CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

381 Advanced MRI Predict Cognition at 1-Year Follow-up in Mildly Impaired HIV+ Patients Cristina Granziera 1 ; Gaetano Perrotta 2 ; Djalel Eddine Meskaldji 3 ; Guillaume Bonnier 3 ; Samanta Simioni 4 ; Matthias Cavassini 5 ; Melanie Metral 6 ; Gunnar Krueger 7 ; Renaud Du Pasquier 6 ; for the Swiss HIV Cohort Study 1 Massachusetts General Hosp, Boston, MA, USA; 2 Hosp Erasme, Univ Libre de Bruxelles, Bruxelles, Belgium; 3 École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland; 4 Inst de Lavigny, Lavigny, Switzerland; 5 Univ Hosp Lausanne, Univ of Lausanne, Lausanne, Switzerland; 6 CHUV, Lausanne, Switzerland; 7 Siemens, Renens, Switzerland

Background: Multi-contrast quantitative MRI has shown to be highly sensitive to detect brain damage in virally suppressed HIV patients. In this study we assessed the value of quantitative MRI in (i) monitoring the longitudinal evolution over 1 year of brain structural abnormalities in aviremic patients with and without minor cognitive impairment (MND+ and MND-) and in (ii) predicting cognitive outcome at one year in HIV+MND+ patients. Methods: We performed T1/T2* relaxometry and Magnetization Transfer Imaging at 3T MRI in 36 aviremic HIV+ patients, at tree time points (tp): study entry (tp0), 6 months (tp1) and 1 year follow-up (tp2). All patients underwent extensive cognitive and psychiatric examination at all time points. We applied three generalized linear models to fit each cognitive score at tp1 and tp2 as a function of (i) the information obtained by structural MRI data in each region of interest (T1 or T2* or magnetization transfer ratio-MTR) at tp0 and (ii) covariates (age and gender ). Multiple comparison correction was done using Bonferroni and cross-validation was performed with a leave-one-out-test. Multivariate Analysis of Variance was used to study the influence of independent variables (age, gender, patients group, MRI sequence and time-points) on three dependent variables (regional mean T1 relaxation time (rt), T2* rt and magnetization transfer ratio) at the 3 time points. Results: At 6 months, cognitive performances were mainly predicted by T1/T2* in basal ganglia and thalamus (table 1); at 1 year, they were also predicted by T1/T2* rt in global white and grey matter (WM and GM, table 1). No significant evolution of microstructural brain properties was observed in both MND+ and MND- over 1 year. These results suggest that in HIV+ patients (i) the integrity of the cortico-striatal- thalamic loop highly influence executive function, working memory and reaction time and that (II) WM integrity plays a major role in processing speed. Conclusions: Multi-contrast MRI metrics appeared to be predictors of cognitive function in well-treated HIV patients. Predicting the evolution of cognitive deficits in HIV patients is important as it might help selecting patients benefitting of more aggressive antiretroviral therapy or of targeted cognitive intervention. Future studies should aim at developing tools based on quantitative metrics to assess microstructural brain alterations on a single-patient basis. 382 Topographies of Cortical and Subcortical Volume Loss in HIV and Aging in the cART Era Anika Guha 1 ; Matthew R. Brier 2 ; Beau M. Ances 2 ; Mario Ortega 1 ; ElizabethWesterhaus 1 ; Brittany Nelson 1 1 Washington Univ Sch of Med, St. Louis, MO, USA; 2 Washington Univ in St. Louis, St. Louis, MO, USA

