In the first five years of life the neurological development of HIV-exposed but uninfected infants, exposed to maternal antiretrovirals before and after birth and throughout breastfeeding, is comparable to that of HIV-unexposed uninfected infants from similar socio-economic backgrounds findings, a prospective cohort study in Uganda and Malawi, shows.
As increasing numbers of women living with HIV take antiretrovirals on an ongoing basis, the number of infants exposed to HIV as well as prolonged ART (during pregnancy, throughout breastfeeding and infant prophylaxis with nevirapine for about six weeks) will continue to grow.
While protecting children exposed to HIV from infection, the effect of prolonged ART exposure on their neurological development is a major concern for parents and healthcare workers.
Initial findings from the randomised PROMISE-BF study showed that while triple antiretroviral regimens were significantly more effective in preventing transmission there was a greater risk of adverse pregnancy outcomes compared to zidovudine alone. These outcomes, including low birthweight and preterm delivery, could potentially affect growth development.
Subsequent findings showed that infant exposure to ART during a prolonged breastfeeding period (up to 18 months) was safe with minimal transmission, minimal adverse health effects and high HIV-free survival at two years of age. However, neurodevelopmental outcomes throughout infancy and early childhood were not assessed.
For the new study, Professor Michael Boivin at the department of psychiatry, research division, Michigan State University and colleagues compared neuro-developmental outcomes of antiretroviral exposure before and after birth in HIV-exposed and uninfected children to HIV-unexposed and uninfected children at 12, 24, 48 and 60 months of age. All children were HIV negative, but some were born to mothers living with HIV (“HIV exposed”) and some were not (“HIV unexposed”).
The cohort was identified from two research sites in Malawi and Uganda of the PROMISE-BF trial. These are resource-poor settings which have adopted Option B+ and where most HIV-exposed infants are breastfed for prolonged periods of time. The two groups of infants were matched for age, sex and socioeconomic background. Primary outcomes were the Mullen Scales of Early Learning (MSEL) cognitive composite score at age 12, 24 and 48 months and the mental processing index for the Kaufman Assessment Battery for Children (KABC-II) global score at 48 and 60 months.
The MSEL test is used for longitudinal development assessment (visual reception, gross and fine motor skills, receptive and expressive language). The KABC II assesses cognitive ability outcomes from a neuro-psychological perspective (memory, visual-spatial processing and problem solving, learning (immediate and delayed memory), non-verbal index and mental processing index). Both have been validated in sub-Saharan Africa.
A total of 861 children were enrolled.
There were no differences in MSEL cognitive composite scores according to ART exposure at 12 and 24 months of age, (p=0.19 and p=0.24, respectively for comparisons of all groups). The groups were triple ART plus infant nevirapine, triple ART plus maternal triple ART, zidovudine plus infant nevirapine, zidovudine plus maternal triple ART and HIV-unexposed controls.
At 48 months, however, scores for children of mothers who did not remain on triple ART both before and after birth were not as good as those remaining on triple antiretrovirals throughout. Adjusted means were 80.6, 81.3 and 85.9, respectively (p=0.049 for the comparison of all groups).
The KABC-II composite scores (mental processing index) did not differ at 48 or 60 months of age according to exposure (p=0.81 and 0.89, respectively).
These findings, based on both longitudinal development and neuro-psychological assessments are unique. This is the first time such outcomes are available for HIV-exposed and uninfected children with prolonged exposure to antiretrovirals, note the authors.
A longer-term follow-up PROMOTE cohort study (of former PROMISE children) at seven sites will assess their later neuro-cognitive performance during their early school years.
Public health implications
Dr Peter Kazembe, of the Baylor College of Medicine Children’s Foundation in Malawi, in an accompanying comment notes the significant public health importance of these findings.
