Roberts, Susan S PhD*†; Martinez, Marisol MD‡; Covington, Deborah L Dr PH†; Rode, Richard A PhD‡; Pasley, Mary V RNC-NP‡; Woodward, William C DO‡
From the *Clinical Research Department, University of North Carolina Wilmington, Wilmington, NC; †Registries and Epidemiology, Kendle International Inc, Wilmington, NC; and ‡Abbott Laboratories, Abbott Park, IL.
Received for publication October 23, 2008; accepted January 29, 2009.
Supported by an unrestricted grant from Abbott Laboratories to Kendle International Inc to develop, organize, and prepare information related to this publication. The research did not receive funding from any of the following organizations: National Institutes of Health, Wellcome Trust, Howard Hughes Medical Institute, or other(s).
Presented as a poster at the XVII International AIDS Conference (AIDS 2008). August 3-8, 2008, Mexico City, Mexico.
Correspondence to: Susan Sinclair Roberts, PhD, University of North Carolina Wilmington, 601 South College Road, Wilmington, NC 28403-5995 (e-mail: firstname.lastname@example.org).
Antiretroviral therapy is recommended for use in HIV-infected pregnant women to reduce maternal HIV RNA levels and decrease vertical transmission.1 The Antiretroviral Pregnancy Registry (APR) was implemented in 1989 to evaluate the risk of major teratogenic effects associated with antiretroviral therapy.2 Enrollment in the APR has met the minimal sample size needed to assess risk for 11 individual drugs among the 33 drugs monitored.3 Recently, lopinavir/ritonavir reached this enrollment threshold. Lopinavir/ritonavir is currently the preferred protease inhibitor recommended by the US Department of Health and Human Services for use during pregnancy.4 Human teratogenicity studies of lopinavir/ritonavir have not been published to date. To ensure the safety of pregnant women and their offspring, continuous clinical and safety data are needed to support current guidelines.
Lopinavir/ritonavir (Kaletra, Aluvia; Abbott Laboratories, Abbott Park, IL) is approved for use in combination antiretroviral therapy for HIV infection in adults and children.5 As coformulated, ritonavir inhibits the cytochrome P3A-mediated metabolism of lopinavir providing increased plasma levels of lopinavir.5 Lopinavir/ritonavir is assigned Food and Drug Administration (FDA) Pregnancy Category C indicating risk to pregnancy cannot be ruled out; adequately powered human studies are lacking; and animal studies are either positive for fetal risk or unavailable.5,6 Ritonavir is designated a Pregnancy Category B indicating animal studies have not demonstrated a risk to the fetus, although adequate well-controlled studies in human pregnancy are unavailable.6,7 In animal studies of lopinavir/ritonavir, no treatment-related malformations were reported, however, embryonic developmental toxicities occurred in rats receiving maternally toxic doses.5 Using data from the APR, this study estimates the risk for birth defects after pregnancy exposures to lopinavir/ritonavir. In addition, prematurity and other negative birth outcomes are examined.
The analysis population includes all reported pregnancy exposures to lopinavir/ritonavir, with complete data on exposure and birth outcome, enrolled in the APR from September 2000, at the time of FDA approval, through July 31, 2007.
Data were extracted from the APR, an international prospective exposure-registration cohort study of women exposed to antiretroviral (ARV) drugs during pregnancy. Exposed women are voluntarily and anonymously enrolled by their health care providers. Detailed methodology of the APR is published elsewhere.2,3 The APR has Institutional Review Board approval, which includes a waiver of informed consent based on the APR's process for protecting patient anonymity.2,3
Key variables were extracted from the APR for all lopinavir/ritonavir-exposed pregnancies: pregnancy outcome, infant birth weight, gestational age, earliest trimester of lopinavir/ritonavir exposure, and birth defects. Pregnancy outcomes are assigned to one of the following mutually exclusive categories: live birth, spontaneous loss, induced abortion, and stillbirth (ie, delivery of a deceased fetus after 20 weeks gestation or weighing more than 500 g). The analyses of prematurity and low birth weight are restricted to singleton live births without birth defects because of the strong association between prematurity and both multiple gestation and congenital anomalies.8
A birth defect is defined as any major structural or chromosomal abnormality (ie, defects that are incompatible with life or requiring major surgical intervention) or any cluster of 2 or more minor abnormalities (ie, minor physical defects with no obvious medical, surgical, or cosmetic consequences) occurring in infants or fetuses of at least 20 weeks gestation.2,3 In the external comparator, the Metropolitan Atlanta Congenital Defects Program (MACDP), the definition includes minor defects only in the presence of a major defect.2,3,9 Pregnancy exposures are expressed as the trimester of earliest exposure (ie, gestational age corresponding to the start date of lopinavir/ritonavir).2,3
