prevalence of birth defects in this cohort, although our study was not adequately powered to detect this difference. EFV has a long half-life and can be detected in plasma up to 8 weeks later following its discontinuation .
The majority of women (95%) presented at/after 6 weeks' gestation (4 weeks after conception), suggesting EFV use for at least 11 days into the embryonic phase (day 17–56 postconception), which is a critical period of organogenesis.
Switching therapy during or after the embryonic phase may not alter the teratogenic risk and may compromise the woman's ability to achieve undetectable plasma viral load at delivery, as was observed in this cohort.
The prevalence of neural tube defects (NTD) in the United Kingdom is 1.4 per 1000 live births , but in some regions of South Africa, it is estimated at 3.55 per 1000 live births [19,28]. In our setting, follow-up of more than 3000 first trimester EFV-based ART exposures would be required to detect a doubling in the rate of NTD.
It is unlikely that any single centre will amass sufficient number of first trimester EFV exposures to enable the magnitude of its risk to be defined. Healthcare workers caring for women who conceive on ART are, therefore, encouraged to register themselves with the APR antenatally .
In resource-poor countries where ART options are limited, coupled with the high burden of TB , it is likely that EFV will remain an important component of ART. It is equally anticipated that more women will present to antenatal clinics, having conceived on treatment. The declining international use of EFV-based ART in pregnancy has made the establishment of a national South African registry necessary. Such calls have been made previously .
Our study has some limitations. First, a relatively small number of infants were exposed to EFV and the duration of follow-up was limited to 6 weeks. In our routine clinical setting, newborns are generally examined only at birth for defects. Only about a quarter of birth defects diagnosable at 5 years of age are detected at birth .
Hence, long-term outcomes of children exposed to in-utero ART still require evaluation. Second, we lacked a control group for the women initiated on EFV and the comparator group of women who conceived on NVP-based ART was very small.
Third, the risk of birth defects following the use of drugs in combination with EFV was not considered. However, international data on the prevalence of birth defects following ZDV, d4T, 3TC, and LPV/r use in pregnancy are reassuring .
Our data provide some reassurance on second/third trimester EFV use, facilitating the possibility of expanding treatment options for pregnant women, especially those in whom the currently recommended drugs may not be suitable. No conclusions can be drawn from this study on the safety or teratogenicity of EFV following first trimester use, given the limited number of exposed infants. We underscore the importance of establishing national surveillance programs in resource-poor countries for pregnant women taking EFV-based ART.
Professor Gary Maartens from the Department of Medicine, Division of Clinical Pharmacology, University of Cape Town, for critically reviewing this manuscript.
Dr Kim Harper, Dr Felicity Goosen, and Dr Verena Linder from the Department of Paediatrics and Child Health, East London Hospital Complex; Dr Rhoda Bennett, Prof. G.J. Homeyr and Dr Eyob Nigussie from the Fetal Medicine Centre, East London Hospital Complex.
Mrs Nomvula Kwadjo, Dr PM Shweni and Mrs Ndileka Gaba from Saving Mothers Saving Babies (SMSB), Eastern Cape Department of Health, Bhisho. Jennifer Dohrn, Celestine Lottering, Nkosazana Jwacu, Thozeka Mancotywa, Thozama Nkuntayi, Zoe Sofute, Yoliswa Hlati, Ndileka Ngcelwane, Freda Pauls, Bongiwe Luzipho, and Lungiswa Kahlane from mothers-to-mothers (M2M) at Frere Hospital. A sincere thanks to all the pregnant women with HIV infection who agreed to participate.
Author contributions: E.B., K.M., and R.N. prepared the initial draft paper. E.B. performed the statistical analysis; all authors contributed to interpretation of results. All authors including B.Mg., S.C., and B.Ma. submitted comments on all drafts and approved the final version.
Presented as a poster birth defects and perinatal outcomes from 748 women who took efavirenz as part of lifelong antiretroviral therapy during pregnancy: a prospective study.
The study was presented at the 4th South African AIDS Conference, Durban, South Africa, 31 March to 3 April 2009, Abstract PS1-36.
1. Lewis-Hall FC. Important Change in SUSTIVA (efavirenz) Package Insert – Change from Category C to D. Bristol-Myers Squibb Company
; 2005. http://www.fda.gov
. (Accessed 19 April 2008)
2. Bristol-Myers Squibb Company. Efavirenz drug label
3. De Santis M, Carducci B, De Santis L, Cavaliere AF, Straface G. Periconceptional exposure to efavirenz and neural tube defects. Arch Intern Med 2002; 162:355.
4. Fundaro C, Genovese O, Rendeli C, Tamburrini E, Salvaggio E. Myelomeningocele in a child with intrauterine exposure to efavirenz. AIDS 2002; 16:299–300.
5. Saitoh A, Hull AD, Franklin P, Spector SA. Myelomeningocele in an infant with intrauterine exposure to efavirenz. J Perinatol 2005; 25:555–556.
6. Urban MF, Chersich MF. Dandy-Walker variant in an infant prenatally exposed to antiretroviral medication. S Afr Med J 2007; 97:947–948.
