HIV/AIDS
MAJOR ARTICLE
In Utero Exposure to Zidovudine and Heart Anomalies in the ANRS French Perinatal Cohort and the Nested PRIMEVA Randomized Trial Jeanne Sibiude,1,2 Jérôme Le Chenadec,2 Damien Bonnet,3 Roland Tubiana,4 Albert Faye,5,6 Catherine Dollfus,7 Laurent Mandelbrot,1,2,6 Sandrine Delmas,2 Nathalie Lelong,8 Babak Khoshnood,8 Josiane Warszawski,2,9,10 and Stéphane Blanche11,12; for the French National Agency for Research on AIDS and Viral Hepatitis French Perinatal Cohort/Protease Inhibitor Monotherapy Evaluation Triala 1
Department of Obstetrics and Gynecology, Hôpital Louis Mourier, Colombes, 2Department of Epidemiology, Centre de Recherche en Épidémiologie et Santé des Populations, Institut National de la Santé et de la Recherche Médicale (INSERM) U1018, Le Kremlin-Bicêtre, 3M3C-Pediatric Cardiology, Assistance Publique, Hôpitaux de Paris (APHP) Hôpital Necker Enfants malades, Université Paris Descartes, 4Department of Infectiology, AP-HP, GH Pitié-Salpêtrière, Pierre Louis Institute of Epidemiology and Public Health, Sorbonne University, UPMC, INSERM UMR-S1136, 5Department of Pediatrics, AP-HP Hôpital Robert Debré, 6Université Diderot Paris 7, 7Department of Pediatrics AP-HP Hôpital Trousseau, 8 INSERM, UMR S953, Université Paris-6, 9Institut National d’études Démographiques, Paris, 10Université Paris Sud, Le Kremlin-Bicêtre, 11Department of Pediatrics, Hôpital Necker, and 12 EA 7223: Évaluation Thérapeutique et Pharmacologie Périnatale et Pédiatrique, Université Paris Descartes, France
Background. Antiretroviral (ARV) regimens during pregnancy are highly effective in preventing mother-tochild transmission of human immunodeficiency virus (HIV). Congenital heart defects (CHDs) and anomalies in cardiac function have been reported in zidovudine (ZDV)–exposed uninfected children. We explored these associations in a large observational cohort and a randomized clinical trial. Methods. Since 1986, the French Perinatal Cohort prospectively enrolled all HIV-infected women in 90 centers and collected follow-up on their children through 2 years of age. All CHDs were reviewed by a specialist blinded to exposures. Additionally, in a randomized trial (PRIMEVA ANRS 135) of 2 ARV regimens during pregnancy, 1 of which was without nucleoside reverse transcriptase inhibitors, infants had a specific follow-up including echocardiography at 1 month and 12 months. Results. Among 12 888 children included, ZDV exposure in the first trimester was significantly associated with CHD (1.5% vs 0.7%; adjusted odds ratio, 2.2 [95% confidence interval, 1.3–3.7]; P < .001). This association was significant for ventricular septal defects (1.1% vs 0.6%; P = .001) and other CHDs (0.31% vs 0.11%; P = .02). In the randomized trial, among 50 infants, girls (but not boys) exposed in utero to ZDV/lamivudine/ritonavir-boosted lopinavir (LPV/r) had a higher left ventricular shortening fraction at 1 month (40% vs 36%; P = .008), and an increased posterior wall thickness at 1 year (5.4 mm vs 4.4 mm; P = .01) than the LPV/r group. Conclusions. This study confirms a specific association between in utero exposure to ZDV and CHDs, and a longlasting postnatal myocardial remodeling in girls. A potential common mechanism, including the involvement of mitochondrial dysfunction, must be explored, and long-term consequences on cardiac function warrant specific attention. Clinical Trials Registration. NCT00424814. Keywords. PMTCT; congenital heart defects; zidovudine; randomized trial; heart function.
Received 3 December 2014; accepted 24 March 2015; electronically published 1 April 2015. a The members of the ANRS French Perinatal Cohort study group are listed in the Appendix. Correspondence: Jeanne Sibiude, MD, MPH, Equipe VIH/IST CESP 1018 INSERM, 82 rue du Général Leclerc, 94276 Le Kremlin-Bicêtre Cedex, France (
[email protected]). Clinical Infectious Diseases® 2015;61(2):270–80 © The Author 2015. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail:
[email protected]. DOI: 10.1093/cid/civ260
270
•
CID 2015:61 (15 July)
•
HIV/AIDS
The use of antiretroviral (ARV) medications during pregnancy has led to a spectacular decrease in the mother-to-child transmission rate of human immunodeficiency virus type 1 (HIV-1), from about 20% to 700 000 children are exposed worldwide [5].
