Quantitative prediction of integrase inhibitor resistance from genotype through consensus linear regression modeling.
ABSTRACT: BACKGROUND: Integrase inhibitors (INI) form a new drug class in the treatment of HIV-1 patients. We developed a linear regression modeling approach to make a quantitative raltegravir (RAL) resistance phenotype prediction, as Fold Change in IC50 against a wild type virus, from mutations in the integrase genotype. METHODS: We developed a clonal genotype-phenotype database with 991 clones from 153 clinical isolates of INI naïve and RAL treated patients, and 28 site-directed mutants.We did the development of the RAL linear regression model in two stages, employing a genetic algorithm (GA) to select integrase mutations by consensus. First, we ran multiple GAs to generate first order linear regression models (GA models) that were stochastically optimized to reach a goal R2 accuracy, and consisted of a fixed-length subset of integrase mutations to estimate INI resistance. Secondly, we derived a consensus linear regression model in a forward stepwise regression procedure, considering integrase mutations or mutation pairs by descending prevalence in the GA models. RESULTS: The most frequently occurring mutations in the GA models were 92Q, 97A, 143R and 155H (all 100%), 143G (90%), 148H/R (89%), 148K (88%), 151I (81%), 121Y (75%), 143C (72%), and 74M (69%). The RAL second order model contained 30 single mutations and five mutation pairs (p?
Project description:BACKGROUND: Different high-dimensional regression methodologies exist for the selection of variables to predict a continuous variable. To improve the variable selection in case clustered observations are present in the training data, an extension towards mixed-effects modeling (MM) is requested, but may not always be straightforward to implement.In this article, we developed such a MM extension (GA-MM-MMI) for the automated variable selection by a linear regression based genetic algorithm (GA) using multi-model inference (MMI). We exemplify our approach by training a linear regression model for prediction of resistance to the integrase inhibitor Raltegravir (RAL) on a genotype-phenotype database, with many integrase mutations as candidate covariates. The genotype-phenotype pairs in this database were derived from a limited number of subjects, with presence of multiple data points from the same subject, and with an intra-class correlation of 0.92. RESULTS: In generation of the RAL model, we took computational efficiency into account by optimizing the GA parameters one by one, and by using tournament selection. To derive the main GA parameters we used 3 times 5-fold cross-validation. The number of integrase mutations to be used as covariates in the mixed effects models was 25 (chrom.size). A GA solution was found when R2MM > 0.95 (goal.fitness). We tested three different MMI approaches to combine the results of 100 GA solutions into one GA-MM-MMI model. When evaluating the GA-MM-MMI performance on two unseen data sets, a more parsimonious and interpretable model was found (GA-MM-MMI TOP18: mixed-effects model containing the 18 most prevalent mutations in the GA solutions, refitted on the training data) with better predictive accuracy (R2) in comparison to GA-ordinary least squares (GA-OLS) and Least Absolute Shrinkage and Selection Operator (LASSO). CONCLUSIONS: We have demonstrated improved performance when using GA-MM-MMI for selection of mutations on a genotype-phenotype data set. As we largely automated setting the GA parameters, the method should be applicable on similar datasets with clustered observations.
Project description:BACKGROUND: Human Immunodeficiency Virus type 2 is naturally resistant to some antiretroviral drugs, restricting therapeutic options for patients infected with HIV-2. Regimens including integrase inhibitors (INI) seem to be effective, but little data on HIV-2 integrase (IN) polymorphisms and resistance pathways are available. MATERIALS AND METHODS: The integrase coding sequence from 45 HIV-2-infected, INI-naïve, patients was sequenced and aligned against the ROD (group A) or EHO (group B) reference strains and polymorphic or conserved positions were analyzed.To select for raltegravir (RAL)-resistant variants in vitro, the ROD strain was cultured under increasing sub-optimal RAL concentrations for successive rounds. The phenotype of the selected variants was assessed using an MTT assay. RESULTS: We describe integrase gene polymorphisms in HIV-2 clinical isolates from 45 patients. Sixty-seven percent of the integrase residues were conserved. The HHCC Zinc coordination motif, the catalytic triad DDE motif, and AA involved in IN-DNA binding and correct positioning were highly conserved and unchanged with respect to HIV-1 whereas the connecting residues of the N-terminal domain, the dimer interface and C-terminal LEDGF binding domain were highly conserved but differed from HIV-1. The N155 H INI resistance-associated mutation (RAM) was detected in the virus population from one ARV-treated, INI-naïve patient, and the 72I and 201I polymorphisms were detected in samples from 36 and 38 patients respectively. No other known INI RAM was detected.Under RAL selective pressure in vitro, a ROD variant carrying the Q91R+I175M mutations was selected. The Q91R and I175M mutations emerged simultaneously and conferred phenotypic resistance (13-fold increase in IC50). The Q91R+I175M combination was absent from all clinical isolates. Three-dimensional modeling indicated that residue 91 lies on the enzyme surface, at the entry of a pocket containing the DDE catalytic triad and that adding a positive charge (Gln to Arg) might compromise IN-RAL affinity. CONCLUSIONS: HIV-2 polymorphisms from 45 INI-naïve patients are described. Conserved regions as well as frequencies of HIV-2 IN polymorphisms were comparable to HIV-1. Two new mutations (Q91R and I175M) that conferred high resistance to RAL were selected in vitro, which might affect therapeutic outcome.
