CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

Poster and Themed Discussion Abstracts

371 CSF ESCAPE IN TREATED HIV INFECTION SHOWS X4 TROPISM AND DEFECTIVE VIRAL SEQUENCES Bryan Smith 1 , Guanhan Li 1 , Hiromi Imamichi 2 , Brad Sherman 2 , Manori De Alwis 1 , Govind Nair 1 , Joseph Snow 1 , Daniel S. Reich 1 , H. Clifford Lane 3 , Avindra Nath 1 1 NIH, Bethesda, MD, USA, 2 NIH, Frederick, MD, USA, 3 NIAID, Bethesda, MD, USA Background: There is often clinical concern when HIV is detectable in cerebrospinal fluid despite virologic suppression in plasma in well-treated patients with HIV infection. We describe such “asymptomatic” CSF escape in participants enrolled in a cohort study at the National Institutes of Health (NIH) and sequenced the envelope gene. Methods: Participants with HIV infection and a plasma viral load less than the limit of detection for at least 12 months were enrolled in a natural history study at the NIH evaluating cognitive and neurologic outcomes. All participants completed thorough neurologic and psychiatric evaluations, brain magnetic resonance imaging, a detail neuropsychological battery, and lumbar puncture. MRI brain volumes were determined using Freesurfer. CSF HIV RNA was determined using Roche COBAS Taqman HIV-1 v2.0 and plasma HIV was determined using the Abbott RealTime PCR. CSF escape was defined as a CSF HIV RNA level ≥40 copies/ml with a plasma level <40 copies/ml. Measures of cognition (HIV-associated neurocognitive disorder [HAND], average T-score, and Global Dementia Scale [GDS]) were determined using consensus criteria as previously described. Pac Bio systemwas used for sequencing of envelope gene, and viral outgrowth assay for detection of replicating virus. Results: All participants had a plasma viral load <40 copies/ml at the time of evaluation. Of 62 participants who completed all procedures, 9 (14.2%) had CSF escape with a mean CSF viral load of 74.2 copies/ml (range 41-289). Of demographic variables, there were no differences in age (mean 53.3 years without CSF escape and 54.4 years with escape) or sex (36% female without escape and 50% female with escape. There were no significant differences in brain volume or measures of cognition or depression between the two groups. In two individuals, defective viral sequences were predominantly found and analysis of V3 loop showed X4 tropic virus. In one patient, X4 tropic virus capable of infecting astrocytes was found by viral outgrowth assay. Conclusion: X4 tropic viral sequences predominate in the CSF in asymptomatic CSF escape, the clinical significance of which needs to be determined by longitudinal monitoring of changes in neurologic, cognitive, and psychiatric outcomes. 372 NEUROPSYCHIATRIC ADVERSE EVENTS ASSOCIATED WITH INTEGRASE STRAND TRANSFER INHIBITORS Prabha Viswanathan , Elande Baro, Guoxing Soon, Adam Sherwat, Kimberly Struble FDA, Silver Spring, MD, USA Background: The increased use of INSTI based regimens over PI or NNRTI based regimens is largely due to the potency and safety of INSTIs. Case reports of new-onset or worsening of neuropsychiatric conditions, particularly depression and suicidality, have emerged with INSTIs. These reports prompted FDA to investigate the relationship between INSTIs and neuropsychiatric adverse events (NPAE), and compare the frequency of NPAEs with INSTIs, PIs and EFV. Methods: FDA conducted a meta-analysis of 6 randomized, active-controlled Phase 3 trials in treatment-naïve subjects (TN) comparing raltegravir, dolutegravir or elvitegravir to: EFV, PIs (atazanavir or darunavir) or another INSTI. These trials were submitted by industry to FDA to support approval of the INSTIs. The frequency of NPAEs in each treatment group was assessed over 96 weeks. NPAEs were identified using the Standardized MedDRA Query (SMQ) Version 18.0. A broad search for terms in the Depression and Suicide/Self- Injury (DSS) SMQ and the Psychosis and Psychotic Disorders (PPD) SMQ was done. The event of highest toxicity grade was counted for each subject. Overall risk difference for INSTI vs EFV and INSTI vs PI was computed based on fixed effects and random effect analyses using inverse variance weights in both models. In the random effects model (DerSimonian- Laird estimate), the study variable was included as a random effect. Analyses were also performed by demographic subgroups (age, sex, race, region, and IV drug use).

CROI 2017 149