CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

386

NeuroHIV: A Novel High-Resolution Subcortical Shape Analysis Christopher R. Ching 1 ;Talia M. Nir 2 ; Boris A. Gutman 3 ; Neda Jahanshad 4 ; Jaroslaw Harezlak 5 ; Assawin Gongvatana 6 ; Ronald A. Cohen 7 ; Bradford Navia 8 ; Paul M.Thompson 3 ; for the HIV Neuroimaging Consortium 1 Univ of California Los Angeles Sch of Med, Marina del Rey, CA, USA; 2 Imaging Genetics Cntr, Univ of Southern California, Marina del Rey, CA, USA; 3 Imaging Genetics Cntr, Univ of Southern California, Marina del Rey, CA, USA; 4 Univ of Southern California, Los Angeles, CA, USA; 5 Indiana Univ Fairbanks Sch of PH, Indianapolis, IN, USA; 6 Univ of California San Diego, San Diego, CA, USA; 7 Univ of Florida, Gainesville, FL, USA; 8 Tufts Univ Sch of Med, Boston, MA, USA Background: Over 50% of HIV+ individuals experience neurocognitive impairment with brain injury having been documented in those on stable antiretroviral therapy. We applied a novel high-resolution subcortical shape analysis technique to better understand the pattern of brain change in a large cohort of HIV+ adults. We hypothesized that shape analysis would detect subcortical morphometric variations associated with clinical and neurocognitive metrics with higher sensitivity compared to traditional volumetric analysis. Methods: Cross-sectional T1-weighted brain MRI scans from 253 chronic HIV+ patients on stable cART were collected through the HIV Neuroimaging Consortium and HIV- associated Brain Dysfunction Study (Age: 47.1± 8.5yrs; M/F: 197/56; Infection Duration: 12± 15yrs). Two shape metrics of local thickness (radial distance) and surface area (Jacobian determinant) were derived from structural MRI across thousands of homologous surface points for bilateral nucleus accumbens, amygdala, caudate, hippocampus, putamen, pallidum, and thalamus shape models. A linear mixed effects model was fit separately for gross subcortical volume (a single value) and shape analysis metrics (thousands of thickness and surface area points) to evaluate associations with current and nadir CD4 counts, and neurocognitive domain-specific T-scores (processing speed (PS); learning (LT); motor function (MF)). Analyses were adjusted for age, sex, ethnicity, intracranial volume, HIV duration, antiretroviral treatment and history of drug use as fixed effects, and scan site as a random effect. All results were corrected for multiple comparisons and a standard FDR correction was applied ( q =0.05). Results: There were no significant associations between single value gross volumes and CD4 measures or neurocognitive scores. However, subcortical shape modeling revealed clusters of largely lower local thickness and surface area metrics associated with lower nadir and current CD4 values. Furthermore, lower thickness and surface area metrics were associated with more impaired PS, LT and MF neurocognitive performance scores (Table 1) . Conclusions: To our knowledge, this study represents the largest subcortical shape analysis of HIV. Our high-resolution subcortical shape analysis technique was more sensitive to associations between brain volume and CD4 counts as well as neurocognitive scores over traditional whole volume subcortical analyses and represents a powerful new technique to track HIV-associated brain change.

387 Cigarette Pack-Years May Affect Cingulate and Frontal Cortex in Chronic HIV Infection Kalpana Kallianpur ;Tracie M. Umaki; Lindsay M. Kohorn; Dominic C. Chow; Beau K. Nakamoto; Scott C. Souza; Cecilia M. Shikuma Univ of Hawaii, Honolulu, HI, USA

Poster Abstracts

Background: Worldwide, smoking is twice as prevalent in the HIV-infected community as in the general population. Smoking exposure is linked to cortical gray matter loss in HIV+ and healthy individuals. However, enhanced frontal and executive functions have been associated with smoking in HIV patients. Chronic nicotine treatment improves memory in HIV transgenic but not in healthy rats. Methods: Cortical volume and neuropsychological (NP) data, obtained in HIV+ subjects on antiretroviral therapy, were cross-sectionally examined for associations with cigarette pack-years. Subjects provided smoking data by self-report, underwent T1-weighted magnetic resonance imaging (MRI) at 3.0 Tesla, and completed an NP test battery that yielded global and domain-specific (psychomotor; executive function; learning and memory [NPZ_lm]; working memory) z-scores. Non-smokers (defined as never-smokers) had zero pack-years. Analysis of covariance compared volumes of frontal and cingulate cortex (CCx) between smokers and non-smokers, adjusting for age and intracranial volume (ICV). Effects of pack-years on volumes were assessed by linear regression controlling for age and ICV. Results: We evaluated 53 subjects [45 male; age 51 ± 7 years; median (IQR) current and nadir CD4 count 466 (299-592) and 125 (39-250) cells/mm 3 ; 44 with plasma HIV RNA<50 copies/mL; 19 non-smokers; 34 smokers (19 ± 12 pack-years)]. Smokers and non-smokers did not differ in demographics, clinical variables, cortical volumes or NP z-scores (p>0.05). Pack-years were associated with volumes of isthmus (β=−0.34, p=0.006), caudal anterior (β=−0.27, p=0.04) and total (β =−0.34, p=0.002) CCx; lateral orbitofrontal cortex (β=−0.30, p=0.02); rostral anterior (β =−0.21, p=0.05) and posterior (β=−0.23, p=0.08) CCx; and total cortex (β =−0.19, p=0.07). Higher NPZ_lm correlated with smaller caudal anterior (r=−0.34, p=0.02) and total (r=−0.26, p=0.07) CCx volume. Pack-years did not relate to NP z-scores. Conclusions: Cingulate cortex may play a crucial role in nicotinic effects on cognitive performance in chronic HIV disease. HIV+ individuals demonstrated negative associations between cigarette pack-years and volumes of cingulate and orbitofrontal cortex. Paradoxically, we found that reduced cingulate cortical volume correlated with better learning and memory. Future research (e.g., with functional connectivity MRI) may elucidate the underlying mechanism of this association, possibly leading to effective treatment for HIV- related neurocognitive impairment.

146

CROI 2016