Project description:Failure of antiretroviral regimens containing elvitegravir (EVG) and raltegravir (RAL) can result in the appearance of integrase inhibitor (INI) drug-resistance mutations (DRMs). While several INI DRMs have been identified, the evolution of EVG DRMs and the linkage of these DRMs with protease inhibitor (PI) and reverse transcriptase inhibitor (RTI) DRMs have not been studied at the clonal level. We examined the development of INI DRMs in 10 patients failing EVG-containing regimens over time, and the linkage of INI DRMs with PI and RTI DRMs in these patients plus 6 RAL-treated patients. A one-step RT-nested PCR protocol was used to generate a 2.7 kB amplicon that included the PR, RT, and IN coding region, and standard cloning and sequencing techniques were used to determine DRMs in 1,277 clones (mean 21 clones per time point). Results showed all patients had multiple PI, NRTI, and/or NNRTI DRMs at baseline, but no primary INI DRM. EVG-treated patients developed from 2 to 6 strains with different primary INI DRMs as early as 2 weeks after initiation of treatment, predominantly as single mutations. The prevalence of these strains fluctuated and new strains, and/or strains with new combinations of INI DRMs, developed over time. Final failure samples (weeks 14 to 48) typically showed a dominant strain with multiple mutations or N155H alone. Single N155H or multiple mutations were also observed in RAL-treated patients at virologic failure. All patient strains showed evidence of INI DRM co-located with single or multiple PI and/or RTI DRMs on the same viral strand. Our study shows that EVG treatment can select for a number of distinct INI-resistant strains whose prevalence fluctuates over time. Continued appearance of new INI DRMs after initial INI failure suggests a potent, highly dynamic selection of INI resistant strains that is unaffected by co-location with PI and RTI DRMs.
Project description:Background:Integrase inhibitors (INI) induce a rapid decline of HIV-RNA in plasma and CD4+ T-cell recovery in blood. Both characteristics are also associated with immune reconstitution inflammatory syndrome (IRIS). Whether the use of INI-containing combination antiretroviral therapy (cART) increases the risk of IRIS is being questioned. Methods:Study within the Dutch ATHENA HIV observational cohort. HIV-1 infected late presenters initiating cART after March 2009 were included if they had <200 CD4+ T-cells per ?L and were diagnosed with an opportunistic infection. IRIS was defined either according to the criteria by French et al. (IRISFRENCH) or by a clinical IRIS diagnosis of the physician (IRISCLINICAL). The primary outcomes were the association between INI and the occurrence of IRISFRENCH and IRISFRENCH+CLINICAL in multivariable logistic regression. Findings:672 patients with a median CD4+ T-cell count of 35 cells per ?L were included. Treatment with INI was independently associated with IRISFRENCH as well as IRISFRENCH+CLINICAL (OR 2·43, 95%CI:1·45-4·07, and OR 2·17, 95%CI:1·45-3·25). When investigating INI separately, raltegravir (RAL) remained significantly associated with IRISFRENCH (OR 4·04 (95%CI:1·99-8·19) as well as IRISFRENCH+CLINICAL (OR 3·07, 95%CI:1·66-5·69), while dolutegravir (DTG) became associated with IRISFRENCH+CLINICAL after it replaced RAL as preferred INI in the cohort after 2015 (OR 4·08, 95%CI:0·99-16·82, p=0·052). Too few patients used elvitegravir to draw meaningful conclusions. Steroid initiation for IRIS was more likely in those who initiated INI versus in those who did not, but no increased hospital (re)admission or mortality rates were observed. Interpretation:In HIV late presenters from a resource rich setting, INI based treatment initiation increased the risk of IRIS. This was observed for RAL and DTG when being initiated as preferential INI in the presence of specific AIDS-conditions, indicative of channeling bias. Although we controlled for all relevant measured confounders, we cannot exclude that the observed association is partially explained by residual confounding. INI use was not associated with mortality nor hospitalization. Therefore, our observation is no reason to avoid INI in late presenters. Funding:The ATHENA database is maintained by Stichting HIV Monitoring and supported by a grant from the Dutch Ministry of Health, Welfare and Sport through the Centre for Infectious Disease Control of the National Institute for Public Health and the Environment.