This analysis follows the recommendations set forth from the FDA; both internal and external comparison groups are used.10 First, the prevalence of birth defects among lopinavir/ritonavir-exposed women is compared with the MACDP. The MACDP actively searches for birth defects among the 50,000 annual births to residents of metropolitan Atlanta's 5 counties and abstracts medical records at all associated Atlanta hospitals, genetics laboratories, and vital records.9 The total prevalence of birth defects in this population was 2.67 per 100 live births; because MACDP is population-based and does not involve sampling, confidence intervals (CIs) are not provided.11 In the second comparison group, first-trimester exposures are compared with combined second/third-trimester lopinavir-ritonavir exposures. Because most structural defects have their origins in the first-trimester, second/third-trimester exposure presents a group that is similar in important covariates but considerably different in exposure.12 Fetal development progresses considerably by the end of the first trimester; therefore, second and third-trimester exposures are grouped together.12 Consistent with the MACDP, the prevalence of birth defects is calculated by dividing the number of infants with birth defects (live births + stillbirths) by the total number of live births.9,11
CIs, constructed around birth defect rates, are based on exact confidence limits for the binomial proportion.13,14 Risk ratios are calculated using methods suggested by Kleinbaum et al.15 CIs for the risk ratio of birth defects for first-trimester exposures relative to second/third trimester exposures are calculated using methods for asymptotic CIs for relative risk.14 Data analysis was performed using SAS software, Version 9.1.3 of the SAS System for Windows XP (SAS Institute Inc, Cary, NC).
Based on sample size calculations, a cohort of 267 live-born infants with first-trimester exposure to lopinavir/ritonavir provides approximately 80% power to detect a 2.4-fold increase in the risk of birth defects (overall) compared with the MACDP birth defect rate of 2.67; this calculation is based on a 2-sided binomial test for a 1-proportion inequality with a Type I error rate of 5%.3 For individual birth defects, sample size requirements are greater; the power to detect an increased risk for a specific defect varies and depends upon the population frequency of the defect. For the internal comparison group, sample sizes of 267 and 688 live-born infants with first-trimester and second/third-trimester lopinavir/ritonavir exposures, respectively, provides approximately 80% power to detect a 2.6-fold increase in the risk of birth defects. This retrospective power calculation is based on a 2-sided normal approximation to the binomial distribution for 2 independent proportions with a Type I error rate of 5% and a population proportion of the comparator group of 2.62%.3
Through July 31, 2007, the APR received 7255 complete prospective reports of pregnancy exposures to 1 or more ARV drugs; an additional 835 (9.8%) were lost to follow-up, and 392 (4.6%) were pending birth outcome. Among the 7255 reports, there were 1051 reports of exposures to lopinavir/ritonavir, of which 987 (93.9%) had complete follow-up and 64 (6.1%) were lost to follow-up (Table 1). Lopinavir/ritonavir-exposed pregnancies reported to the APR represent nearly one fourth of all enrolled pregnancies; 85% were enrolled from within the United States.
Of the 987 lopinavir/ritonavir-exposed pregnancies, there were 1006 pregnancy outcomes. The trimester of exposure was unspecified in 1 pregnancy that resulted in a singleton live birth without birth defects. Most of the remaining 986 pregnancies (n = 968, 98.2%) were singleton pregnancies resulting in 968 outcomes: 920 live births, 20 spontaneous losses, 19 induced abortions, and 9 stillbirths. Multiple pregnancies represented 1.8% of all evaluable pregnancies and resulted in 35 live births and 2 stillbirths.
Birth outcomes are summarized in Table 2. There were no birth defects reported from induced or spontaneous losses; this is common due to early gestation and the mechanisms behind these events. The majority of reported lopinavir/ritonavir-exposed pregnancies received combination antiretroviral therapy (Table 2). Birth defects were reported in 22 of the 955 live births and in 1 of the 11 stillbirths for a prevalence of 2.4% (95% CI = 1.5 to 3.6) (Table 3). Of the 267 live births exposed to lopinavir/ritonavir in the first trimester, there were 5 outcomes with birth defects (1.9%; 95% CI = 0.6 to 4.3); all were live births. These proportions are similar to the MACDP prevalence of 2.67% overall and 2.09% for defects diagnosed before the seventh day after birth.11 Of the 688 live births exposed to lopinavir/ritonavir in the second/third-trimester group, there were 18 outcomes with birth defects; one of which was in a still-born infant (2.6%; 95% CI = 1.6 to 4.1). The prevalence of birth defects among the offspring of women with first-trimester lopinavir/ritonavir exposures is not significantly different from the prevalence of second/third-trimester exposures (risk ratio = 0.72, 95% CI = 0.27 to 1.91); however, the width of the CI suggests low statistical power for making this comparison. All offspring with birth defects were exposed to combination antiretroviral therapy. Among the 23 lopinavir/ritonavir-exposed outcomes with reported birth defects, there was no pattern suggesting a common etiology. All 23 cases had a major abnormality; none of the these cases were classified as having a birth defect due to the presence of 2 or more minor abnormalities.
