Safety of In Utero and Neonatal Antiretroviral Exposure
The results of the present study indicated that mean scores for the cognitive and academic outcomes in this sample of HIV-exposed children, age 5–13 years, fell slightly below the average range. There were no significant associations between any ARV regimen or drug class and any cognitive or academic outcome. In addition, no individual ARV drug was associated with significantly lower cognitive or academic outcomes. Overall, these results support the safety of ARV use during pregnancy and the neonatal period and increase our knowledge regarding the long-term effects of in utero and neonatal ARV exposure on the cognitive and academic functioning of HIV-exposed, uninfected school-age children.
The present study found no differences in verbal intelligence when comparing school-aged children with and without atazanavir (Reyataz; Bristol-Myers Squibb, New York, NY) exposure, contrary to earlier analyses of SMARTT data by Sirois et al and Rice et al that found small but significant negative associations between atazanavir exposure and language development in infants at 1 and 2 years of age. Atazanavir is a relatively new ARV approved for use during pregnancy; the cohorts of infants and young children in the earlier studies were more likely to have been exposed to atazanavir than the cohort of school-age children in the present study. The finding of no difference in our study may therefore be because of the small number of children with atazanavir exposure. This question should be further examined in future studies.
Children in this study with in utero tenofovir exposure attained higher mean WPPSI-III PIQ scores and those exposed to combination neonatal prophylaxis obtained higher WASI VIQ and WIAT-II-A Word Reading and Spelling scores. Although the finding regarding PIQ suggests a possible relative protective effect of tenofovir exposure, it should be interpreted with caution, given the multiple comparisons included in the analysis. In addition, although we controlled for maternal health factors during pregnancy and socioeconomic factors that may have predicted maternal use of specific ARV regimens or individual ARV drugs, as well as receipt of combination prophylaxis in infants, these findings could be at least partially attributable to residual confounding from unmeasured or imperfectly measured factors.
In the general population, many variables have been found to affect cognitive and academic functioning, specifically, maternal cognitive status, prematurity, small for gestational age and household income. This study of children with in utero HIV exposure corroborated the findings of previous research. Our study participants were more likely to live in low income households, with mothers or caregivers whose mean cognitive scores fell below average. Prenatal alcohol exposure was significantly associated with several of the outcomes. Although we could reasonably have expected that HIV and ARV exposure, in addition to these factors, might have been associated with even lower scores, this was not seen in the present study and is an encouraging finding. Thus, although prenatal HIV and ARV exposure may be important factors to consider, this study highlights the significant influence of other biological and environmental factors on development of children exposed to HIV.
The strengths of this study included the large sample of children with perinatal HIV and ARV exposure and valid cognitive and academic data. The study included detailed information concerning maternal immunologic status, ARV history during pregnancy and duration of ARV use. Detailed information about the children's birth history and exposure to ARVs in the neonatal period and other relevant confounders was also included.
The generalizability of our results may be limited by demographic differences between the children included versus not included in the analyses. The requirement for fluency in English limited the number of Spanish-speaking children who completed the assessments for this study. Although there was a significant difference between the children included and not included in the WPPSI-III sample with respect to their exposures to cARV in utero, the proportion of children with prenatal exposure to cARV was high in both groups; thus, the difference was unlikely to affect study conclusions. Additionally, we did not have information on maternal adherence to ARV regimens and the analyses did not include all possible combinations of ARVs, thus limiting our ability to detect potential effects of any specific combination.
In conclusion, these results support the safety of ARV use during pregnancy and the neonatal period with respect to later cognitive and academic outcomes in school-aged children. As access to ARVs to prevent mother-to-child transmission increases across the globe, it is anticipated that an increasing number of children will experience in utero ARV exposure. Thus, it is critical that more information be obtained on the potential short- and long-term toxicities of such exposures on the child. New drugs and combinations of drugs continue to become available and will be used during pregnancy for treatment of women with HIV. Therefore, continued follow-up studies of HIV- and ARV-exposed youth are critical to monitor for potential neurobehavioral consequences of such exposure and to help determine whether there are optimal approaches to choice of ARV drug regimens during pregnancy.
