CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

GI and a specific drug class intensification, except ABC for which we observed deterioration of Working Memory +1.5 vs 0 at W96 (p=0.008). Conclusion: Intensification strategy demonstrated clear improvement in GDS and HC, particularly with effective CPE 2016 ≥9, and correlation with enhanced CPE 2016 score. For ARV controlled pts, an updated CPE score taking in account potential drug toxicity could be useful. 416 CROSS-SECTIONAL AND CUMULATIVE LONGITUDINAL CPE SCORES ARE NOT ASSOCIATED WITH HAND Galia Santos 1 , Katharine Darling 1 , Isabella Locatelli 1 , Melanie Metral 1 , Alexandra Calmy 2 , Thanh Doco-Lecompte 2 , Christoph Hauser 3 , Helen Kovari 4 , Philip E. Tarr 5 , Marcel Stoeckle 6 , Caroline Di Benedetto 7 , Patrick Schmid 4 , Renaud Du Pasquier 1 , Matthias Cavassini 1 1 Lausanne University Hospital, Lausanne, Switzerland, 2 University Hospitals of Geneva, Geneva, Switzerland, 3 University of Bern, Bern, Switzerland, 4 University Hospital Zurich, Zurich, Switzerland, 5 University of Basel, Basel, Switzerland, 6 University Hospital Basel, Basel, Switzerland, 7 Ospedale Regionale di Lugano, Lugano, Switzerland Background: Higher Central nervous system Penetration Effectiveness (CPE) scores have been associated with better control of HIV replication in the central nervous system (CNS) but the link between CPE score and neurocognitive function remains controversial. Studies to date examining CPE score and HAND have used the CPE score as a cross-sectional variable. The potential association between HAND and a cumulative CPE score, representing the entire longitudinal history of a patient’s cART exposure over years, has not been explored. We hypothesized that a cumulative CPE score (summarizing scores from cART initiation to the time of neurocognitive assessment) would better predict the presence of HAND than a cross-sectional CPE score (a single measure at the time of neurocognitive assessement). Methods: The Neurocognitive Assessment in the Metabolic and Aging COhort (NAMACO) is an ongoing prospective longitudinal observational sub-study of the Swiss HIV Cohort Study (SHCS), set up to study the cognitive and neurological impact of HIV infection on an aging HIV-positive population. In this study, 981 HIV-positive SHCS participants aged ≥45 years underwent a standardized neurocognitive assessment. The cross-sectional and cumulative CPE scores were tested as potential predictors of the presence of HAND in multivariate logistic regression models. Results: The majority of patients were male (80%) and Caucasian (92%). Undetectable HIV-RNA in the plasma was recorded in 96%. Neurocognitive impairment was present in 40%. Overall, 27%were considered to have HAND with 26%, 0.8% and 0.6% classified as asymptomatic neurocognitive impairment, mild neurocognitive disorder and HIV-associated dementia, respectively. Non-HIV-associated factors contributed to neurocognitive impairment in 13% of the population. None of the cross-sectional and cumulative CPE scores tested was statistically significantly associated with neurocognitive impairment or with HAND. Conclusion: In this large and well-treated HIV-positive cohort, 27%were diagnosed with HAND. Cross-sectional and cumulative CPE scores were not associated with HAND or with non-HAND neurocognitive impairment. However, further longitudinal analysis will provide a valuable means of establishing factors in the future which may be associated with the appearance, persistence or resolution of neurocognitive impairment. The impact of modifying cART, and potentially CPE score, on neurocognitive performance will be examined in future longitudinal studies within the NAMACO study. Tess O’Meara 1 , Yong Kong 1 , Brinda Emu 1 , Kevin Robertson 2 , Richard W. Price 3 , Sarah Yosief 2 , Xinran N. Liu 1 , Rabib Chaudhury 1 , Serena S. Spudich 1 , Lingeng Lu 1 1 Yale University, New Haven, CT, USA, 2 University of North Carolina Chapel Hill, Chapel Hill, NC, USA, 3 San Francisco General Hospital, San Fransisco, CA, USA Background: Mechanisms of persistent neurological impairment during antiretroviral therapy (ART) are poorly understood. Exosomes, nanoparticles exocytosed from cells, contain bioactive molecules that affect cellular activity. We performed deep sequencing and pathway analyses of plasma exosomal miRNA (exo-miRNA) to examine associations between exo-miRNA signaling and neuropsychological (NP) performance after ART initiated in early infection. Methods: Participants who started ART in early infection were divided into 2 two groups based on Total-Z (TZ) score (motor, executive function, processing 417 EXOSOMAL MICRO-RNA EXPRESSION ASSOCIATES WITH NEUROPSYCHOLOGICAL FUNCTION DURING ART

