CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

an increased burden of aging-related phenotypes including frailty, which we have shown to be heightened in HIV and predictive of mortality. Cognitive impairment is a key aging-related phenotype prevalent in HIV. However, limited data exist on the relationship of cognitive impairment to frailty and its impact on mortality in the ART era. Methods: Standard neurocognitive assessments were performed cross- sectionally among HIV-infected and uninfected PWID in the ALIVE cohort from 2010 through 2012 in 5 domains: executive function, attention, learning/ memory, information processing and motor processing. Global cognitive performance was determined as the average of z scores from each domain. Frailty was assessed based on the 5 physical frailty phenotype domains – weight loss, low physical activity, exhaustion, decreased grip strength, and slow gait speed. Mortality was ascertained through 2016 through linkage to the National Death Index. Cox proportional hazards models were used to estimate the risk (hazard ratios [HR] with 95% confidence intervals [CI]) for all-cause mortality. Results: Among 519 ALIVE participants with a median age of 52 years, 41% were HIV positive. In multivariate analyses, older age and hazardous alcohol use were significantly associated with impairments in executive function, information processing, motor processing and global cognitive impairment. Being both frail and HIV-infected was associated with heightened information and motor processing impairments. Adjusting for sociodemographics, premorbid IQ, comorbidity, substance use and HIV disease stage, impaired information processing (aHR 1.41; 95% CI, 1.06, 1.88), motor processing (aHR 1.61; 95% CI, 1.30, 1.98) and global cognitive impairment (aHR 1.67; 95% CI, 1.10, 2.56) were significantly associated with increased mortality; global cognitive impairment and frailty (aHR 2.32; 95% CI, 1.03, 5.20) were independently associated with mortality. Conclusion: Cognitive impairment is a significant predictor of death among persons with HIV, independent of HIV disease stage, chronic disease comorbidity, and frailty. Further elucidation of the epidemiologic and biological underpinnings of cognitive impairment in HIV and PWID could facilitate interventions to improve survival for these populations. Davide De Francesco 1 , Jonathan Underwood 2 , Emmanouil Bagkeris 1 , Daphne S. Babalis 2 , Xinzhu Wang 2 , Laura Dickinson 3 , Saye Khoo 3 , Marta Boffito 4 , Caroline Sabin 1 , Alan Winston 2 , for the POPPY study group 1 University College London, London, UK, 2 Imperial College London, London, UK, 3 University of Liverpool, Liverpool, UK, 4 Chelsea and Westminster NHS Foundation Trust, London, UK Background: Limited data exist on the effects, either beneficial or detrimental, of nucleoside-reverse transcriptase inhibitor (NRTI) exposure on cognitive function in people living with HIV (PLWH). We investigated the associations of plasma tenofovir (TDF), emtricitabine (FTC), lamivudine (3TC) and abacavir (ABC) pharmacokinetics (PK) with cognitive function among PLWH recruited in the POPPY study. Methods: PK sampling and cognitive function (6 domains) were obtained from 638 PLWH on TDF, FTC, 3TC or ABC. For each drug, four different PK parameters were considered: area under the curve over 24 hours (AUC), maximum concentration (CMAX), trough concentration (CT) and clearance (CL/F). Cognitive scores were standardized into Z-scores (mean=0, sd=1) and averaged to obtain domain and global Z-scores. Associations between PK parameters and Z-scores were assessed using rank regression adjusting for age, gender, race, education, BMI, weight, recreational drug use, alcohol consumption, use of boosted protease inhibitors or efavirenz, as appropriate. Results: The 638 PLWH were predominantly male (87%), with a median (IQR) age of 52 (47, 59) years and 93% had a HIV RNA <50 copies/mL. 520 were on TDF, 485 on FTC, 125 on 3TC and 93 on ABC. The median (IQR) global Z-score was 0.06 (-0.31, 0.40), 0.08 (-0.29, 0.40), 0.09 (-0.32, 0.32) and 0.11 (-0.36, 0.33) in recipients of TDF, FTC, 3TC and ABC, respectively. After adjusting for potential confounders, including efavirenz use, none of the four TDF and FTC PK parameters were associated with global cognitive scores, with only weak associations with 3TC PK parameters (Table). Higher ABC AUC and CT were associated with better cognitive scores (both p’s=0.02), while increased CL/F was associated with poorer scores (p=0.04). In particular, ABC AUC [adjusted rho: 0.26 (0.05, 0.47), p=0.02] and CT [adjusted rho: 0.24 (0.03, 0.45), p=0.03] were associated with better visual attention, while associations with other

