|Home | About | Journals | Submit | Contact Us | Français|
We evaluated the association between maternal antiretrovirals (ARVs) during pregnancy and infant congenital anomalies (CAs), utilizing data from the NISDI Perinatal Study.
The study population consisted of first singleton pregnancies on study, ≥ 20 weeks gestation, among women enrolled in NISDI from Argentina and Brazil who delivered between September 2002 and October 2007. CAs were defined as any major structural or chromosomal abnormality, or a cluster of two or more minor abnormalities, according to the conventions of the Antiretroviral Pregnancy Registry. CAs were identified from fetal ultrasound, study visit, and death reports. The conventions of the Antiretroviral Pregnancy Registry were used. Prevalence rates [number of CAs per 100 live births (LBs)] were calculated for specific ARVs, classes of ARVs, and overall exposure to ARVs.
Of 1229 women enrolled, 995 pregnancy outcomes (974 LBs) met the inclusion criteria. Of these, 60 infants (59 LBs and 1 stillbirth) had at least one CA. The overall prevalence of CAs (per 100 LBs) was 6.2 (95%CI = 4.6, 7.7). The prevalence of CAs after first trimester ARVs (6.2; 95%CI = 3.1, 9.3) was similar to that after second (6.8; 95%CI = 4.5, 9.0) or third trimester (4.3; 95%CI = 1.5, 7.2) exposure. The rate of CAs identified within seven days of delivery was 2.36 (95%CI: 1.4–3.3).
The prevalence of CAs following first trimester exposure to ARVs was similar to that following second or third trimester exposure. Continued surveillance for CAs among children exposed to ARVs during gestation is needed.
The increasing complexity of antiretroviral (ARV) regimens used during pregnancy for treatment of HIV-1-infected women and for prevention of mother-to-child transmission of HIV-1 raises concerns regarding potential ARV-related adverse events such as low birth weight, preterm birth, stillbirth, and teratogenicity.1–5 Also, many HIV-1-infected women use other drugs (in addition to ARVs) during pregnancy with potentially teratogenic effects, such as trimethoprim/sulfamethoxazole prescribed for Pneumocystis jiroveci pneumonia prophylaxis.
The estimated proportion of children in Latin America with congenital anomalies (CAs) ranges from 0.4% to 8.4%.6–11 There are few data addressing the relationship between maternal ARVs during pregnancy and infant CAs in Latin America. Therefore, we analyzed data from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) International Site Development Initiative (NISDI) Perinatal Study, a prospective cohort study of HIV-1-infected women and their children conducted at multiple sites in Latin America and the Caribbean, in order to determine the prevalence of CAs in this population, overall and according to in utero exposure to ARVs.
The NISDI Perinatal Study is a prospective cohort study conducted at Latin American and Caribbean clinical sites where, at a minimum, ARV prophylaxis and alternatives to breastfeeding are available. The primary objectives include characterizing adverse events among infants born to HIV-1-infected women.12 Enrollment began in September 2002, and is ongoing. Prior to enrollment, all women provide signed informed consent for enrollment of themselves and their infants. Infant study visits are conducted before hospital discharge after birth, at 6–12 weeks, and six months of age. During each of these study visits, the infant’s parent or guardian is interviewed, a physical examination is conducted, and laboratory samples are obtained. The protocol was approved by the ethical review boards of each clinical site enrolling subjects, the sponsoring institution (NICHD), and the data management and statistical center (Westat).
Clinical, immunologic, and virologic characteristics of the women are assessed during pregnancy, at the time of hospital discharge after delivery, and at the 6–12 week postpartum visit. Maternal clinical disease staging13 is performed at each study visit. A maternal history of substance use during the index pregnancy is ascertained through maternal interview at enrollment. Infant gestational age at birth (in completed weeks) is determined either by obstetric estimation (dates of last menstrual period with or without ultrasonography)or by pediatric newborn examination (Ballard et al14, Dubowitzet al15 or Capurro et al16).