Discussion
The results of the present study indicated that mean scores for the cognitive and academic outcomes in this sample of HIV-exposed children, age 5–13 years, fell slightly below the average range. There were no significant associations between any ARV regimen or drug class and any cognitive or academic outcome. In addition, no individual ARV drug was associated with significantly lower cognitive or academic outcomes. Overall, these results support the safety of ARV use during pregnancy and the neonatal period and increase our knowledge regarding the long-term effects of in utero and neonatal ARV exposure on the cognitive and academic functioning of HIV-exposed, uninfected school-age children.
The present study found no differences in verbal intelligence when comparing school-aged children with and without atazanavir (Reyataz; Bristol-Myers Squibb, New York, NY) exposure, contrary to earlier analyses of SMARTT data by Sirois et al and Rice et al that found small but significant negative associations between atazanavir exposure and language development in infants at 1 and 2 years of age. Atazanavir is a relatively new ARV approved for use during pregnancy; the cohorts of infants and young children in the earlier studies were more likely to have been exposed to atazanavir than the cohort of school-age children in the present study. The finding of no difference in our study may therefore be because of the small number of children with atazanavir exposure. This question should be further examined in future studies.
Children in this study with in utero tenofovir exposure attained higher mean WPPSI-III PIQ scores and those exposed to combination neonatal prophylaxis obtained higher WASI VIQ and WIAT-II-A Word Reading and Spelling scores. Although the finding regarding PIQ suggests a possible relative protective effect of tenofovir exposure, it should be interpreted with caution, given the multiple comparisons included in the analysis. In addition, although we controlled for maternal health factors during pregnancy and socioeconomic factors that may have predicted maternal use of specific ARV regimens or individual ARV drugs, as well as receipt of combination prophylaxis in infants, these findings could be at least partially attributable to residual confounding from unmeasured or imperfectly measured factors.
In the general population, many variables have been found to affect cognitive and academic functioning, specifically, maternal cognitive status, prematurity, small for gestational age and household income. This study of children with in utero HIV exposure corroborated the findings of previous research. Our study participants were more likely to live in low income households, with mothers or caregivers whose mean cognitive scores fell below average. Prenatal alcohol exposure was significantly associated with several of the outcomes. Although we could reasonably have expected that HIV and ARV exposure, in addition to these factors, might have been associated with even lower scores, this was not seen in the present study and is an encouraging finding. Thus, although prenatal HIV and ARV exposure may be important factors to consider, this study highlights the significant influence of other biological and environmental factors on development of children exposed to HIV.
The strengths of this study included the large sample of children with perinatal HIV and ARV exposure and valid cognitive and academic data. The study included detailed information concerning maternal immunologic status, ARV history during pregnancy and duration of ARV use. Detailed information about the children's birth history and exposure to ARVs in the neonatal period and other relevant confounders was also included.
The generalizability of our results may be limited by demographic differences between the children included versus not included in the analyses. The requirement for fluency in English limited the number of Spanish-speaking children who completed the assessments for this study. Although there was a significant difference between the children included and not included in the WPPSI-III sample with respect to their exposures to cARV in utero, the proportion of children with prenatal exposure to cARV was high in both groups; thus, the difference was unlikely to affect study conclusions. Additionally, we did not have information on maternal adherence to ARV regimens and the analyses did not include all possible combinations of ARVs, thus limiting our ability to detect potential effects of any specific combination.
In conclusion, these results support the safety of ARV use during pregnancy and the neonatal period with respect to later cognitive and academic outcomes in school-aged children. As access to ARVs to prevent mother-to-child transmission increases across the globe, it is anticipated that an increasing number of children will experience in utero ARV exposure. Thus, it is critical that more information be obtained on the potential short- and long-term toxicities of such exposures on the child. New drugs and combinations of drugs continue to become available and will be used during pregnancy for treatment of women with HIV. Therefore, continued follow-up studies of HIV- and ARV-exposed youth are critical to monitor for potential neurobehavioral consequences of such exposure and to help determine whether there are optimal approaches to choice of ARV drug regimens during pregnancy.