domains were non-significant [adjusted rho’s ranging from 0.12 and 0.08 (executive function) to 0.18 (psychomotor) for AUC and CT, respectively; all p’s>0.05]. Conclusion: Whilst we found no evidence of detrimental effects of NRTI exposure on cognitive function, greater ABC (but not TDF, FTC and 3TC) plasma exposure was associated with better cognitive scores. Although confounding due to adherence and other unmeasured factors may exist, these results could have implications for the design of future research programmes for PLWH with cognitive disorders.

420 VARIABILITY IN COGNITIVE IMPAIRMENT OVER TIME IN PEOPLE WITH HIV AND MATCHED CONTROLS Davide De Francesco 1 , Caroline Sabin 1 , Jonathan Underwood 2 , Magnus Gisslén 3 , Ferdinand Wit 4 , Rosan van Zoest 4 , Judith Schouten 5 , Ben Schmand 4 , Peter Portegies 5 , Peter Reiss 4 , Alan Winston 2 , for the Co-morBidity in Relation to AIDS (COBRA) collaboration 1 University College London, London, UK, 2 Imperial College London, London, UK, 3 University of Gothenburg, Gothenburg, Sweden, 4 Amsterdam Institute for Global Health and Development, Amsterdam, Netherlands, 5 OLVG, Amsterdam, Netherlands Background: Although cognitive impairment (CI) is frequently reported in people with HIV (PWH), limited data exist on the dynamics of change in treated individuals meeting definitions of CI over time. A greater understanding of these dynamics would assist in targeting individuals at risk of cognitive decline. Methods: Treated PWH with HIV RNA <50 copies/mL for ≥1 year and comparable controls underwent assessment of cognitive function (six domains) at baseline and after two years. Demographically-adjusted T-scores at baseline and follow-up were derived and the multivariate normative comparison (MNC) criterion was used to determine CI. Fisher’s exact test was used to assess differences in the number of people moving over time from CI to no CI (and vice versa) between PWH and controls. Individuals showing a significant change over time in their cognitive function, whilst accounting for practice effect, were identified by applying the MNC criteria to the differences between follow-up T-scores and those expected given baseline T-scores, socio-demographics and time between testing. Results: The 123 PWH and 77 controls were predominantly male (93% in both), with a median (IQR) age of 55 (50-61) and 56 (51-63) years, respectively. At baseline, the prevalence of CI was 20% in PWH and 4% in controls (p<0.001). Whilst none of the controls with CI at baseline improved to no CI at follow-up, 9/21 (43%) PWH moved from CI to no CI (p=0.50, Table). Moreover, 2% and 4% of PWH and controls without CI at baseline were classified as impaired at follow- up (p=0.70). Twelve (10%) PWH and 5 (6%) controls experienced a significant decline in cognitive function over two years (p=0.42): 6/12 (50%) PWH and 2/5 (40%) controls were classified with no CI at both baseline and follow-up; 1 (8%) PWH and 2 (40%) controls moved from not having CI to have CI. Conclusion: We observed different dynamics of change in cognitive function within this cohort. A substantial proportion of PWH who were classified as having CI initially, did not meet criteria for CI after 2 years; only less than half of both PWH and controls who significantly declined, stably met the definition of CI. Linkage of these detailed cognitive phenotypes with biomarker and neuroimaging findings may assist in understanding the underlying pathogenic mechanisms and developing future targeted management approaches.

Poster Abstracts

419 ASSOCIATIONS BETWEEN PLASMA NRTI CONCENTRATIONS AND COGNITIVE FUNCTION

CROI 2019 152