Among singleton births overall, 13.4% were born before 37 weeks gestation and 19.2% weighed less than 2500 g at birth (Table 4). Infants with first-trimester exposures seemed more likely to be born prematurely and with low birth weight compared with those whose earliest exposure occurred in the second or third trimester.
The APR reported an overall prevalence of birth defects of 2.7% among 6893 ARV-exposed pregnancies and 2.8% among 2673 pregnancies exposed during the first trimester.3 Other large observational studies have reported similar findings.16-18 This study found an overall birth defect prevalence of 2.4% among women exposed to lopinavir/ritonavir at any time during pregnancy and a birth defect rate of 1.9% among those exposed during the critical first trimester. These rates are consistent with those found in the internal and external comparison groups. The MACDP reports an overall prevalence of 2.67% and an “early diagnosis” rate of 2.09%.9 This “early diagnosis” rate may be a more appropriate comparator because the data from this study are provided primarily by obstetricians who often have limited access to diagnoses made after the first few days of birth.2 In the internal comparison group, the birth defect rate among first trimester exposures (1.9%) is consistent with those exposed in the second and third trimesters (2.6%). In all comparisons, birth defect rates in the lopinavir/ritonavir exposure group were lower than the comparator rates; however, the differences were not statistically significant.
Published preterm delivery rates of infants exposed to protease inhibitors in utero varies from 18% to 36%.19-23 A study of nelfinavir-exposed infants enrolled in the APR reported a prematurity rate of 11.8%.19 As a secondary analysis, this study reports a prematurity rate of 13.4%.
Several factors should be considered when interpreting these results. Because the APR employs voluntary reporting, clinicians may selectively enroll patients based on perceived risk. Women may be enrolled after prenatal testing if no abnormalities are identified; this practice could lower the estimated risk of birth defects.24 To reduce these potential sources of bias, the APR has recruited a group of providers agreeing to enroll all of their eligible patients into the APR.2,3 Obstetricians, often the primary reporters, may have incomplete health information beyond newborn period; this could lead to ascertainment bias.24
The MACDP has been identified by the APR Advisory Committee as the best available external comparison group; however, this comparison is not without limitations.10,24 The MACDP consists of 1,231,191 women (1,232,192 births) enrolled from within the metropolitan Atlanta area between 1968 and 2003; it is an active surveillance system with more rigorous case finding than the APR.10
The APR is a global registry; however, over 85% of enrollments come from the United States. The ability to generalize these results globally is limited, and comparable global cohorts are not available. The sample size in this study is sufficient to detect a 2.4-fold increase in the overall risk for birth defects but too small to assess the risk of individual birth defects. Most of the enrolled women received combination therapy; therefore, the impact of individual drugs remains difficult to assess. The APR Advisory Committee provides the following consensus statement3: “For the overall population exposed to ARV drugs in this Registry, no increases in risk of overall birth defects or specific defects have been detected to date when compared with observed rates for “early diagnoses” in population-based birth defects surveillance systems or with rates among those with earliest exposure in the second or third trimester. In analyzing individual drugs with sufficient data to warrant a separate analysis, an increased frequency for birth defects has been detected for didanosine only. Although no pattern of birth defects has been detected with didanosine, the Committee continues to monitor this increase. For abacavir, efavirenz, lopinavir/ritonavir, nelfinavir, nevirapine, ritonavir, stavudine, and tenofovir, sufficient numbers of first trimester exposures have been monitored to detect at least a 2-fold increase in risk of overall birth defects. No such increases have been detected to date. For lamivudine and zidovudine, sufficient numbers of first-trimester exposures have been monitored to detect at least a 1.5-fold increase in risk of overall birth defects and a 2.2-fold increase in risk of birth defects in the more common classes, cardiovascular and genitourinary systems. No such increases have been detected to date. Although the Registry population exposed and monitored to date is not sufficient to detect an increase in the risk of relatively rare defects, these findings should provide some assurance when counseling patients.”