Background: Studies of HIV-associated brain atrophy often focus on a priori brain regions of interest, which can introduce bias. We used a minimally-biased multivariate approach to analyze changes in brain volumetrics associated with HIV and its relationship to aging, viral factors, and combination antiretroviral therapy (cART). Methods: A cross-sectional, comparative study of 51 HIV-negative (HIV–) and 146 HIV-infected (HIV+) participants was conducted. Structural MRI was acquired and analyzed from participants using principal component analysis (PCA) to reduce dimensionality and determine topographies of volumetric changes. Neuropsychological (NP) assessment was examined using global and domain-specific scores. The effects of HIV-disease factors (e.g. viral load, CD4, duration infection, etc.) on brain volumes and cognitive function were investigated using least absolute shrinkage and selection operator (LASSO) regression. Results: Two components of interest were visualized using PCA. An aging effect predominated both components. The first component, a cortically-weighted topography, accounted for a majority of variance across subjects (43.5% of variance) and showed an independent HIV effect. A secondary, subcortically-weighted topography (4.6%) showed HIV-status accentuated age-related atrophy. In HIV patients, the cortical topography correlated with global NP scores and nadir CD4, while subcortical volume loss correlated with recent viral load. Conclusions: Cortical regions showed the most prominent volumetric changes due to aging and HIV. Within HIV+ participants, cortical volumes were associated with immune history while subcortical changes were associated with current immune function. Cognitive function was primarily associated with cortical volume changes. Observed changes in chronically infected patients may reflect both past history of infection and current viral status. Future longitudinal studies should follow HIV patients soon after seroconversion. 383 Meta-Analysis of Large HIV Cohort Reveals CD4 Effects on Longitudinal Brain Atrophy Daniel Schonfeld 1 ;Talia M. Nir 1 ; Christopher R. Ching 2 ; Neda Jahanshad 3 ; Jaroslaw Harezlak 4 ; Giovanni Schifitto 5 ;Tong Zhu 6 ; Ronald A. Cohen 7 ; Bradford Navia 8 ; Paul M.Thompson 1 1 Imaging Genetics Cntr, Univ of Southern California, Marina Del Rey, CA, USA; 2 Univ of California Los Angeles Sch of Med, Marina del Rey, CA, USA; 3 Univ of Southern California, Los Angeles, CA, USA; 4 Indiana Univ Fairbanks Sch of PH, Indianapolis, IN, USA; 5 Univ of Rochester, Rochester, NY, USA; 6 Univ of Michigan, Ann Arbor, MI, USA; 7 Univ of Florida, Gainesville, FL, USA; 8 Tufts Univ Sch of Med, Boston, MA, USA Background: While advances in ART have dramatically improved life expectancies of HIV+ people, chronic infection is still associated with neurological deficits and brain injury. By pooling datasets across independent studies of HIV, we can boost statistical power to map brain disease progression and generalize findings to the world-wide HIV epidemic. In a pilot tensor based morphometry (TBM) study for the HIVNC consortium, we examined the effects of baseline CD4 count on longitudinal maps of local volumetric changes in 169 chronically HIV+ individuals on stable cART from 7 independent sites. Methods: For each site, anatomical T1-weighted image pre-processing included removal of extra-cerebral tissue, N3 bias field correction, ONLM denoising, and linear registration to a target brain template. For each subject, follow up T1-weighted scans were linearly and elastically registered to their respective baseline scans and 3D Jacobian expansion factor maps were created that characterized voxel-wise percent annual rate of volumetric changes. Longitudinal Jacobian maps were then spatially normalized to a common baseline study-specific template. For each site, voxel-wise regression models, adjusting for effects of sex and age, were used to test for effects of baseline CD4 count on structural variations in Jacobian maps. Across sites, an inverse variance weighted fixed-effects meta-analysis was used to combine results from independent measurements. The standard FDR method was used to correct for multiple comparisons ( q =0.05). Results: While no single site yielded significant findings, the meta-analysis took advantage of the consistent associations across sites and revealed that HIV+ individuals (mean baseline CD4: 381 +/- 235, Age: 48 +/- 8.6, HIV duration: 12 yrs +/- 6.7) with a lower baseline CD4 count displayed significantly higher rates of atrophy (expanded ventricles and white matter atrophy) than healthier subjects (critical p =0.0003675).

Poster Abstracts

144

CROI 2016