7. Public Health Service Task Force. Recommendations for Use of Antiretroviral Drugs in Pregnant HIV-Infected Women for Maternal Health and Interventions to Reduce Perinatal HIV Transmission in the United States
. Supplement: Safety and Toxicity of Individual Antiretroviral Agents in Pregnancy
; 29 April 2009. http://AIDSinfo.nih.gov
. (Accessed 6 July 2009)
8. Reiff-Eldridge R, Heffner CR, Ephross SA, Tennis PS, White AJ, Andrews EB. Monitoring pregnancy outcomes after prenatal drug exposure through prospective pregnancy registries: a pharmaceutical company commitment. Am J Obstet Gynecol 2000; 182:159–163.
9. Antiretroviral Pregnancy Registry Steering Committee. Antiretroviral Pregnancy Registry International Interim Report for 1 January 1989 through 31 January 2009
. Wilmington, NC: Registry Coordinating Center; 2008. www.APRegistry.com
. (Accessed 31 July 2009)
10. 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 2005; 40:116–118.
11. Watts HD, Li D, Handelsman E, Tilson H, Paul M, Foca M, et al
. Assessment of birth defects according to maternal therapy among infants in the women and infants transmission study. J Acquir Immune Defic Syndr 2007; 44:299–305.
12. Bussmann H, Wester WC, 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 2007; 45:269–273.
13. 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. AIDS 2009; 23:519–524.
14. Nachega JB, Hislop M, Dowdy DW, Gallant JE, Chaisson RE, Regensberg L, et al
. Efavirenz versus nevirapine-based initial treatment of HIV infection: clinical and virological outcomes in Southern African adults. AIDS 2008; 22:2117–2125.
15. National Department of Health. National HIV and Syphilis Antenatal Seroprevalence Survey in South Africa 2007
. Pretoria, South Africa: Directorate: Health Systems Research, Research Coordination and Epidemiology, Department of Health; 2008. http://www.doh.gov.za
. (Accessed 20 April 2008)
16. De Lazzari E, Leon A, Arniaz JA, Martinez E, Knobel H, Negredo E, et al. Risk of hepatotoxicity in virologically suppressed HIV patients switching to nevirapine according to gender and CD4 count.
46th ICAAC, San Francisco, CA. September 27–30, 2006. Abstract H-1064.
17. Watts HD, Covington DL, Beckerman K, Garcia P, Scheuerle A, Dominguez K, et al
. Assessing the risk of birth defects associated with antiretroviral exposure during pregnancy. Am J Obstet Gynecol 2004; 191:985–992.
18. Royal College of Obstetricians and Gynaecologists. The management of early pregnancy loss.
Guideline No. 25. London: RCOG; 2006.
19. Venter PA, Christianson AL, Gericke GS, Hutamo CM, Makhura MP. Congenital anomalies in rural black South African neonates: a silent epidemic? S Afr Med J 1995; 85:15–20.
20. Human Genetics Policy Guidelines for the Management and Prevention of Genetic Disorders, Birth Defects and Disabilities
. Pretoria, South Africa: Sub-Directorate, Human Genetics, National Department of Health; 2008. www.doh.gov.za
. (Accessed 29 December 2008)
21. National Department of Health. National Antiretroviral Treatment Guidelines
. 1st ed. Pretoria, South Africa: Jacana; 2004.
22. World Health Organization. Antiretroviral drugs for treating pregnant women and preventing HIV infection in infants: towards universal access
. Geneva, Switzerland: WHO; 2006.
23. Back D, Gibbons S, Khoo S. Pharmacokinetic Drug Interactions with Nevirapine. J Acquir Immune Defic Syndr 2003; 34(Suppl 1):S08–S14.
24. van Leth F, Andrews S, Grinsztejn B, Wilkins E, Lazanas MK, Lange JMA, et al
. The effect of baseline CD4 cell count and HIV-1 viral load on the efficacy and safety of nevirapine or efavirenz-based first-line HAART. AIDS 2005; 19:463–471.
25. Baylor MS, Johann-Liang R. Hepatotoxicity associated with nevirapine use. J Acquir Immune Defic Syndr 2004; 35:538–539.
26. Tariq Sadiq S, Fredericks S, Khoo SH, Rice P, Holt DW. Efavirenz detectable in plasma 8 weeks after stopping therapy and subsequent development of nonnucleoside reverse transcriptase inhibitor-associated resistance. AIDS 2005; 19:1716–1717.
27. Rankin J, Pattenden S, Abramsky L, Boyd P, Jordan H, Stone D, et al
. Prevalence of congenital anomalies in five British regions, 1991–99. Arch Dis Child Fetal Neonatol Ed 2005; 90:F374–F379.
28. Robertson H-L, Steyn NP, Venter PA, Christianson AL. Neural tube defects and folic acid: a South African perspective
. S Afr Med J
29. Lawn SD, Myer L, Bekker LG, Wood R. Burden of tuberculosis in an antiretroviral treatment programme in sub-Saharan Africa: impact on treatment outcomes and implications for tuberculosis control. AIDS 2006; 20:1605–1612.
30. Rossouw T. Quantifying antiretroviral risk in pregnancy. S Afr Med J 2007; 97:1014–1016.