Table 1.

Clinical characteristics

Author (year)	Country	Study design	Study population (clinical)	Exposures	Biological outcome	Clinical neurodevelopmental outcomes	Main outcomes	Limitations
Jahanshad (2015) 11	Thailand	Cross‐sectional	n = 30 HEU n = 33 age‐matched HUU Ages: five to fifteen years	In utero: HIV ART exposure only in those born after 2000 (regimen unknown) Infant: Some received AZT prophylaxis	Brain Magnetic Resonance Imaging with Diffusion Tensor Imaging	Wechsler Preschool and Primary Scale of Intelligence‐III Wechsler Intelligence Scale for Children‐III	No volumetric differences or differences in white matter integrity or brain structure between HEU and HUU. The diffusion tensor imaging measures were associated with Full Scale IQ and Performance IQ scores, but not Verbal IQ scores.	Small sample size and no power calculations.
Jankiewicz (2017) [12]	South Africa	Prospective Cohort	n = 65 HIV+ (51 with ART prior to 12 weeks of age, 14 with ART after 12 weeks) n = 19 HEU n = 27 HUU All were seven years old	In utero: HIV ART exposure for HIV+ infants varied between two regimens: single dose NVP alone or with AZT after 34 weeks gestation ART exposure for HEU was unknown Infant: Early versus late ART for HIV+ only	Brain Magnetic Resonance Imaging with Diffusion Tensor Imaging	n/a	In HEU children compared to HUU a cluster in the right posterior corona radiata with higher fractional anisotropy and increases and decreases in diffusivity metrics was found, while bilateral regions in the corticospinal tract demonstrated reduced diffusivity	Small sample size and no power calculations. Cohort was not assessed for prenatal or perinatal HIV infection.
Poblano (2004) 13	Mexico	Prospective Cohort	n = 37 HEU (12 AZT alone, 25 AZT/3TC) n = 37 HUU Age of HEU: 40.31 ± 6.72 weeks Age of HUU: 42.56 ± 4.79 weeks	In utero: HIV 12 AZT alone, 25 AZT/3TC	Brainstem auditory evoked potentials	n/a	Comparison of wave latencies showed significant delay of wave I and I‐III interwave interval in the AZT‐3TC treated group (p < 0.05). This subclinical effect on the auditory pathway would affect lower brainstem function.	Small sample size and no power calculations. Short follow‐up period.
Tardieu (2005) 14	France	Retrospective chart review	n = 49 HEU Ages: 10 to 44 months	In utero: HIV AZT (some as monotherapy, others with AZT included in various combinations)	Brain magnetic resonance imaging	Neurological assessment	50% of MRIs were abnormal with diffuse hyperintensity in the supratentorial white matter (n = 13) and in the pontine tegmentum (n = 14) as the most common findings. Other children had cerebral volume loss (n = 10), abnormality in the basal ganglia (n = 4), and some had necrosis in the white matter (n = 3).	Patients were selected based on being symptomatic, biasing the sample towards abnormality.
Tran (2016) 15	South Africa	Cross‐sectional study nested within an observational study	n = 15 HEU n = 24 HUU Ages: two to four weeks	In utero: HIV Triple therapy ART (not defined)	Brain Magnetic resonance imaging with diffusion tensor imaging	Dubowitz neurobehavioral scale	HEU were found to have higher fractional anisotropy in the middle cerebellar peduncles compared to HUU neonates, after correction for age and sex. Scores on the Dubowitz abnormal neurological signs subscales were positively correlated with fractional anisotropy (r = 0.58, p = 0.038) in the left uncinate fasciculus in HEU infants and negatively correlated with diffusivity metrics in the right inferior cerebellar peduncle and bilateral hippocampal cingulum in HEU infants.	Small sample size and no power calculations. Unclear how these results will translate over time.