The source population was children enrolled in PACTG protocols 219 and 219C, a multisite US cohort of children born to HIV-infected women initiated to study the long-term effects of in utero ARV exposure and complications of pediatric HIV infection. Protocol 219 followed HIV-infected and HIV-uninfected perinatally exposed children at clinics across the United States from May 1993 through August 2000. Children currently or previously enrolled in another PACTG protocol and children whose mothers were enrolled in a PACTG perinatal protocol during pregnancy were eligible. In September 2000, a revised protocol was initiated, PACTG 219C, and the eligibility criterion mandating enrollment in another PACTG protocol was removed. The present study was restricted to children enrolled in 219 or 219C before 1 year of age to improve the accuracy of birth defect information recorded on protocol case report forms. The study was approved by site institutional review boards, and parents or guardians provided informed consent.
Study visits, which included physical examinations, were scheduled every 3 months for HIV-infected children, and every 6 months until 2 years of age (protocol 219), or every 3 months through 1 year of age (protocol 219C) and annually thereafter for HIV-uninfected children. Protocol 219 did not include a direct question regarding the presence of defects, but birth defects were a primary outcome and were recorded on diagnosis case report forms. Protocol 219C included a direct question regarding birth defects. Detailed data on birth defects also were collected in PACTG perinatal protocols 076, 185, 249, 250, 316, 332, 353, 354, 358, and 386 and the International Maternal Pediatric and Adolescent AIDS Clinical Trials (IMPAACT) protocol P1025. Forty-two percent of mother-infant pairs in protocol 219 and 219C participated in one of these perinatal protocols during pregnancy-gestation; these data were used to supplement 219 and 219C data.
Gestational age at birth was estimated from the date of last menstrual period, ultrasound measurement before 22 weeks gestation, or newborn examination. Trimesters were defined as first trimester, conception to <14 weeks gestation; second trimester, 14 weeks to <28 weeks gestation; and third trimester, 28 weeks to delivery. The primary determinant was first trimester in utero ARV exposure. We considered overall ARV exposure, ARV classes, and specific ARV agents to which at least 1 child with a birth defect was exposed in the first trimester. The reference group consisted of children unexposed to the particular ARV drug (or class) during the first trimester, and thus included ARV unexposed children, children exposed to ARV in labor only, children unexposed to the particular ARV drug but to other ARV, and children exposed to the particular ARV drug in the second and/or third trimester only. We also examined ARV exposure by trimester of first exposure (unexposed, first trimester, second or third trimester); however, since the first trimester estimates were substantially unchanged in this model from the former classification, results from the more parsimonious models were presented.
The outcome was the presence of a birth defect documented within the first year of life. Clinicians blinded to ARV exposure reviewed and classified the reported defects according to the MACDP guidelines as major defects or conditional defects. To further prevent misclassification, we followed a modified version of MACDP guidelines employed by the APR, in which children with 2 or more conditional defects in the absence of a major defect were considered a case. Therefore, a child with at least 1 major defect or at least 2 conditional defects in the absence of a major defect was considered a case. Children classified as having birth defects solely based on conditional MACDP defects were categorized separately from those with major defects.
The prevalence and exact 95% CI of birth defects per 100 live births was estimated overall, by cohort (219 vs. 219C), and infant HIV-infection status. Differences in birth defect prevalence across these and other characteristics were assessed using the χ2 test, Fisher exact test, and Cochran-Armitage trend test for categorical variables, and the Wilcoxon rank sum test for continuous variables. Logistic regression models were used to estimate associations between first trimester in utero ARV exposure of any drug and of specific drugs and the most common categories of birth defects (all birth defects, musculoskeletal defects, and heart defects), including both HIV-infected and uninfected children. Potential confounders with a P < 0.25 in univariate analysis were initially included in adjusted models, but only those that produced at least a 10% change in the estimated odds ratio were retained in final models. Children with recognized chromosomal abnormalities or congenital infections such as toxoplasmosis were excluded from regression analyses.
Pediatr Infect Dis J. 2010;28(8):721-727. © 2010 Lippincott Williams & Wilkins
Cite this: Birth Defects among Children Born to Human Immunodeficiency Virus-infected Women: Pediatric AIDS Clinical Trials Protocols 219 and 219C - Medscape - Aug 01, 2010.