Cognitive impairment affects 30-50% of perinatally-infected children with HIV in Sub-Saharan Africa. Pathophysiological models of HIV-associated cognitive impairment suggest that immune activation and chronic inflammation contribute to neuronal injury, eventually leading to neuronal dropout and cortical atrophy.

Thirty-three perinatally-infected subjects with HIV between the ages of 8-17 were recruited from the Pediatric Center of Excellence in Lusaka, Zambia as part of the HIV-associated Neurocognitive Disorders in Zambia (HANDZ) study. Six age-matched HIV-exposed uninfected controls were recruited from the community. T1 sequences were acquired on a 1.5T scanner. Study participants underwent a neuropsychological battery sampling 8 cognitive domains commonly affected in HIV, and a composite "NPZ8" score was generated to summarize cognitive test performance. Regional cortical thickness and volume were estimated and modeled with Freesurfer software.

We identified decreases in cortical volume (502 cm³ vs 539 cm³, p=0.118), left hippocampal volume (3.7 cm³ vs 3.9 cm³, p=0.113), and total gray matter volume (669 cm³ vs 704 cm³, p=0.1379) between HIV infected subjects and controls. Within HIV positive cases, we identified positive correlations between cortical thickness and NPZ8 score in the right parahippocampal gyrus (r=0.5051, clusterwise p-value (CWP)=0.0010) and medial temporal lobe (r=0.4983, CWP=0.0656). Among cases, we identified an increase in CSF volume in those with malaria history (1055 mm³ vs 845 mm³, p=0.004).

Although limited by the small sample size, these data support the hypothesis that total and regional brain volumes might be utilized as a biomarker for cognitive outcomes in future studies. These data also suggest that malaria coinfection may contribute to brain atrophy or altered cerebrospinal fluid pressure dynamics in pediatric HIV patients.