The study population was restricted to women enrolled in the NISDI Perinatal Protocol as of October 2007 for the first time (second pregnancies on-study excluded) who delivered a singleton infant (live born or stillborn) ≥ 20 weeks gestation. Additionally, the study population was restricted to infants of mothers from Brazil and Argentina, since the great majority of subjects were enrolled in these two countries.
ARV regimens were determined by the subject’s clinician, independent of participation in the NISDI Perinatal Study. Both Argentinean and Brazilian guidelines for the use of antiretrovirals during pregnancy17–18 indicate that all HIV-infected women should be evaluated for the need of ARV treatment during pregnancy. ARV treatment should be continued or initiated according to the HIV disease stage of the pregnant woman. For those women who do not require ARV treatment, ARV prophylaxis should be administered throughout pregnancy and continued intrapartum. If prophylaxis is not initiated during pregnancy, intrapartum zidovudine should be administered intravenously.
We first described the overall receipt of ARVs during pregnancy. ARV regimens received during pregnancy were categorized in the following manner: none; one or two nucleoside or nucleotide analogue reverse transcriptase inhibitors (NRTIs; two NRTIs with one non nucleoside reverse transcriptase inhibitor (NNRTI) (HAART-NNRTI); two NRTIs with one protease inhibitor (PI) (HAART-PI), and other. If two HAART regimens were received for 28 days or more during pregnancy, the regimen received later in pregnancy took precedence overa regimen received earlier during pregnancy.
We analyzed the relationship between ARV use during the first trimester of pregnancy and CAs, since this is the trimester of fetal organogenesis. For these analyses, we focused on the earliest regimen used for at least 28 days during the first trimester of pregnancy and specific ARVs used during the first trimester.
A CA was defined as any major structural or chromosomal abnormality, or any cluster of two or more minor (conditional) abnormalities, occurring in infants or fetuses of at least 20 weeks gestational age based on the same criteria as used by the Antiretroviral Pregnancy Registry (APR)19 and the Metropolitan Atlanta Congenital Defects Project (MACDP).20 CAs were identified through fetal ultrasound, study visit, and death reports up to the child’s six month follow-up visit. Infants with multiple congenital anomalies were counted as a single outcome for analysis, and CAs were grouped according to organ system21 for reporting.
In addition to the primary exposure and outcomes variables of interest (maternal ARV use during pregnancy, especially during the first trimester, and infant CAs), we examined other variables in relation to the presence or absence of CAs. These variables were: maternal demographic characteristics (marital status; country of residence; age; years of formal education; gainful employment outside of the home); maternal HIV disease stage (CD4 count and CD4 percentage; plasma HIV-1 RNA concentration; clinical disease stage – both at enrollment and prior to delivery); obstetrical characteristics (gravidity; parity), other medical characteristics (body mass index; toxoplasmosis, rubella, cytomegalovirus, Herpes simplex virus, and syphilis during pregnancy; sexually transmitted infections; diabetes; hypertension; infectious and non-infectious renal disease during pregnancy); maternal alcohol, tobacco, and illicit drug use during pregnancy; folate antagonist and other class D drug exposures during pregnancy; and folic acid supplementation during pregnancy.
The prevalence of CAs was calculated by dividing the number of CAs reported among pregnancy outcomes ≥ 20 weeks of gestation (stillborn and live born) by the total number of live births (LBs). Prevalence of CAs were calculated for those ARV exposures with more than 200 mother-infant pairs, following the APR convention.19 Ninety-five percent confidence intervals (95%CIs) were calculated using a Poisson distribution. Fisher’s exact test was used to assess associations between CAs and study characteristics, in particular ARV and folate antagonist use during pregnancy. Variables at least marginally associated with CA (p < 0.20) were considered candidates for multivariable logistic regression modeling. SAS software (Cary, NC) was used for all analyses.