Policymakers can be reassured that the antiretrovirals they advocate do no neuro-developmental harm to the uninfected infant. These data further support lifelong ART: antiretrovirals are good for the mother’s health, prevent HIV transmission to the infant and cause no harm to the infant in the medium term. Healthcare workers can reassure mothers on ART that prolonged breastfeeding is scientifically sound and will not harm the infant.
Background: Antiretroviral medication during pregnancy and breastfeeding substantially decreases the risk of HIV transmission from mothers to infants, but its effects on the child’s neurodevelopment are unknown. This study compared neurodevelopmental outcomes of ante-partum and post-partum antiretroviral exposure in HIV-exposed and uninfected children with HIV-unexposed and uninfected children at ages 12, 24, 48, and 60 months.
Methods: For this study, a prospective cohort of HIV-exposed and uninfected children was identified from two research sites in the PROMISE-BF trial (at Blantyre, Malawi, and Kampala, Uganda), in which pregnant HIV-infected mothers were randomly assigned to triple antiretroviral prophylaxis (lopinavir–ritonavir plus either lamivudine and zidovudine or emtricitabine and tenofovir), versus zidovudine alone. Post partum, the mother–infant pairs were randomly assigned to maternal triple antiretroviral treatment or infant nevirapine during breastfeeding. HIV-unexposed and uninfected children matched for age, sex, and socioeconomic background were enrolled at vaccination and well-child clinics at the study sites. We included only children without a history of documented brain infection or injury or substantial malnutrition, and whose mothers were randomly assigned and maintained within their assigned ante-partum and post-partum phases throughout their treatment arm periods. Primary outcomes were the Mullen Scales of Early Learning (MSEL) cognitive composite score at age 12 months, 24 months, and 48 months; and the mental processing index for the Kaufman Assessment Battery for Children, second edition (KABC-II) global score at 48 months and 60 months. Repeated measures were analysed using a linear mixed-effects model controlling for data collection site.
Findings: Between Aug 23, 2013, and Dec 17, 2014, we co-enrolled 861 children. For MSEL assessments, 738 were eligible for inclusion at age 12 months, 790 at age 24 months, and 692 at age 48 months. For KABC-II assessments, 685 were eligible for inclusion at age 48 months and 445 at age 60 months. There were no differences in MSEL cognitive composite scores according to exposure at age 12 and 24 months (p=0·19 and 0·24, respectively, for comparison of all groups). At 48 months, MSEL cognitive composite scores were worse for children of mothers who did not remain on triple antiretroviral treatment throughout both the ante-partum and post-partum treatment phases (adjusted means 80·64 [95% CI 77·74–83·54] and 81·34 [78·19–84·48], respectively), compared with those who did remain on triple treatment (adjusted mean 85·93, 95% CI 83·05–88·80; p=0·0486 for the comparison of all groups). The KABC-II composite scores (mental processing index) did not differ at 48 or 60 months according to exposure (p=0·81 and 0·89, respectively, for comparison of all groups). Scores for MSEL and KABC-II for children of mothers on triple antiretrovirals in both the ante-partum and post-partum treatment phases were similar to those for children in the HIV-unexposed and uninfected reference group at all timepoints.
Interpretation: Maternal triple antiretroviral exposure during both the ante-partum and post-partum phases did not result in greater developmental risks for the mothers’ HIV-exposed and uninfected children through age 60 months, compared with children who were HIV-unexposed and uninfected. This might be because ante-partum triple antiretroviral protection of the health of mothers with HIV during pregnancy might be neuroprotective for the child, and when continued post partum, could enhance the quality of caregiving for the child through better clinical care for the mother.
Michael J Boivin, Limbika Maliwichi-Senganimalunje, Lillian W Ogwang, Rachel Kawalazira, Alla Sikorskii, Itziar Familiar-Lopez, Agatha Kuteesa, Mary Nyakato, Alex Mutebe, Jackie L Namukooli, MacPherson Mallewa, Horacio Ruiseñor-Escudero, Jim Aizire, Taha E Taha, Mary G Fowler