CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

analysis was used to identify co-occurring trajectories of psychiatric disorders, and multinomial logistic regression was used to examine sociodemographic predictors of the trajectories. A log-binomial model was used to examine the association between trajectories and a viremic event. Results: We identified 3 psychiatric trajectories spanning a median of 10 years. 1) AYA with “consistent low disorder” (63%) had no mood or behavioral disorders, few and decreasing anxiety disorders, and increasing but relatively few substance use disorders. 2) AYA with “persistent anxiety” (26%) had persistent anxiety disorders, low and decreasing behavioral disorders, and low but increasing mood and substance use disorders. 3) AYA with “escalating comorbidity” (11%) had substantial comorbidity at enrollment, with increasing substance use disorders, anxiety and mood disorders over time. At last FU, more than half (62%) of AYA had a viremic event. Compared to AYA with “consistent low disorder,” AYA with “escalating comorbidity” were significantly older (OR=1.62; 95% CI=1.26-2.10), reported higher neighborhood stress at enrollment (OR=4.28; 95% CI=1.67-11.0), and had 51% higher risk of a viremic event (RR=1.51; 95% CI=1.08, 2.12), while AYA in the “persistent anxiety” trajectory were more likely to be female (OR=2.05; 95%CI=1.17-3.61) and had 26% higher risk of a viremic event (RR=1.26; 95% CI=0.93, 1.72). Conclusion: PHIV AYA are at high risk for mental health and substance use problems, with more comorbidity over time associated with a viremic event. Addressing the substantial and evolving mental health challenges among AYA is critical to meeting 90-90-90 treatment goals. 820 TRAJECTORY ANALYSIS OF COGNITIVE OUTCOMES IN CHILDREN WITH PERINATAL HIV Payal Patel 1 , Tanakorn Apornpong 2 , Stephen J. Kerr 2 , Thanyawee Puthanakit 3 , Kulvadee Thongpibul 4 , Pope Kosalaraksa 5 , Pradthana Ounchanum 6 , Suparat Kanjanavanit 7 , Ly Penh Sun 8 , Saphonn Vonthanak 9 , Serena S. Spudich 1 , Jintanat Ananworanich 10 , Robert Paul 11 , for the PREDICT/Resilience Study Group 1 Yale University, New Haven, CT, USA, 2 HIV–NAT, Thai Red Cross AIDS Research Centre, Bangkok, Thailand, 3 Chulalongkorn University, Bangkok, Thailand, 4 Chiang Mai University, Chiang Mai, Thailand, 5 Khon Kaen University, Khon Kaen, Thailand, 6 Chiang Rai Prachanukroh Hospital, Chiang Rai, Thailand, 7 Nakornping Hospital, Chiang Mai, Thailand, 8 National Centre for HIV/AIDS Dermatology and STDs, Phnom Penh, Cambodia, 9 University of Health Sciences, Phnom Penh, Cambodia, 10 Henry M Jackson Foundation, Bethesda, MD, USA, 11 University of Missouri St Louis, St Louis, MO, USA Background: Cognitive impairment is common in children with perinatal HIV (pHIV). HIV-related neuropathogenesis may produce distinct cognitive phenotypes as children age. We used trajectory modeling to identify clusters of children with pHIV following unique developmental trajectories and identified predictors of belonging to select cognitive trajectory groups. Methods: Participants included Thai and Cambodian children with pHIV enrolled in the PREDICT study. Children ages 4 to 12 years, with CD4% between 15-24% and no history of AIDS defining illnesses were included. Cognitive measures included intelligence tests, Children’s Color Trails, and Beery- Buktenica Developmental Test of Visual-Motor Integration and were conducted annually with a minimum follow-up of 3 years (median 5 years). Children with similar cognitive trajectories were classified using maximum likelihood estimation and Bayesian Information Criterion. Joint estimation was used to assess the influence of time varying co-variates of treatment initiation and viral suppression on trajectory course. Multiple logistic regression was employed to identify baseline factors (age, household income, parent as a caregiver, CD4 nadir, and treatment arm) associated with trajectory group membership. Results: At baseline assessment, 286 children had a median age of 8 years, median CD4% of 20%, and 51%were on ART. Trajectory analyses revealed a 3-cluster classification for cognitive data representing high, medium and low scoring groups. Figure 1 shows an example of trajectory groups for Children’s Color Trails 1. Scores in the low and medium trajectory groups were stable across adolescence. In contrast, the highest scoring group demonstrated a 10-point increase in scores from baseline. Children in the lowest scoring trajectory groups were more likely to enroll at an older age (p=0.01) and report lower household income (p<0.005). Neither CD4 nadir nor treatment arm (immediate versus deferred until immunosuppression ART initiation) was associated with cognitive trajectory status. Conclusion: Trajectory modeling succinctly classifies cohort heterogeneity in cognitive outcomes in pHIV. Most trajectory scores remained stable across age suggesting that cognitive potential is likely determined at an early age with

