Cognitive impairment in treated HIV-disease
File(s)
Author(s)
Underwood, Jonathan
Type
Thesis or dissertation
Abstract
Background
Combination antiretroviral therapy has dramatically improved the outlook for people living with HIV-infection worldwide. As such, the focus of care in well-treated individuals has shifted to the management of long-term comorbidities, such as cognitive impairment. However, the pathogenesis of cognitive impairment in virologically suppressed individuals is unclear.
Aims
To determine the prevalence, characteristics and understand the pathogenesis of cognitive impairment in well-treated HIV-positive individuals and to assess biomarkers for their ability to predict cognitive impairment and longitudinal changes in cognition.
Methods
Cross-sectional analysis of two European cohorts (POPPY and COBRA) and longitudinal analysis of the CHARTER cohort using blood, cerebrospinal fluid, clinical, cognitive and neuroimaging data with advanced statistical techniques including machine-learning.
Results
Firstly, cognitive impairment was prevalent in ~20% of successfully treated patients compared to ~5% in demographically comparable controls. However, it was mild, not clearly associated with symptoms and remained stable over time. Additionally, the prevalence depended on the diagnostic method used, with simulation data demonstrating that the commonly used ‘Frascati criteria’ classifies impairment in ~25-50% of a normative control population. Secondly, cognitive impairment in well-treated patients was predominantly associated with white matter microstructural injury rather than grey or white matter atrophy and using multivariate machine learning techniques could be predicted with up to 80% accuracy. Thirdly, greater exposure to efavirenz and nevirapine were associated with clinical and neuroimaging signals of CNS neurotoxicity. However, these results should be interpreted with caution given their cross-sectional nature and limited sample size (n=60). Nevertheless, they justify further, prospective study given that millions are prescribed these drugs worldwide.
Conclusions
Cognitive impairment was more prevalent in well-treated HIV-positive patients compared to well matched controls, with white matter microstructural injury sustained before sustained suppression of HIV-viraemia the likely pathogenic driver. Reassuringly however, this impairment was generally mild, asymptomatic and remains stable over time.
Combination antiretroviral therapy has dramatically improved the outlook for people living with HIV-infection worldwide. As such, the focus of care in well-treated individuals has shifted to the management of long-term comorbidities, such as cognitive impairment. However, the pathogenesis of cognitive impairment in virologically suppressed individuals is unclear.
Aims
To determine the prevalence, characteristics and understand the pathogenesis of cognitive impairment in well-treated HIV-positive individuals and to assess biomarkers for their ability to predict cognitive impairment and longitudinal changes in cognition.
Methods
Cross-sectional analysis of two European cohorts (POPPY and COBRA) and longitudinal analysis of the CHARTER cohort using blood, cerebrospinal fluid, clinical, cognitive and neuroimaging data with advanced statistical techniques including machine-learning.
Results
Firstly, cognitive impairment was prevalent in ~20% of successfully treated patients compared to ~5% in demographically comparable controls. However, it was mild, not clearly associated with symptoms and remained stable over time. Additionally, the prevalence depended on the diagnostic method used, with simulation data demonstrating that the commonly used ‘Frascati criteria’ classifies impairment in ~25-50% of a normative control population. Secondly, cognitive impairment in well-treated patients was predominantly associated with white matter microstructural injury rather than grey or white matter atrophy and using multivariate machine learning techniques could be predicted with up to 80% accuracy. Thirdly, greater exposure to efavirenz and nevirapine were associated with clinical and neuroimaging signals of CNS neurotoxicity. However, these results should be interpreted with caution given their cross-sectional nature and limited sample size (n=60). Nevertheless, they justify further, prospective study given that millions are prescribed these drugs worldwide.
Conclusions
Cognitive impairment was more prevalent in well-treated HIV-positive patients compared to well matched controls, with white matter microstructural injury sustained before sustained suppression of HIV-viraemia the likely pathogenic driver. Reassuringly however, this impairment was generally mild, asymptomatic and remains stable over time.
Version
Open Access
Date Issued
2016-10
Date Awarded
2017-05
Advisor
Winston, Alan
Sponsor
European Union
Grant Number
FP-7-HEALTH 305522
Publisher Department
Department of Medicine
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)