As of October 2007, 1229 women had enrolled in the NISDI Perinatal Study, of whom 1079 (87.8%) were enrolled in Argentina or Brazil. Of these enrollments, two women were not followed to delivery (one died, one was lost to follow-up). Of the remaining 1077 enrollments, 55 (5.1%) represented the second (or greater) pregnancy on study, and were excluded. Of the 1022 first pregnancies on study, 21 (2.1%) multiple gestation pregnancies were excluded. Of the remaining 1001 women with singleton outcomes, six (0.6%) experienced pregnancy losses before 20 weeks gestation, and were excluded. Among the remaining 995 pregnancy outcomes ≥ 20 weeks gestation, there were 974 (97.9%) LBs, one (0.1%) therapeutic abortion, and 20 (2.0%) stillbirths.
Overall, 988 (99.3%) of the 995 women in the study population received one or more ARVs during pregnancy, with 249 (25.0%) receiving at least one ARV during the first trimester of pregnancy. The seven women categorized as not receiving ARVs during pregnancy included four women who received ARVs only at the time of delivery and three who had no record of receiving ARVs during the index pregnancy. Six of the seven women had a live birth.
CAs were detected among 60 infants (59 LBs and one stillbirth) in utero or postnatally through the six month study visit, for an overall prevalence of 6.16 per 100 LBs (95%CI = 4.60–7.72). Excluding infants with only minor (conditional) anomalies, the prevalence was 5.75 per 100 LBs (95%CI = 4.24–7.24). The prevalence of CAs detected within the first seven days of life only was 2.36 per 100 LBs (95%CI = 1.40–3.33), including two infants with patent foramen ovale (PFO). There were no infants included only because of minor (conditional) anomalies. Of the 60 infants with CAs, 41 had a single anomaly, 16 had two anomalies, and three had multiple anomalies. The specific CAs detected, grouped by organ system and first ARV exposure, are listed in Table 1. As shown in this table, the cardiovascular and musculoskeletal systems were most often affected.
Characteristics of the study population, overall and according to the presence or absence of CAs, are shown in Tables 2 and and3.3. Of the covariates examined, only marital status of the mother (P = 0.01) was associated with CAs. The remaining variables were not associated with CAs, including: use of ARVs at conception or at enrollment; the number of ARV regimens used for 28 days or more during pregnancy, the first ARV regimen used for 28 days during the first trimester and the last ARV regimen used for 28 days or more during pregnancy. There was no statistically significant association between the trimester of first ARV exposure and infant CAs (P = 0.52). Multivariable analyses were not performed since the main exposures of interest were not statistically associated with infant CAs, and did not meet our minimal entry criteria of P < 0.20 for multivariable modeling.
The prevalence of CAs among HIV-1-infected women who first received ARVs at the time of conception or during the first trimester was 15 of 242 LBs (6.20 per 100 LBs; 95% CI = 3.06–9.34); second trimester exposure: 35 of 518 LBs (6.76 per 100 LBs, 95% CI = 4.52–9.00); third trimester exposure: nine of 208 LBs (4.33 per 100 LBs, 95% CI = 1.50–7.15); and, no ARV exposure during pregnancy: one of 6 LBs (16.67 per 100 LBs, 95%CI = 0–49.33) (Table 4). Prevalence of CAs according to ARV class, specific ARVs, and any ARV exposure, by trimester, are shown in Table 5. The 95% confidence interval estimates for the prevalence rates generally overlap, suggesting that there are no significant differences in the prevalence of CAs according to ARV class or specific ARVs.
In this study of HIV-1-infected women and their infants in Argentina and Brazil, the overall prevalence of CAs was 6.16/100 LBs. The prevalence of CAs following first trimester exposure to ARVs (6.20/100 LBs) did not appear significantly different from that following second (6.76/100 LBs) or third trimester (4.33/100 LBs) exposure. In addition, the prevalence of CAs did not appear to differ significantly according to ARV class, specific ARVs, and any ARV exposure. Marital status, associated with CAs in univariate analysis, presumably is a proxy for other environmental and/or socioeconomic factors.