Project description:Antiretroviral therapy for HIV-2 infection is hampered by intrinsic resistance to many of the drugs used to treat HIV-1. Limited studies suggest that the integrase inhibitors (INIs) raltegravir and elvitegravir have potent activity against HIV-2 in culture and in infected patients. There is a paucity of data on genotypic variation in HIV-2 integrase that might confer intrinsic or transmitted INI resistance.We PCR amplified and analyzed 122 HIV-2 integrase consensus sequences from 39 HIV-2-infected, INI-naive adults in Senegal, West Africa. We assessed genetic variation and canonical mutations known to confer INI-resistance in HIV-1.No amino acid-altering mutations were detected at sites known to be pivotal for INI resistance in HIV-1 (integrase positions 143, 148 and 155). Polymorphisms at several other HIV-1 INI resistance-associated sites were detected at positions 72, 95, 125, 154, 165, 201, 203, and 263 of the HIV-2 integrase protein.Emerging genotypic and phenotypic data suggest that HIV-2 is susceptible to the new class of HIV integrase inhibitors. We hypothesize that intrinsic HIV-2 integrase variation at "secondary" HIV-1 INI-resistance sites may affect the genetic barrier to HIV-2 INI resistance. Further studies will be needed to assess INI efficacy as part of combination antiretroviral therapy in HIV-2-infected patients.
Project description:BACKGROUND: HIV integrase inhibitor use is limited by low genetic barrier to resistance and possible cross-resistance among representatives of this class of antiretrovirals. The aim of this study was to analyse integrase sequence variability among antiretroviral treatment naive and experienced patients with no prior integrase inhibitor (InI) exposure and investigate development of the InI drug resistance mutations following the virologic failure of the raltegravir containing regimen. METHODS: Sequencing of HIV-1 integrase region from plasma samples of 80 integrase treatment naive patients and serial samples from 12 patients with observed virologic failure on raltegravir containing treatment whenever plasma vireamia exceeded >50 copies/ml was performed. Drug resistance mutations were called with Stanford DB database and grouped into major and minor variants. For subtyping bootstrapped phylogenetic analysis was used; Bayesian Monte Carlo Marcov Chain (MCMC) model was implemented to infer on the phylogenetic relationships between the serial sequences from patients failing on raltegravir. RESULTS: Majority of the integrase region sequences were classified as subtype B; the remaining ones being subtype D, C, G, as well as CRF01_AE , CRF02_AG and CRF13_cpx recombinants. No major integrase drug resistance mutations have been observed in InI-treatment naive patients. In 30 (38.5%) cases polymorphic variation with predominance of the E157Q mutation was observed. This mutation was more common among subtype B (26 cases, 54.2%) than non-B sequences (5 cases, 16.7%), p=0.00099, OR: 5.91 (95% CI:1.77-22.63)]. Other variants included L68V, L74IL, T97A, E138D, V151I, R263K. Among 12 (26.1%) raltegravir treated patients treatment failure was observed; major InI drug resistance mutations (G140S, Q148H and N155H, V151I, E92EQ, V151I, G163R) were noted in four of these cases (8.3% of the total InI-treated patients). Time to the development of drug resistance ranged from 2.6 to 16.3 months with mean increase of HIV viral load of 4.34 (95% CI:1.86-6.84) log HIV-RNA copies/ml at the time of emergence of the major mutations. Baseline polymorphisms, including E157Q were not associated with the virologic failure on raltegravir. CONCLUSIONS: In InI treatment naive patients polymorphic integrase sequence variation was common, with no major resistance mutants. In the treatment failing patients selection of drug resistance occurred rapidly and followed the typical drug resistance pathways. Preexisting integrase polymorphisms were not associated with the treatment failure.
Project description:Viruses were sequenced from 75 antiretroviral therapy (ARV)-naïve and from 75 ARV-treated patients who subsequently received a raltegravir-containing regimen. No major integrase inhibitor (INI)-resistance mutations were present in the 150 integrase (IN) sequences. Four ARV-naïve (5.3%) and two ARV-treated patients (2.7%) had one of the following minor INI-resistance mutations: L74M, E157Q, G163R, and R263K but there was no association between baseline raltegravir genotype and subsequent response to raltegravir treatment. We also combined our sequences with 4170 previously published group M IN sequences from INI-naïve patients to assess IN sequence variability and compared our findings with those of a study we performed in 2008 using data from 1563 patients. The number of polymorphic IN positions increased from 40% to 41% between the two studies. However, none of the major INI-resistance mutations was found to be polymorphic in either study and there were no significant changes in the prevalence of any of the minor INI-resistance mutations.
