Personalized risk index for neurocognitive decline among people with well-controlled HIV infection

T Yuen, MJ Brouillette, LK Fellows… - JAIDS Journal of …, 2017 - journals.lww.com
T Yuen, MJ Brouillette, LK Fellows, RJ Ellis, S Letendre, R Heaton, N Mayo
JAIDS Journal of Acquired Immune Deficiency Syndromes, 2017journals.lww.com
Background: Little is known about the predictors of neurocognitive decline in HIV+
individuals with good virological control. Identification of modifiable risk factors would allow
targeted interventions to reduce the risk of decline in higher risk individuals. The objective of
this study was to develop a risk index to predict neurocognitive decline over 3 years in
aviremic HIV+ individuals. Methods: As part of the CNS HIV Anti-Retroviral Therapy Effects
Research (CHARTER) study, HIV+ adults completed clinical evaluation and …
Abstract
Background:
Little is known about the predictors of neurocognitive decline in HIV+ individuals with good virological control. Identification of modifiable risk factors would allow targeted interventions to reduce the risk of decline in higher risk individuals. The objective of this study was to develop a risk index to predict neurocognitive decline over 3 years in aviremic HIV+ individuals.
Methods:
As part of the CNS HIV Anti-Retroviral Therapy Effects Research (CHARTER) study, HIV+ adults completed clinical evaluation and neuropsychological tests every 6 months. Group-based trajectory analysis was used to detect patterns of neurocognitive change; individuals who deteriorated≥ 0.5 SD on at least one neuropsychological test were considered decliners. Multiple logistic regression was used to identify baseline sociodemographic, clinical, biological, and lifestyle factors associated with decline in the subgroup that was consistently aviremic during the first 3 years. A risk index was developed using the beta-coefficients from the final regression model.
Results:
Neurocognitive decline occurred in 23 of 191 (12%) participants followed longitudinally. The baseline factors that predicted decline were glomerular filtration rate≤ 50 mL/min, known duration of HIV infection≥ 15 years, education≤ 12 years, and cerebrospinal fluid protein> 45 mg/dL.
Conclusions:
Using this analytic approach, neurocognitive decline was uncommon in this sample of aviremic HIV+ individuals. The 3-year risk of decline ranged from 2% in those with no risk factors to 95% in those with all 4. The strongest predictor was glomerular filtration rate, also a predictor of cardiovascular disease. This raises the possibility that controlling vascular risk factors could reduce the risk of neurocognitive decline.
Lippincott Williams & Wilkins