Two separate analyses of prospectively collected data on prenatal exposure to dolutegravir were performed using data from the APR and EPPICC. Data from EPPICC were supplemented with patient data collected during the same time frame in the European Treatment Network for HIV, Hepatitis, and Global Infectious Diseases and obstetric sites participating in the Pharmacokinetics of Newly Developed Antiretroviral Agents in HIV-Infected Pregnant Women network. These analyses included pregnant women who received dolutegravir-containing regimens, prospectively enrolled into the studies before any knowledge of pregnancy outcome. In addition to data on dolutegravir exposure, we also assessed sociodemographic and clinical characteristics, delivery details, frequency of birth defects, birth weight, gestational age, and other neonatal outcomes.
APR receives annual institutional review board (IRB) approval from Western IRB, including waivers of informed consent and authorization for use and disclosure of protected health information. No patient identifiers were collected, ensuring the confidentiality of APR registrants.
Participating studies in EPPICC obtained ethical approval from national, local, or both levels, as well as informed consent where required, within these approvals.
Antiretroviral Pregnancy Registry. APR is an international, voluntary, prospective exposure registry that contains case reports from 70 countries, with women from the United States and its territories comprising 77.3% of those reports. The methods and analyses from APR have been previously published. Pregnant women with prenatal exposure to ARV medication were registered by their health care providers, who reported the women's exposure to ARVs during pregnancy and the pregnancy and birth outcome data. All data were semiannually reviewed by an independent advisory committee. Exposures to ARV were classified and analyzed by the earliest trimester of exposure to each ARV medication. The proportion of birth defects was compared with internal and external comparator groups. External comparators were 2 population-based surveillance systems, Metropolitan Atlanta Congenital Defects Program (MACDP) of the Centers for Disease Control and Prevention and the Texas Birth Defects Registry. Internal comparators included exposures to other drugs and exposures in the second or third trimester of pregnancy relative to exposures during the first trimester (when organogenesis occurs). Only singleton births without defects were analyzed for spontaneous and induced abortions, stillbirths, and premature birth [<37 weeks of gestation, low birth weight (LBW), and very LBW (VLBW)]. Multiple births were excluded because of the increased risk of adverse outcomes associated with multiple births. The reporting period for APR data in this analysis began on January 1, 1989, and ended on January 31, 2018.
European Pregnancy and Paediatric HIV Cohort Collaboration. EPPICC is an international network of cohort studies conducting epidemiologic research on pregnant women and children infected with HIV and children exposed to HIV during pregnancy.[12,13] Participating studies collect prospective data on pregnant women and their infants, with exposure data collected before the pregnancy outcome is known, and pregnancies with birth defects were not excluded. All pregnant women reporting exposure to dolutegravir during their pregnancy to participating European studies by August 2016 were included in this analysis (see Appendix, Supplemental Digital Content, https://links.lww.com/QAI/B304, which lists these cohorts/studies/networks).
Anonymous, individual-level data were merged at University College London Great Ormond Street Institute of Child Health using a standard operating procedure for data formatting according to a modified HIV Cohorts Data Exchange Protocol data specification using study-specific unique identifiers. In addition to standard pregnancy and neonatal outcomes, mode of HIV acquisition, including vertical transmission, was reported.
The merged database underwent a comprehensive set of data-quality checks to prevent duplication and ensure accuracy and completeness.
In both studies, LBW was defined as <2500 g and VLBW was defined as <1500 g. Preterm delivery was defined as birth occurring at <37 completed gestational weeks. Birth defect prevalence was calculated by dividing the number of birth defects by the number of live births.
In the APR analysis, birth defects were reported based on the MACDP classification. The first- and second-trimester cutoffs were 14 and 28 weeks of gestation, respectively. Preterm delivery was defined as birth before 37 weeks of gestation by the health care provider's best obstetric estimate. Spontaneous abortion was defined as death of a fetus or expulsion of the products of conception before 20 weeks of gestation. Stillbirth was defined as the death of a fetus at ≥20 weeks of gestation or a fetus weighing ≥500 g (in cases for which the gestational age was unavailable).
In the EPPICC analysis, birth defects were classified according to the World Health Organization's International Classification of Diseases, 10th Revision, and European Surveillance of Congenital Anomalies (EUROCAT) classification. The first, second, and third trimesters were defined as <13, 13–24, and ≥25 completed gestational weeks, respectively. Small for gestational age (SGA) was defined according to sex-specific US standards. Induced abortion was defined as voluntary termination of pregnancy at <22 weeks of gestation. Spontaneous abortion was defined as the death of a fetus or expulsion of the products of conception at <22 weeks of gestation. Stillbirth was defined as death of a fetus occurring at ≥22 weeks of gestation.
Separate analyses were conducted for data from the EPPICC and APR cohorts. Data were not combined because of differences in data collection and methods used for analysis. For the EPPICC and APR data sets, standard descriptive statistics were used to summarize the data using STATA v12.0 (StatCorp, College Station, TX) and SAS v9.3 (Cary, NC), respectively. For the APR, statistical inference was based on exact methods for binomial proportions, with 80% power and a type 1 error rate of 5% for detecting a doubling of overall risk of birth defects. For specific defects, the power varied with the frequency of the defect and the size of the exposed group.
J Acquir Immune Defic Syndr. 2019;81(4):371-378. © 2019 Lippincott Williams & Wilkins