CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

495 ARE INTEGRASE INHIBITORS A RISK FACTOR FOR IRIS IN THE ANRS 146 OPTIMAL TRIAL? Jean-Daniel Lelievre 1 , Lambert Assoumou 2 , Esther Aznar 3 , Federico Pulido 4 , Giuseppe Tambussi 5 , Manuel Crespo 6 , Agnes Meybeck 7 , Jean-Michel Molina 8 , Fanny Cardon 9 , Constance Delaugerre 8 , Rémi Lancar 2 , Lydie Béniguel 2 , Yves Levy 2 , Dominique Costagliola 2 1 University of Paris-Est, Paris, France, 2 INSERM, Paris, France, 3 Fundación SEIMC- GeSIDA, Madrid, Spain, 4 Hospital Universitario 12 de Octubre, Madrid, Spain, 5 Ospedale San Raffaele, Milano, Italy, 6 Vall d’Hebron Research Institute, Barcelona, Spain, 7 Centre Hospitalier de Tourcoing, Tourcoing, France, 8 Hôpital Saint-Louis, Paris, France, 9 France Recherche Nord & Sud Sida-HIV Hépatites, Paris, France Background: At CROI 2017, two observational studies based on retrospective assessment of IRIS occurring within 6 to 12 months after first antiretroviral treatment (ART) initiation have reported an increased risk of IRIS with integrase inhibitor (II) based regimen. We evaluated this association in the context of the ANRS 146 trial, an European, multicentre, randomized, double-blind, phase III trial, in France, Spain and Italy, in ART-naïve HIV1-infected adults with CD4+ count<200/μL or an ADE, in which IRIS was a component of the composite primary endpoint prospectively validated by an event review committee. Methods: Participants were randomized (1:1) to receive cART plus placebo or maraviroc for 72 weeks and no significant effect of maraviroc was evidenced on the composite primary endpoint (any new ADE, serious non-AIDS-defining event, IRIS, or death from any cause). IRIS was defined as 2 major criteria (A=atypical presentation of opportunistic infection or tumour in patients responding to ART, B=decrease in plasma HIV RNA >1log/mL) or major criterion A plus 2 minor criteria (increase in CD4 count after starting ART, increase in immune response to the relevant pathogen, spontaneous resolution of the disease without antimicrobial or tumour therapy with ART continuation). We compared the risk of IRIS using Kaplan-Meier estimates and multivariable Cox proportional-hazards models. Results: Between October 2011 and November 2014, 409 patients were included. At baseline, median HIV viral load (VL) was 5.39 log 10 copies/mL, median CD4+ count 80 cells/μL and 42% of participants had an ADE. Sixty-two individuals initiated with II (55 with raltegravir) and 347 did not (PI/r=283, NNRTI=64). No difference was seen in baseline characteristics of individuals (treatment arm, baseline CD4, VL, geographical origin) depending on whether initiating with II or not. Overall, 28 documented IRIS occurred, 26 within the first six months including 6 KS, 5 PCP, 4 folliculitis, 2 CMV, 2 cryptococcal infections. The 72 week rate of IRIS was 9.7% in individuals initiating with II (6) versus 6.5% in those not initiating with II (22), with a crude hazards ratio of 1.52(95%CI:0.61- 3.74) and an adjusted one (on treatment arm, origin, baseline CD4 count and VL) of 1.49(95%CI:0.60-3.71). Conclusion: No strong association was detected between the risk of IRIS and initiating ART with an integrase inhibitor in individuals presenting at an advanced stage in the ANRS Optimal trial, where IRIS was a prospectively validated endpoint. 496 RALTEGRAVIR-BASED ART IS EFFECTIVE AND SAFE IN HIV+ LIVER TRANSPLANT RECIPIENTS Christian Manzardo 1 , Lucia Serrano 2 , Jesús Fortún 3 , Marino Blanes 4 , Montserrat Tuset 1 , Miguel Montejo 5 , Miguel Jiménez 6 , Elisa Cordero 7 , Raquel Lastras 8 , Esteban Ribera 9 , Maria Luisa Gonzalez-Dieguez 10 , Asunción Moreno 1 , Antoni Rimola 1 , Jose M. Miro 1 1 Hospital Clinic of Barcelona, Barcelona, Spain, 2 Fundación SEIMC-GeSIDA, Madrid, Spain, 3 Hospital Ramon y Cajal, Madrid, Spain, 4 Hospital Universitario La Fe, Valencia, Spain, 5 Hospital Universitario de Cruces, Barkaldo, Spain, 6 Hospital Regional Carlos Haya, Malaga, Spain, 7 Hospital Universitario Virgen del Rocio, Sevilla, Spain, 8 Hospital Universitario de Bellvitge, Barcelona, Spain, 9 Hospital Universitario de la Vall d’Hebron, Barcelona, Spain, 10 Hospital Universitario Central de Asturias, Oviedo, Spain immunosuppressants (IS) and some antiretroviral families (especially protease inhibitors [PI] and non-nucleoside reverse transcriptase inhibitors [NNRTI]) remains a challenge. Raltegravir (RAL) is a non-boosted integrase inhibitor that did not interact with IS in a small trial in HIV-infected transplant recipients (Tricot, Am J Transplant 2009). Nevertheless, clinical experience in this setting is limited. The aim of this study was to analyze the efficacy and safety of RAL plus Background: Liver transplantation (LT) is safe in selected HIV- infected individuals. However, management of interactions between