The prevalence of CAs within seven days of birth (2.4%) in our study population is similar to that at delivery (2.8%) reported by the Latin American Collaborative Study of Congenital Malformations (ECLAMC)10, a hospital-based program for the clinical and epidemiological investigation of CAs. ECLAMC reports on CAs among approximately 200,000 births per year, including both major and minor anomalies diagnosed prior to hospital discharge after birth for infants weighing 500 grams or more. Participating countries include Argentina, Brazil, Chile, Ecuador, Peru, and Venezuela.
The overall prevalence of CAs in our study population is within the range reported among the general population in Latin America. Very low rates have been reported in studies utilizing birth certificate data or medical records: 0.4% in Vitoria, Brazil6, 0.8% in Rio de Janeiro7, and 1.4% in Pelotas, Brazil8. In relatively small, retrospective cohort studies of HIV-1-infected women, the observed prevalence of CAs was 2.3% in Buenos Aires, Argentina22 and 2.2% in Rio de Janeiro, Brazil.23 Other studies have reported CA rates ranging from 1.7% in Rio de Janeiro, Brazil9; 2.2% in Argentina (personal communication, Silvina Ivalo); 4.7% in Brazil24; and 8.4% in Chile.11
The overall prevalence of CAs in our study population is higher than that reported in other, large studies of HIV-1-infected women and their infants19, 25–27, and the higher prevalence of CAs observed in our study could be attributed to its prospective design, with follow-up of infants until six months after birth and with reporting of both minor and major anomalies. However, the prevalence of CAs detected within the first seven days of life (2.36/100 LBs) is similar. Most importantly, our results showed no difference in the prevalence of CAs according to trimester of exposure to ARVs, consistent with previous studies of in utero ARV exposure and infant CAs.
One of the largest sources of data regarding in utero ARV exposure and CAs is the APR (130 CAs among 4530 LBs, or 2.9 CAs/100 LBs)19, a voluntary registry, including data from the U.S. and other countries, with prospective assessment of exposure. The great majority of data utilized within the APR are collected at the time of birth or shortly thereafter. However, in some cases data collected through the first 1–6 years after birth are used for categorizing infant CAs. For the overall population exposed to ARVs in this registry, no increase in risk of either CAs overall or specific CAs has been detected to date when compared with observed prevalence for “early diagnoses” in population-based birth defects surveillance systems or with prevalence among those with earliest ARV exposure in the second trimester or third trimester. In analyzing individual drugs with sufficient data to warrant separate analyses, an increased frequency for CAs has been detected for didanosine, but without an increase in the prevalence of a specific anomaly.19 Analyses of data collected within the first 18 months of life for HIV-exposed infants enrolled in the Women and Infants Transmission Study (WITS) (90 CAs among 2527 LBs, or 3.6 CAs/100 LBs) revealed a statistically significant elevated rate of hypospadias after first trimester exposure to zidovudine.25 For the National Study of HIV in Pregnancy and Childhood (NSDHPC) in the United Kingdom and Ireland, most reports of CAs are collected with the first few weeks of life.26 The observed rate of CAs was 232 CAs per 8242 LBs, or 2.8%.26 No increased risk of CAs after in utero exposure to specific ARVs or different classes of ARVs was observed in this study26, nor in the European Collaborative Study (55 CAs among 3740 children, or 1.5%).27
HIV-1-infected women may take other potentially teratogenic drugs besides ARVs. Dihydrofolate reductase inhibitors, such as trimethoprim, pyrimethamine, and sulfadiazine, and other folate antagonists, such as carbamazepine, phenytoin, and phenobarbital, may increase the risk of neural tube defects as well as cardiovascular, oral clefts, urinary tract and limb-reduction defects.28–30 A retrospective study concluded that there was no evidence of teratogenicity of ARVs if given alone during the first trimester, but exposure to the combination of ARVs and folate antagonists (n=13) was associated with a significantly higher risk of CA when compared to no exposure (n=198). 31 In the present study, CAs were not associated with exposure to ARVs alone or in combination with folate antagonists. This could be partially explained because 86% of women in the study population were asymptomatic and because of food fortification policies (extra synthetic folic acid in wheat flour) established in several South American countries in recent years.32
Our observed stillbirth rate of 2% is consistent with the rate observed among 859,750 Latin American hospital births from 1982–1986 (2.0%).33 Our 2% stillbirth rate also is consistent with published stillbirth rates among HIV-infected mothers in the US (2%)34 and in Argentina (1.7%).22
This was the first large prospective study in Latin America to address the prevalence of CAs among HIV-1-exposed infants, and to accurately collect data regarding not only in utero exposure to ARVs, but also exposure to folic acid and folate antagonists. However, despite the large sample size, there was limited power to pursue detailed analyses of the prevalence of CAs according to specific ARV exposures, and to analyze specific types of CAs. The results of our analyses do not support changes to current recommendations for the use of ARVs during pregnancy for treatment of HIV-1-infected women and for prevention of mother to child transmission. Continued monitoring of the prevalence of CAs among children of HIV-1-infected women should be pursued.