The APR is ongoing; exposures to all antiretroviral therapies during pregnancy may be reported by contacting the APR at 800-258-4263 in the United States and Canada, 910-256-0238 internationally, or electronically, at www.APRegistry.com.
1. Watts DH. Management of human immunodeficiency virus infection in pregnancy. N Engl J Med. 2002;346:1879-1891.
2. Covington DL, Tilson H, Elder J, et al, and APR Steering Committee. Assessing teratogenicity of antiretroviral drugs: monitoring and analysis plan of the APR. Pharmacoepidemiol Drug Saf. 2004;13:537-545.
3. Antiretroviral Pregnancy Registry Steering Committee. Antiretroviral Pregnancy Registry International Interim Report for 1 January 1989 Through 31 July 2007. Wilmington, NC: Registry Coordinating Center; 2007.
4. Perinatal HIV Guidelines Working Group. 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. July 8, 2008 1-96 (Table 3, page 69-72). Available at: http://aidsinfo.nih.gov/ContentFiles/PerinatalGL.pdf
. Accessed July 19, 2008.
8. Kotelchuck MC, Wise P. Epidemiology of prematurity and goals for prevention. In: Taeusch HW, Yogman MW, eds. Follow-Up Management of the High-Risk Infant. Boston, MA: Little, Brown & Co; 1987: 3-14.
9. Centers for Disease Control and Prevention. Metropolitan Atlanta Congenital Defects Program. Surveillance Procedure Manual and Guide to Computerized Anomaly Record. Atlanta, GA: US Department of Health and Human Services, Public Health Service; 1998.
11. Correa A, Cragan J, Kucik J, et al. Metropolitan Atlanta congenital defects program 40th anniversary edition surveillance report: reporting birth defects surveillance data 1968-2003. Birth Defects Research. (Part A) 2007;79:65-93.
12. Stevenson RE, Hall JG, Goodman RM, et al. Human Malformations and Related Anomalies, Volume 1. Oxford Monographs on Medical Genetics, no. 27. New York, NY: Oxford University Press; 1993.
13. Collett D. Modelling Binary Data. London, United Kingdom: Chapman and Hall; 1991.
14. Fleiss JL, Levin B, Paik MC. Statistical Methods for Rates and Proportions. 3rd ed. New York, NY: Wiley Interscience; 2003.
15. Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic Research: Principles and Quantitative Methods. New York, NY: Van Nostrand Reinhold Company; 1982.
16. European Collaborative Study. Exposure to antiretroviral therapy in utero or early life: the health of uninfected children born to HIV-infected women. J Acquir Immune Defic Syndr. 2003;32:380-387.
17. Townsend CL, Tookey PA, Cortina-Borja M, et al. Antiretroviral therapy and congenital abnormalities in infants born to HIV-1 infected women in the United Kingdom and Ireland, 1990-2003. J Acquir Immune Defic Syndr. 2006;42:91-94.
18. Watts DH, Danner L, Handelsman E, 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.
19. Covington DL, Conner SD, Doi PA, et al. Risk of birth defects associated with nelfinavir exposure during pregnancy. Obstet Gynecol. 2004;103:1181-1189.
20. Cotter AM, Garcia, AG, Duthely ML, et al. Is antiretroviral therapy during pregnancy associated with an increased risk of preterm delivery, low birth weight, or stillbirth? J Infect Dis. 2006:193:1195-1201.
21. Morris AB, Dobles AR, Cu-Vin S, et al. Protease inhibitor use in 233 pregnancies. J Acquir Immune Defic Syndr. 2005:40:30-33.
22. Schulte J, Dominguez K, Sukalac T, et al. Pediatric Spectrum of HIV Disease Consortium. Declines in low birth weight and preterm birth among infants who were born to HIV-infected women during an era of increased use of maternal antiretroviral drugs: pediatric spectrum of HIV disease, 1989-2004. Pediatrics
. 2007;119:e900-e906. Available at: www.pediatrics.org/cgi/content/full/peds.2006-1123v1
. Accessed December 17, 2007.
23. Townsend CL, Cortina-Borja M, Peckham CS, et al. Antiretroviral therapy and premature delivery in diagnosed HIV-infected women in the United Kingdom and Ireland. AIDS. 2007;21:1019-1026.
24. Honein MA, Paulozzi LJ, Cragan JD, et al. Evaluation of selected characteristics of pregnancy drug registries. Teratology. 1999;60:356-364.
© 2009 Lippincott Williams & Wilkins, Inc.