CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

Background: Integrase inhibitors (INIs) including raltegravir (RAL), elvitegravir (EVG), and dolutegravir (DTG) now play a key role in antiretroviral therapy. Studies evaluating selection of drug resistance-associated mutations (DRAMs) in patients failing INI-based regimen are often performed by Sanger sequencing and may underestimate the prevalence of DRAMs. Moreover, few studies using ultra-deep sequencing (UDS) have been conducted to evaluate impact of pre- existing DRAMs on virological failure to INI treatment. Therefore, in this study, we aimed at evaluating the prevalence of DRAMs at failure and also the clinical impact of baseline DRAMs on INI response by UDS. Methods: Sanger and UDS were performed on samples at failure of 134 patients failing an INI-based regimen (65, 49, and 20 patients failed RAL, DTG, and EVG respectively) to study the prevalence of DRAMs. Samples at baseline of 34 patients with virological failure (VF) and of 31 with virological success (VS) under INIs were also sequenced by UDS to evaluate the clinical impact of pre-existing DRAMs on INI response. UDS data were analyzed by Smartgene platform and resistance was interpreted following the ANRS resistance algorithm V26. Results: At failure, the prevalence of at least one INI DRAM was 39.6% (53/134) by Sanger, 47% (63/134) by UDS at 5% detection threshold, and 57.5% (77/134) at 1% detection threshold. Among 53 patients harbouring at least 1 DRAM detected by both techniques, the most dominant DRAMs are the N155H, Q148H/K/R, and T97A observed in 24 (45.2%), 12 (22.6%), and 10 (18.9%) patients respectively whereas the Y143C was detected in 6 (11.3%) patients. UDS detected additionally DRAMs such as T66A/I (n=7), E92Q (n=3), T97A (n=2), E138K (n=3), G140S (n=1), Y143C/H (4), S147G (n=2), Q148R (n=2), N155H (n=3), S230G/R (n=2), and R263K (n=2). Overall, the presence of DRAMs detected only by UDS has led to changes in resistance interpretation to INI class in 13% (17/134) of patients at 1% of detection threshold. There was no difference in prevalence (14.7% vs 12.9%, p value = 0.817) of baseline DRAMs between patients with VF and those with VS under INI regimen. DRAMs found at baseline by UDS in patients with VF were not detected at failure either in majority or in minority. Conclusion: In this study, UDS was more sensitive than Sanger to detect minority DRAMs to INIs at failure. However, minority DRAMs identified at baseline did not emerge at failure and had no significant impact on the virological response to INI-based regimen. 546 INTEGRASE INHIBITOR RESISTANCE SELECTIONS INITIATED WITH DRUG RESISTANT HIV-1 Kristen Andreatta , Silvia Chang, Ross Martin, Madeleine Willkom, Kirsten L. White Gilead Sciences, Inc, Foster City, CA, USA Background: The integrase strand transfer inhibitors (INSTIs) raltegravir (RAL) and elvitegravir (EVG) have overlapping resistance pathways that develop during treatment failure while dolutegravir (DTG) and bictegravir (BIC) retain antiviral efficacy against many HIV-1 isolates with primary INSTI resistance (-R) substitutions in vitro. Here, the emergence of drug resistance in wild-type (WT) and mutant HIV-1 was assessed in parallel under in vitro selective pressure by the four INSTIs to compare viral evolution from preexisting drug resistant mutants. Methods: The INSTI-R substitutions E92Q, Q148R, N155H, and R263K in integrase (IN) and the common reverse transcriptase (RT) resistance substitution M184V (RT-M184V) were introduced into HIV-1 as single site- directed mutants. In vitro dose-escalation resistance selections were performed starting with these mutants or WT viruses using RAL, EVG, DTG, BIC, and emtricitabine (FTC). Viral supernatants were genotyped by population and deep sequencing. Site-directed mutants representing the final viral genotypes observed in the selections were phenotyped for drug resistance. Results: In selections initiated with WT HIV-1 IN (± RT-M184V), S153F/Y or R263K substitutions developed in the presence of BIC, S153Y or N144D developed in the presence of DTG, and Q148K/R and/or T66I developed in the presence of RAL and EVG (Table 1). In selections initiated with IN mutants (E92Q, Q148R, N155H, R263K), many had no further accumulation of INSTI-R substitutions: EVG and RAL selections were stopped after 16-26 weeks due to high levels of resistance while BIC and DTG selections continued for ~30 weeks with growth curves similar to the WT selections. In the presence of DTG, Q148R added the secondary INSTI-R substitution E138K, E92Q added the primary INSTI-R substitution S147G, and N155H added S147N. In the presence