2 nucleos(t)ide analogs (NUCs) vs. other antiretroviral therapy (ART) regimens in LT HIV-infected recipients. Methods: We performed a nationwide, multicenter cohort study, including 272 consecutive patients who underwent LT from 2002 to 2012 and were followed until December 2016. For the efficacy analysis, the study population comprised 211 patients who had started any of 4 post-LT ART regimens (table) and had completed at least 1 year of follow-up. For the safety analysis, we included 246 patients who died or underwent retransplantation (reLT) during the first year. Results: Patients receiving the 4 ART regimens (table) had comparable baseline donor and recipient characteristics (data not shown). In terms of virological suppression, no differences were found between the 4 ART regimens at 1 year after LT. However, a trend towards better CD4+ T-cell count recovery at 48 weeks was observed in the RAL group (table). As for safety, the survival analysis did not reveal any differences in mortality and/or reLT rates after 1 year between the 4 ART groups (p=.204 at 1 year for the combined endpoint by the log-rank test). Nevertheless, patients receiving RAL+2 NUCs had a significantly lower cumulative probability of experiencing acute graft rejection during the first 6 months after LT (12% [95%CI 3-21%] for RAL-based ART vs. 28% [95%CI 22-35%] for the other combined ART regimens; p=.021 [log-rank]). Conclusion: A post-LT ART regimen based on RAL+2 NUCs was as virologically effective as other ART regimens (PI, NNRTI) at 48 weeks. In addition, the regimen showed a trend towards better immune reconstitution and was associated with significantly lower rates of acute rejection. One-year mortality and reLT were similar between the 4 ART regimens. Whenever possible, RAL+2 NUCs should be the preferred ART regimen for HIV-infected individuals undergoing LT. 497 INSTI-BASED INITIAL ART IN A US COHORT OF HIV-INFECTED ADULTS Stockton Mayer 1 , Nabil Rayeed 2 , Rachel Hart 2 , Richard Novak 1 , Bienvenido Yangco 3 , Frank J. Palella 4 , Benjamin Young 5 , Kate Buchacz 6 1 University of Illinois at Chicago, Chicago, IL, USA, 2 Cerner Corp, Kansas City, MO, USA, 3 Infectious Disease Research Institute, Tampa, FL, USA, 4 Northwestern University, Chicago, IL, USA, 5 APEX Family Medicine, Denver, CO, USA, 6 CDC, Atlanta, GA, USA Background: The duration of antiretroviral (ART) regimens is important to HIV management. We compared duration of integrase strand transfer inhibitor (INSTI)-based ART regimens versus other contemporary regimens among adults in routine HIV care. Methods: We analyzed data records of HIV Outpatient Study (HOPS) cohort participants seen at ten United States (U.S.) HIV clinics during January 1, 2007 to June 30, 2016. We studied patients who initiated ART (baseline date) and had ≥2 HOPS clinic visits thereafter. We assessed the probability of remaining on a regimen (i.e., no drug additions or substitutions other than nucleoside reverse transcriptase inhibitors) and rates of achieving virologic suppression (HIV RNA < 200 copies/ml) on initial INSTI versus non-INSTI regimens by performing Kaplan Meier survival and Cox proportional hazards regression analyses. Results: Among 1,140 eligible patients, 290 were prescribed an INSTI regimen (163 raltegravir, 84 elvitegravir, 43 dolutegravir) and 850 were prescribed non-INSTI regimen. In both groups, most patients were male (76% INSTI, 79% non-INSTI), non-white (59% INSTI, 64% non-INSTI), and under 50 years old (85% INSTI vs. 86% non-INSTI). For the INSTI and non-INSTI groups, respectively, median baseline viral load (VL, copies/ml) was 36,120 (interquartile range [IQR], 11065, 139196) vs. 33,635 (IQR 7448, 127229) and median CD4 + cell count (CD4, cells/mm 3 ) was 335 vs. 305. In Kaplan-Meier analysis, the estimated probabilities of remaining on initial regimens at 2 and 4 years were 60% and 55% for INSTI and 50% and 40% for non-INSTI regimens, respectively (overall Log rank test p = 100,000: HR 0.6; CI 0.4,0.9). INSTI regimens were also associated with higher rates of achieving viral suppression (HR 1.5; CI 1.3, 1.8), irrespective of baseline VL ( 100,000: HR 2.0; CI 1.5, 2.6).

Poster Abstracts

CROI 2018 179