Neurodevelopmental effects of HIV exposure: A prospective neuroimaging study of uninfected children born to mothers living with HIV

BACKGROUND: Globally, over one million children are born every year to mothers living with HIV. The successful scale-up of antiretroviral therapy (ART) has reduced vertical HIV transmission, which means that most children are uninfected. Despite avoiding HIV infection, antenatal HIV exposure may still have serious consequences. Understanding the neurodevelopment of children who are HIV-exposed and uninfected (HEU) is therefore of substantial public health importance. Advances in neuroimaging provide insights into biological pathways underlying neurodevelopmental outcomes. The aim of this thesis was to examine the effects of in utero exposure to HIV on child brain structure and function. METHODS: The research includes a prospective neurodevelopment and neuroimaging study of HEU and HIV-unexposed (HU) children embedded in the Drakenstein Child Health Study (DCHS). The DCHS is a large population-based birth cohort of over 1000 mother-child pairs in a well-characterised South African community with high HIV prevalence. Validated developmental assessments using the Bayley Scales of Infant and Toddler Development-III and high-resolution structural magnetic resonance imaging (MRI) were undertaken in HEU and HU children from birth to three years. In addition, a conceptual framework and a systematic review and meta-analysis were developed to contextualise the results. FINDINGS: Neurodevelopment: In the DCHS, children who were HEU had two-fold increased risk of language delay compared to HU children at two years; receptive and expressive language scores were associated with maternal CD4 count in pregnancy. Across studies, the systematic review and meta-analysis found HEU children were at risk of poorer expressive language as well as gross motor development relative to HU children, with no consistent effects of maternal ART regimens. Potential HIV-specific and universal mechanisms for impaired development were identified. Neuroimaging: The MRI study demonstrated the feasibility of neuroimaging young children without sedation in a sub-Saharan African setting and significant associations between brain structure and neurocognitive function. On average, at age 2-6 weeks, infants who were HEU had smaller total grey matter and basal ganglia volumes compared to HU infants; volumes were associated with maternal CD4. At age 2-3 years, HEU children continued to have lower subcortical brain volumes, notably in the basal ganglia and hippocampus. Regional brain volumes correlated with language scores and were associated with maternal HIV severity. Additionally, HEU children had greater mean cortical thickness in the prefrontal region than HU children, which mediated the relationship between HIV exposure and language outcomes. CONCLUSION: This thesis provides novel data that children who are exposed to HIV in utero are at risk of impaired language development and altered brain structure in early life. Identified associations between HIV exposure, frontostriatal brain structure, and neurocognition suggest neurobiological pathways may underlie language deficits in children who are HEU. While findings indicate early neurodevelopmental differences, future research will delineate mechanisms and determine long-term implications. Overall, this work expands our understanding of the brain structure-function relationship, provides unique insights into the effects of HIV exposure on the brain during a critical developmental window, and highlights the importance of optimising antenatal HIV care to ensure children thrive.