Birth defects and their potential association with ARVs have been investigated previously, with discrepant results. Most studies did not have the power necessary to evaluate the relation with congenital heart defects (CHDs); however 2 studies found a significant association between CHDs and ARV use [6, 7]. Recently, in a study of >13 000 mother–child pairs in the French Perinatal Cohort (EPF), we observed a strong specific association between zidovudine (ZDV) in utero exposure and CHD incidence [8]. In utero drug toxicity has also been shown to induce various organ dysfunctions, some of which may not be apparent at birth [9]. The potential impact of in utero ZDV exposure on myocardial function among HIV-exposed but uninfected (HEU) children was investigated in successive studies from the same group. The first study on a small number of children concluded that there was no impact on myocardial function [10]. In a second study, Lipshultz et al reported an increased left ventricular (LV) shortening fraction at 4 months of age in HEU children (of whom one-third were exposed in utero to ZDV), compared to controls born to HIV-negative mothers [11]. The same authors showed differences in LV morphology and function parameters in a large cohort of HEU children, not all exposed to ARVs. They reported decreased septal wall thickness, smaller LV dimensions, increased contractility, and increased LV shortening fraction, persisting up to 2 years of age. Interestingly, these anomalies were significantly more marked in girls than in boys [12]. However, in a more recent study comparing once again HEU children to children not exposed in utero to HIV, they failed to show a difference in echocardiographic parameters, but described several differences among individual ARV drugs [13]. The objective of this study was to explore the relationship between perinatal ZDV exposure and structural or functional heart anomalies in children. METHODS
enrollment in the cohort or because of the HIV status of the mother; in particular, echocardiography was only requested in case of detection of a heart murmur or clinical suspicion of cardiac anomaly. The study was approved by the Hôpital Cochin institutional review board and the French computer database watchdog commission. The study population was the same as described in our previous study of all types of congenital defects [8], including all live births exposed to ARVs during pregnancy. Terminations of pregnancy and stillbirths were not included because until recently they were not consistently recorded in EPF. We also excluded patients with unknown dates for starting ZDV or with chromosomal abnormalities. For the present study, all files of children diagnosed with CHD by the pediatrician in the previous EPF study [8] were reviewed by a pediatric cardiologist and, when needed, additional information was collected on-site. When the defect was considered to be nonsignificant, the case was considered to have a normal heart, thus explaining the difference in the number of CHDs between the present study and the previous EPF study [8]. Exposure to ZDV was categorized as (1) unexposed to any ARV in the first trimester, (2) exposed to ZDV with or without other ARVs in the first trimester, or (3) exposed to other ARVs in the first trimester. Based on the embryological chronology of heart formation and on previous literature [6, 15], to specifically study exposure to ZDV in the first trimester, categories (1) and (3) were then grouped together as unexposed to ZDV in the first trimester. Since 1994, the recommended dose for ZDV has been 500 mg daily [16]. Maternal variables included age, geographical origin, intravenous drug use, gravidity, parity, CD4 cell count, and HIV-1 RNA load closest to the delivery. Neonatal variables included sex, gestational age, birth weight, and HIV infection status. Alcohol and tobacco use were recorded only after 2005, and data on concomitant medications and preexisting maternal conditions such as diabetes mellitus were not available in the cohort.
Study of Congenital Heart Defects in EPF
Data Source, Study Population, and Variables The EPF–French National Agency for Research on AIDS and Viral Hepatitis [ANRS] CO1/CO11 study has prospectively enrolled pregnant HIV-infected women delivering in 90 centers in France since 1986, and has been described previously [1]. In participating maternity wards, approximately 95% of all HIVinfected pregnant women gave informed consent and were included. Follow-up was done according to French national guidelines, which include trimestrial fetal ultrasound for all pregnant women. For children exposed to maternal HIV, pediatric examination was done at birth and at 1, 3, 6, 12, and 18–24 months [14], and all clinical events were recorded at each visit. No additional imaging was done systematically because of
Statistical Analysis We described the prevalence of CHDs before and after expert reclassification, and detailed each type of defect. We studied associations between reclassified CHD and ZDV in univariate analysis and multivariate logistic regressions. Multivariate models included all variables associated with CHD in the univariate analysis with a P value