544 UNINTEGRATED VIRAL DNA INVOLVED IN DOLUTEGRAVIR RESISTANCE Isabelle Malet 1 , Frédéric Subra 2 , Hervé Leh 2 , Charlotte Charpentier 3 , Gilles Collin 3 , Diane Descamps 3 , Eric Deprez 2 , Vincent Calvez 1 , Anne-Geneviève Marcelin 1 , Olivier Delelis 2 1 Institut Pierre Louis d’Epidémiologie et de Santé Publique, Paris, France, 2 Ecole Normale Supérieure de CachanEcole Normale Supérieure de CachanEcole Normale Supérieure de CachanEcole Normale Supérieur, Cachan, France, 3 Bichat–Claude Bernard Hospital, Paris, France Background: Dolutegravir (DTG), belonging to the strand-transfer inhibitor (STI) family, shows a high genetic barrier. Contrary to the first generation of STI, raltegravir and elvitegravir, no pathway of resistance has been yet described in DTG treated individuals. We previously selected a virus resistant to DTG without mutation in the integrase (IN) gene but with mutations in the 3’ poly-purine tract (PPT). In this study, we characterize the original replication mechanism of this mutant virus. Methods: Previously, a virus highly resistant to DTG was obtained by in-vitro selection. We infected MT4 cells with this virus in the presence or absence of a high DTG, RAL or EVG concentration (500 nM). We compared the replication of the mutant to the WT in the same conditions by quantitative PCR measuring all viral DNA forms (integrated, 2-LTR and 1-LTR circles DNA, linear DNA). We also sequenced integrated viral DNA sites in all conditions. Infections with these settings were also performed in Rev-GFP cells allowing quantification of late viral DNA expression by flow cytometry analysis. Results: We found that the replication of the mutant was important and similar with or without STI, highlighting the resistance of the mutant to all approved STI. As expected, the replication of the WT virus was totally impaired in the presence of STI. Interestingly, no integrated DNA has been detected for the mutant under STI treatment neither with qPCR nor with viral integrated DNA sequencing methods. Quantifications of episomal viral DNA show no accumulation of 2-LTR circles, normally observed when integration is inhibited, but an important amount of 1-LTR circles. Results obtained by infection of Rev-GFP cells show that expression of unintegrated viral DNA from the mutant is much more important than the one reported from unintegrated viral DNA from WT infection under STI treatment. Conclusion: For the first time, we describe the replication of a mutant resistant to all used STI without mutation in the integrase but which is supported by unintegrated viral DNA. Importantly, replication of the mutant under STI treatment, without integration, is due to the higher expression level of its unintegrated viral DNA compared to unintegrated viral DNA from the WT. This study highlights the role of unintegrated viral DNA in HIV-1 replication as a way to escape to IN inhibitors. These data are important to understand why some patients fail to STI inhibitors treatment without mutations in the integrase gene. 545 PREVALENCE AND CLINICAL IMPACT OF MINORITY RESISTANT VARIANTS TO INTEGRASE INHIBITORS Thuy T. Nguyen 1 , Djeneba Bocar Fofana 1 , Charlotte Charpentier 2 , Sidonie Lambert 1 , Laurence Morand-Joubert 1 , Diane Descamps 2 , Marc Wirden 1 , Christine Katlama 1 , Minh Le 2 , Gilles Peytavin 2 , Philippe Flandre 1 , Vincent Calvez 1 , Eve Todesco 1 , Anne-Geneviève Marcelin 1 1 Pierre and Marie Curie University, Paris, France, 2 Bichat–Claude Bernard Hospital, Paris, France

Poster Abstracts

CROI 2018 200