HIV-1 integrase resistance among antiretroviral treatment naive and experienced patients from Northwestern Poland

HIV integrase, being one of the key retroviral enzymes necessary for successful replication, is one of attractive targets in the treatment of HIV infection. Integrase inhibitors, targeting one of the essential steps of the virus life cycle, namely strand transfer have been approved for the clinical practice in 2008, and have proven to be highly efficient in treatment of both antiretroviral-naive and experienced individuals [1‐6]. As potent agents, this class of drugs is not only important part of the salvage regimens but is also useful in patients with therapy complications such as lipodystrophy, dyslipidaemia, or liver injury [7‐9]. Raltegravir (RAL) is currently licensed in Europe for both treatment naive and experienced patients, elvitregravir pending EMEA approval, while newer compounds such as dolutegravir undergo phase III clinical trials [10‐13].

Despite high efficacy observed among treatment experienced patients with drug resistance to other antiretroviral drug classes, low genetic barrier to resistance and possible cross-resistance among integrase inhibitors is a limiting factor in the practical use of these compounds. Virologic failure has been associated with major, signature mutations within the catalytic domain of the enzyme, and include Y143R/C, N155H Q148K/R/H integrase sequence variants associated with significant susceptibility reduction both to RAL and elvitegravir (EVG) [14‐19]. Accumulation of secondary, accessory mutants is responsible for the further increase in the level of the resistance and often restored replication capacity [20‐22]. Major mutations remain uncommon in the antiretroviral treatment naive patients; so far only a few cases of transmitted drug resistance (Q148H, G140S and N155H mutations) have been described [23, 24]. On the other hand, polymorphic mutations in the central core domain positions have been observed in up to 34% of the published sequences [25] and 56% of the patients with recently acquired infection [26]; some of these naturally occurring variants have been observed in patients failing raltegravir and elvitegravir (L74M, T97A, S119G/R, E157Q, G163K/R), with notable the frequency variation across the subtypes [4, 25, 27‐31]. Recent reports also describe high frequency of the minority clades bearing major and accessory mutations, however, clinical significance of this pre-existing low level variability is yet to be determined [20, 27, 32]. Analyses for the secondary drug resistance are currently included in the guidelines for the drug resistance testing among individuals failing integrase containing treatment [33]. This study was designed to investigate the sequence variability in the integrase region with two objectives: firstly, to characterize primary integrase resistance mutations among the treatment naive and experienced patients with no prior integrase inhibitor (InI) exposure; secondly to investigate the development of the InI drug resistance mutations following the virologic failure of the raltegravir-containing regimen.

Integrase inhibitors remain an attractive option and have become a vital component of the modern antiretroviral treatment, especially among patients with preexisting drug resistance or treatment complications [2, 40‐45]. It is necessary to monitor the transmission and de novo development of drug resistance mutations decreasing the susceptibility of HIV against this class of antiretrovirals to provide the virologists and clinicians with the current data allowing for adequate therapeutic strategies.

In the presented study, among the InI naive patients no major drug resistance mutations have been observed, however, accessory mutations have been common (38.5%). Of the noted variants, four (L68V, T97A, V151I, E157Q) have previously been described as polymorphic, occurring in >1% of integrase sequences [25]. These integrase mutations were more prevalent in the subtype B viruses with E157Q occurring at higher frequencies than previously reported [46‐50]. In the previous reports this polymorphism was shown to impair the integrase 3′end processing and strand transfer [17] but was associated with only minimal reduction of the susceptibility to RAL and elvitegravir (<6 fold) [25, 26, 51, 52]. Of note, in our study the E157Q mutation was observed mostly among phylogenetically related subtype B infected intravenous drug users (Figure 1A), and was not associated with higher ratio of the virological failure. This is the largest described so far cluster with this polymorphism in subtype B infected patients. Lack of clinically important, primary resistance mutations for raltegravir, elvitegravir and dolutegravir is consistent with published reports from other studies [27, 47, 53‐59], and is supports the fact that the transmission of the drug resistance is unlikely in the populations previously unexposed to the integrase inhibitor treatment [60].

In the group of the RAL treated patients treatment was successful in 73.9% of cases, while the 8.3% of the failing patients developed major drug resistance mutations significantly reducing susceptibility to both raltegravir and elvitegravir. Number of virologic failures was higher than observed in the STARTMRK and BENCHMRK trials for the patients with high baseline genotypic scores [1, 5, 6, 15], however, in the analysed group integrase inhibitor was usually considered as a second line treatment and was selected due to toxicity, drug resistance or drug-drug interactions, as well as preexisting drug resistance in treatment experienced patients; suboptimal adherence was also commonly noted - only in three (one described below with developed integrase resistance mutations and two with integrase-inhibitor susceptible variants) failing patients estimated adherence exceeded 90%. Another reason for the high frequency of treatment failure on raltegravir-containing regimen may be related to the preexisting drug resistance mutations in the reverse transcriptase region resulting in the lower susceptibility to the background regimen. NRTI drug resistance mutations (thymidyne analog mutations and M184V) were present in six (50%) of the twelve treatment failing patients prior to RAL introduction. Additionally, treatment efficacy was numerically worse among non-B subtypes (with borderline statistical significance), however this may be associated with small group size as in the previous reports raltegravir was comparably efficacious across B and non-B HIV-1 subtypes [4].

Integrase drug resistance profiles in four treatment failing patients with drug resistance followed two typical pathways, Q148H/G140S (one case) and N155H/V151I (three patients), associated with high level resistance to raltegravir and cross-resistance to elvitegravir [16‐18, 61, 62]. Similar cases have been described previously [19, 61, 63]. Of note, in the patient with the Q148H/G140S salvage treatment option with dolutegravir was also lost due to the resistance associated with this pathway [64]. As expected, in cases observed in our study virologic rebound was associated with significant increase of the plasma viral load at the time-point of the occurrence of the major mutation. Stepwise accumulation of mutations was observed in two patients with N155H followed by the secondary E92Q, V151I and/or G163R mutations resulting in further increase raltegravir and elvitegravir resistance and often enhanced replicative capacity [27, 65]. All observed mutations have been previously associated with raltegravir drug resistance observed in in vivo studies [15, 45]. The N155H pathway is also associated with smaller reduction in the RAL susceptibility than the Q148H, moreover, double Q148H/G140S mutants were found to be fitter than the E92Q/N155H ones [66, 67]. This is consistent with the observed higher viral loads at the time of the development of drug resistance in a patient 2 (if compared to the patients 1,3,4) in whom such a highly fit double mutant was observed (Figure 3). In the patient 3 the observed drug resistance is probably associated with the suboptimal treatment and poor drug exposure when combined only with boosted saqinavir, however no other treatment option was possible for the patient at the time of therapy initiation . Time to InI resistance development in the observed patients did not exceed few months, except for one case with the double Q148H/G140S mutant, which is in accordance to the data on the low genetic barrier for the integrase inhibitors and rapid selection of the resistant variants [15, 68].

In summary, the results of this study provide data for the clinical practice and the treatment with integrase inhibitors. No major mutations associated with integrase dug resistance have been found in the pretreatment samples, however the polymorphic variation was common. In the failing patients selection of drug resistance occurred rapidly, and followed the typical pathways with accumulation of the drug resistance. Poor adherence, preexisting drug resistance and suboptimal combination were the key reasons associated with the development of the drug resistance.