the exception of a subgroup of children who experienced developmental gains in select cognitive domains. Poverty and longer duration of untreated HIV may predispose children with pHIV to an increased risk of suboptimal cognitive development.

Poster Abstracts

821 FOCUSED COGNITIVE FUNCTION TESTING IN YOUNG PEOPLE WITH PERINATAL HIV IN ENGLAND

Alejandro Arenas-Pinto 1 , Hannah Castro 1 , Diane Melvin 2 , Marthe Le Prevost 1 , Caroline Foster 2 , Kate Sturgeon 1 , Alan Winston 3 , Lindsay Thompson 1 , Diana Gibb 1 , Ali Judd 1 , for the Adolescents and Adults Living with Perinatal HIV (AALPHI) Steering Committee 1 MRC Clinical Trials Unit at UCL, London, UK, 2 Imperial College Healthcare NHS Trust, London, UK, 3 Imperial College London, London, UK Background: We previously reported that cognitive performance in young people with perinatal HIV (PHIV+) without a CDC C diagnosis (PHIV+/no C) was similar to a comparable group of HIV-negative (HIV-) young people in England, but poorer than normative data, most noticeably in the domains of learning and memory. Here, we assess cognitive performance in the same cohort 2 years later, with expanded testing of these specific cognitive domains. Methods: 234 PHIV+ and 68 HIV- young people completed 9 tests: 5 NIH Toolbox tests measuring executive function (Flanker inhibitory control/ attention, dimensional change card sort), speed of information processing (pattern comparison), and memory (list sorting, picture sequence); 2 Hopkins Verbal Learning Tests (HVLT-R) (learning (L) (sum of 3 trials), delayed recall (R)); and 2 verbal language measures (Weschler Individual Achievement Test word reading (WIAT-II)), British Picture Vocabulary Scale). Z scores for each test were calculated using normative data, adjusted for age (and sex, ethnicity, and education for NIH Toolbox), and averaged by domain where appropriate. Chi squared, Wilcoxon rank sum and ANOVA tests compared proportions, median and means respectively, by HIV and CDC C status. Results: 139(59%) and 48(71%) of PHIV+ and HIV- were female (p=0.09), 202(86%) and 52(76%) were black (p=0.05), and median age was 19[17,21] and 18[16,21] years (p=0.45) respectively. 55(24%) of PHIV+ had a CDC C diagnosis (PHIV+/C). For HVLT-R, PHIV+/C participants had lower mean z scores (L -2.8 (95% CI -3.3, -2.2), R -2.6 (-3.1, -2.0)) than PHIV+/no C (L -2.0 (-2.3, -1.8), R -1.9 (-2.1, -1.6)) and HIV- participants (L -1.7 (-2.1, -1.3), R -1.5 (-1.9, -1.0)), and all were <1 SD below the reference mean (Figure). However, 292(97%) improved their score over the learning trials, and this was seen in all groups (p=0.62). PHIV+/C had poorer scores than the other 2 groups for verbal language measures, however mean scores were within 1 SD below the reference mean for all groups indicating mild impairment. NIH Toolbox executive function, speed of information processing and memory tests were similar for all 3 groups. Conclusion: Cognitive function was similar between PHIV+ and HIV- young people in most domains/tests. However, performance in verbal learning and recall fell below population normative scores, and was more pronounced in PHIV+/C, supporting wider findings that earlier ART initiation may protect aspects of cognitive development.

CROI 2019 319