CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

356 EFFECTS OF AGE AND OBESITY ON NEUROCOGNITIVE PERFORMANCE IN THE MACS COHORT

Kristine Erlandson 1 , Long Zhang 2 , Lisa Jacobson 2 , James T. Becker 3 , Cynthia Munro 2 , Jordan E. Lake 4 , Eileen Martin 5 , Andrew Levine 6 , Todd Brown 2 , Ned Sacktor 2 1 Univ of Colorado, Aurora, CO, USA, 2 The Johns Hopkins Univ, Baltimore, MD, USA, 3 Univ of Pittsburgh, Pittsburgh, PA, USA, 4 Univ Texas Houston, Houston, USA, 5 Rush Univ, Chicago, USA, 6 Univ of California Los Angeles, Los Angeles, CA, USA Background: Age and obesity are both risk factors for dementia and associated with impaired neurocognitive (NC) performance, independent of age or obesity-associated comorbidities. We hypothesized that the effects of age and obesity on NC performance would be greater among HIV+ persons. Methods: Men with or at risk for HIV infection in the Multicenter AIDS Cohort Study (MACS) with 1) no diagnosed dementia or psychosis, 2) receipt of ≥2 antiretroviral drugs from ≥2 classes, 3) HIV-1 RNA <400 copies/mL at >80% of visits and 4) complete NC assessments from≥ 1 visit were included. BMI was categorized as underweight (BMI <18.5 kg/ m2), normal weight (18.5-24.9), overweight (25-29.9), or obese (≥30); under- and normal weight were combined for analyses. Log-transformed Trail Making Test, parts A and B (TrA, TrB) and the Symbol Digit Modalities test (SDMT) scores were adjusted for age, education, race, and repeat administrations. Mixed methods regression models determined associations of age and BMI with neurocognitive testing, adjusted for covariates. Results: HIV+ (n=795) and HIV- (n=1605) men contributed 10,023 and 23,204 visits, respectively. HIV+men contributed to 5148 under/normal weight, 3612 overweight, and 1263 obese visits; HIV- men to 10,157 under/normal weight, 8072 overweight, and 4975 obese visits. Baseline characteristics among HIV+ vs HIV- men, respectively, included: mean age 43.0 (standard deviation 10.1) vs 42.9 (10.9) years; BMI 25.7 (4.3) vs 26.0 (4.7) kg/m2. In multivariate models (Figure), increasing age was associated with poorer NC performance among both HIV+ and HIV- men on TrA and TrB (p<0.001) but not SDMT; there was no HIV by age interaction (p≥0.20). Obesity was associated with improved performance on TrA, TrB, and SDMT among HIV-uninfected participants (all p<0.05), and TrA among HIV-infected participants (p=0.038). There was no age by obesity or HIV serostatus by obesity interaction (all p≥0.11). Conclusion: Increasing age was associated with poorer NC performance on TrA and TrB among both HIV+ and HIV- men, with no difference by HIV serostatus. Furthermore, an unexpected protective effect of obesity on NC performance was observed, and most significant among HIV- controls. These findings could be the result of unbalanced drop-out or mortality differences at the BMI extremes, or may reflect physiologic processes associated with improved psychomotor speed. Further studies are needed to clarify the mechanisms underlying these findings.

Poster and Themed Discussion Abstracts

357 CSF TO PLASMA HIV-RNA RATIO ≥1 IS ASSOCIATED WITH HAND IN UNTREATED HIV PATIENTS Francesca Bai , Francesca Iannuzzi, Esther Merlini, Lidia Borghi, Teresa Bini, Giulia Marchetti, Antonella d’Arminio Monforte Univ of Milan, Milan, Italy Background: The association between cerebrospinal fluid (CSF) HIV-RNA and mild forms of HIV-associated neurocognitive disorders (HAND) is still controversial. We aimed to explore the association between plasma and CSF viral load and HAND in a cohort of untreated HIV+ patients (pts). Methods: We consecutively enrolled untreated HIV+ pts from 01/2011 to 08/2016 to undergo a neurocognitive evaluation (10 tests on 6 cognitive domains to diagnose HAND by Frascati’s criteria), lumbar puncture (CSF HIV-RNA) and blood sample (plasma HIV-RNA, CD38/CD45RA/CD45R0/CD127 on CD4/CD8 by flow cytometry). Pts were divided into HAND (including asymptomatic neurocognitive impairment [ANI], mild neurocognitive disorder [MND] or HIV associated dementia [HAD]) and noHAND (normal cognitive performances). Chi-square, Mann-Whitney test and logistic regression were used. Results: 155 pts were enrolled. HAND was diagnosed in 50/155 (32%) pts: 40 (80%) ANI, 10 (20%) MND; no HAD. Globally, median plasma HIV-RNA was 4.7 log10 cp/ml (IQR 4.11- 5.3) with a CSF HIV-RNA approximately 1 log lower than plasma levels (3.65 log 10 cp/ml, IQR 3-4.16). No differences in age, sex, hepatitis coinfections and time since HIV diagnosis were described between HAND and noHAND group; however, HAND pts were characterized by a higher frequency of AIDS events and CD4+ nadir <200 cells/mmc, in comparison to noHAND pts. No differences in median plasma and CSF HIV-RNA were observed with similar median CSF to plasma HIV-RNA ratio in the two groups; however, HAND was more frequently associated with ratio ≥1, in comparison to noHAND group (CSF to plasma HIV-RNA ratio ≥1: HAND 13, 26%-noHAND 11, 10.5%; p=0.013). Regarding T-cells immuno- phenotypes, lower naïve CD45RA+CD4+% and central memory CD127+CD4+% T-cells were associated with HAND (Table 1). In multivariate analysis, adjusting for CD4+ nadir <200 cells/mmc, CSF to plasma HIV-RNA ratio ≥1 (AOR 3.088, IC95% 1.172-8.138, p=0.023) and AIDS events (AOR 3.446, IC95% 1.173-10.123, p=0.024) were confirmed associated with HAND. Conclusion: in our cohort of untreated HIV+ pts, we reported a 32% of HAND prevalence with a higher frequency in pts with AIDS and CD4+ nadir <200 cells/mmc. Interestingly, mild forms of HAND were independently associated with a CSF to plasma HIV-RNA ratio ≥1, suggesting a compartmentalization of systemic infection into central nervous system with subsequent neuronal damage and neurocognitive impairment.

CROI 2017 141