Source of Financial Support: NICHD Contract # N01-HD-3-3345 and # HHSN267200800001C (NICHD Control # N01-DK-8-0001).
*Principal investigators, co-principal investigators, study coordinators, coordinating center representatives, and NICHD staff include: Argentina: Buenos Aires: Marcelo H. Losso, Adriana S. Durán, Silvina Ivalo (Hospital General de Agudos José María Ramos Mejía); Brazil: Belo Horizonte: Jorge Pinto, Victor Melo, Fabiana Kakehasi (Universidade Federal de Minas Gerais); Caxias do Sul: Ricardo da Silva de Souza, Nicole Golin, Machline Paim Paganella (Universidade de Caxias do Sul/Secretaria Municipal de DST/AIDS de Caxias do Sul - Ambulatorio Municipal DST/AIDS); Nova Iguaçu: Jose Pilotto, Beatriz Grinsztejn, Valdilea Veloso (Hospital Geral Nova de Iguaçu Setor de DST/AIDS; Porto Alegre: Ricardo da Silva de Souza, Breno Riegel Santos, Rita de Cassia Alves Lira (Universidade de Caxias do Sul/Hospital Conceição); Ricardo da Silva de Souza, Mario Peixoto, Marcelo Almeida (Universidade de Caxias do Sul/Hospital Fêmina); Regis Kreitchman, Debora Coelho Fernandes (Irmandade da Santa Casa de Misericórdia de Porto Alegre); Ribeirão Preto: Marisa M. Mussi-Pinhata, Geraldo Duarte, Carolina Sales V. Macedo, Maria A. do Carmo Rego (Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto da Universidade de São Paulo); Rio de Janeiro: Ricardo Hugo S. Oliveira, Elizabeth S. Machado, Maria C. Chermont Sapia (Instituto de Puericultura e Pediatria Martagão Gesteira); Esau Custodio Joao, Leon Claude Sidi, Guilherme Amaral Calvet, Claudete Araújo Cardoso (Hospital dos Servidores do Estado); São Paulo: Regina Celia de Menezes Succi, Prescilla Chow Lindsey (Federal University of São Paulo); Peru: Lima: Jorge Alarcon (Instituto de Medicina Tropical “Daniel Alcides Carrion”-Division de Epidemiología), Carlos Velásquez Vásquez (Instituto Materno Perinatal), César Gutiérrez Villafuerte (Instituto de Medicina Tropical “Daniel Alcides Carrion”-Division de Epidemiología); Data Management and Statistical Center: Yolanda Bertucci, Laura Freimanis Hance, René Gonin, D. Robert Harris, Roslyn Hennessey, James Korelitz, Margot Krauss, Sharon Sothern, Sonia K. Stoszek (Westat, Rockville, MD, USA); NICHD: Rohan Hazra, Lynne Mofenson, Jack Moye, Jennifer S. Read, Heather Watts, Carol Worrell (Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland, USA). Supported by NICHD Contract # HHSN267200800001C (NICHD Control # N01-DK-8-0001)
Presented in part: at the 48th Annual ICAAC/IDSA 46th Annual Meeting; Washington, DC; October 25–28, 2008 [abstract 964]
Conflict of interest: There are no conflicts of interest to disclose.