Project description:The aim of this study was to characterize the prevalence and patterns of genotypic integrase inhibitor (INI) resistance in relation to HIV-1 clade.The cohort comprised 533 INI-naive subjects and 255 raltegravir recipients with viraemia who underwent integrase sequencing in routine care across Europe, including 134/533 (25.1%) and 46/255 (18.0%), respectively, with non-B clades (A, C, D, F, G, CRF01, CRF02, other CRFs, complex).No major INI resistance-associated mutations (RAMs) occurred in INI-naive subjects. Among raltegravir recipients with viraemia (median 3523 HIV-1 RNA copies/mL), 113/255 (44.3%) had one or more major INI RAMs, most commonly N155H (45/255, 17.6%), Q148H/R/K?+?G140S/A (35/255, 13.7%) and Y143R/C/H (12/255, 4.7%). In addition, four (1.6%) raltegravir recipients showed novel mutations at recognized resistance sites (E92A, S147I, N155D, N155Q) and novel mutations at other integrase positions that were statistically associated with raltegravir exposure (K159Q/R, I161L/M/T/V, E170A/G). Comparing subtype B with non-B clades, Q148H/R/K occurred in 42/209 (20.1%) versus 2/46 (4.3%) subjects (P?=?0.009) and G140S/A occurred in 36/209 (17.2%) versus 1/46 (2.2%) subjects (P?=?0.005). Intermediate- to high-level cross-resistance to twice-daily dolutegravir was predicted in 40/255 (15.7%) subjects, more commonly in subtype B versus non-B clades (39/209, 18.7% versus 1/46, 2.2%; P?=?0.003). A glycine (G) to serine (S) substitution at integrase position 140 required one nucleotide change in subtype B and two nucleotide changes in all non-B clades.No major INI resistance mutations occurred in INI-naive subjects. Reduced occurrence of Q148H/R/K?+?G140S/A was seen in non-B clades versus subtype B, and was explained by the higher genetic barrier to the G140S mutation observed in all non-B clades analysed.
Project description:HIV-1 integrase is the third enzymatic target of antiretroviral (ARV) therapy. However, few data have been published on the distribution of naturally occurring amino acid variation in this enzyme. We therefore characterized the distribution of integrase variants among more than 1,800 published group M HIV-1 isolates from more than 1,500 integrase inhibitor (INI)-naïve individuals. Polymorphism rates equal or above 0.5% were found for 34% of the central core domain positions, 42% of the C-terminal domain positions, and 50% of the N-terminal domain positions. Among 727 ARV-naïve individuals in whom the complete pol gene was sequenced, integrase displayed significantly decreased inter- and intra-subtype diversity and a lower Shannon's entropy than protease or RT. All primary INI-resistance mutations with the exception of E157Q--which was present in 1.1% of sequences--were nonpolymorphic. Several accessory INI-resistance mutations including L74M, T97A, V151I, G163R, and S230N were also polymorphic with polymorphism rates ranging between 0.5% to 2.0%.
Project description:The failure of raltegravir (RAL) is generally associated with the selection of mutations at integrase position Y143, Q148, or N155. However, a relatively high proportion of failures occurs in the absence of these changes. Here, we report the phenotypic susceptibilities to RAL and elvitegravir (EVG) for a large group of HIV-infected patients failing on RAL-containing regimens. Plasma from HIV-infected individuals failing on RAL-containing regimens underwent genotypic and phenotypic resistance testing (Antivirogram v2.5.01; Virco). A control group of patients failing on other regimens was similarly tested. Sixty-one samples were analyzed, 40 of which belonged to patients failing on RAL-containing regimens. Full RAL susceptibility was found in 20/21 controls, while susceptibility to EVG was diminished in 8 subjects, with a median fold change (FC) of 2.5 (interquartile range [IQR], 2.1 to 3.1). Fourteen samples from patients with RAL failures showed diminished RAL susceptibility, with a median FC of 38.5 (IQR, 10.8 to 103.2). Primary integrase resistance mutations were found in 11 of these samples, displaying a median FC of 68.5 (IQR, 23.5 to 134.3). The remaining 3 samples showed a median FC of 2.5 (IQR, 2 to 2.7). EVG susceptibility was diminished in 19/40 samples from patients with RAL failures (median FC, 7.71 [IQR, 2.48 to 99.93]). Cross-resistance between RAL and EVG was high (R(2) = 0.8; P < 0.001), with drug susceptibility being more frequently reduced for EVG than for RAL (44.3% versus 24.6%; P = 0.035). Susceptibility to RAL and EVG is rarely affected in the absence of primary integrase resistance mutations. There is broad cross-resistance between RAL and EVG, which should preclude their sequential use. Resistance to EVG seems to be more frequent and might be more influenced by integrase variability.