speed, memory, and learning): higher (NP+, TZ>0) and lower scoring (NP-, TZ<0). Exosomes were precipitated from plasma using a polyethylene glycol solution and assessed with NanoSight nanoparticle tracking analysis (NTA), transmission electron microscopy (TEM), and exoELISA-Ultra assay for CD63 expression. After massive parallel sequencing of extracted exo-miRNA, differential expression between NP- and NP+ was assessed as -1≥log2 fold- change≥1, base-mean expression level>10, p<0.05, with patient age as a covariate. KEGG pathway analyses were performed on microT-CDS-predicted miRNA targets, with p value representing the probability of the pathway enriched by differentially expressed miRNAs. Results: Samples were collected from 19 men with median age 43 years, CD4+ count 641 cells/ul, plasma HIV RNA <50 copies, 0.6 years of pre-ART infection, and 1.2 years-duration of ART. Median TZ scores were 0.35 for NP+ (n=12) and -0.50 for NP- (n=7). Exosome extraction was confirmed by visualization of vesicles with TEM, measurement of their diameters by NTA, and detection of surface CD63 expression. After miRNA sequencing, 15 differentially expressed miRNAs were identified between NP+ and NP- groups (p<0.05). miR-708 was the most differentially expressed exo-miRNA, with 37-fold higher levels in NP- vs NP+ (p=9.6E-4). Let-7a-5p was expressed at 3-fold higher levels in NP+ vs NP- (p=3.7E-3). KEGG analysis implicated 12/15 identified exo-miRNAs in glycosphingolipid biosynthesis pathways (p=1.7E-11). Conclusion: High-throughput analyses revealed differentially expressed plasma exo-miRNAs in higher versus lower NP performing participants despite early suppressive ART. Many of these miRNAs are predicted to target pathways related to neurological function, including sphingolipid biosynthesis, previously related to neurodevelopment and cognition. Exo-miRNAs may associate with cerebral function and deserve further investigation. 418 NO CHANGE IN NEUROCOGNITIVE FUNCTION AFTER SWITCHING FROM EFAVIRENZ TO RILPIVIRINE Giuseppe Lapadula 1 , Davide P. Bernasconi 2 , Francesca Bai 3 , Antonio Di Biagio 4 , Emanuele Focà 5 , Stefano Bonora 6 , Francesco Castelli 7 , Claudio Viscoli 8 , Alessandra Bandera 1 , Antonella D’Arminio Monforte 3 , Andrea Gori 1 1 San Gerardo Hospital, Monza, Italy, 2 University of Milano–Bicocca, Milan, Italy, 3 University of Milan, Milan, Italy, 4 IRCCS AOU San Martino-IST, Genoa, Italy, 5 Spedali Civili di Brescia, Brescia, Italy, 6 University of Turin, Turin, Italy, 7 University of Brescia, Brescia, Italy, 8 University of Genoa, Genoa, Italy Background: Efavirenz (EFV) association with neuropsychiatric side effects is well known. However, a link between EFV and neurocognitive (NC) impairment is controversial. Whether switching from EFV to rilpivirine (RPV) can improve NC performances is still debated. Methods: Randomized open-label controlled trial. Eligible patients (pts) had confirmed HIV-RNA <50 cp/ml for >6 months under treatment with co- formulated TDF/FTC/EFV and at least one among: (i) Z-transformed score <-1 in at least 1 of 6 cognitive domains explored by NC assessment; (ii) depression or anxiety, defined as scores >90th percentile in Beck Depression Inventory- II (BDI-II) or Beck Anxiety Inventory (BAI); (iii) low quality of sleep (i.e., a Pittsburgh Sleep Quality Index [PSQI] score >5). Pts were randomized 1:1 to continuing TDF/FTC/EFV or switching to co-formulated TDF/FTC/RPV. Treatment efficacy, NC function and symptoms were evaluated 12 and 24 weeks after randomization. Results: 74/125 screened patients fulfilled inclusion criteria and were randomized to switch arm (36 patients) or continuation arm (38 patients). 89.2% of pts were male, mean age was 47 (SD:10.3) and current and nadir CD4+ counts were 702 (SD:265) and 299 (SD:139) cells/mm 3 . At baseline, total NC functioning z-score was -0.02 (0.78); 63% and 23.4% of the pts had z-scores below -1 in ≥1 or 2 domains. The proportion of pts with altered scores in BDI-II, BAI and PSQI was 31.1%, 17.6% and 44.6%. At week 24, 71 pts were re-evaluated by NC assessment. Overall, NC performances improved, but no statistically significant difference was found comparing the two arms (table 1). Depression and anxiety reduced over time, with no statistically significant difference between arms (at week 24, 14.8% vs. 17.1%, P=1 and 0% vs. 8.3%, P=0.25 for switch and continuation arm). A significant improvement in CNS symptoms, quality of sleep and self-reported cognitive failures was reported by pts in the switch but not in the continuation arm (table 1). At week 24, 97.1% and 97.3% of pts had HIV-RNA <50 copies/ml in the switch and continuation arm, respectively (P=0.24). No protocol defined virological failure, grade ≥3 laboratory abnormalities nor serious adverse events related to the study drugs were observed.

Poster Abstracts

CROI 2018 147