HIV-1 strand transfer integrase strand inhibitors (INSTIs) are the recommended anchor class for current first-line antiretroviral therapy . Although the resistance profile is thought to be well established for the first-in-class raltegravir (RAL) and for the highly cross-resistant elvitegravir (EVG), the second generation INSTI dolutegravir (DTG) has been approved later and appears to have a partly different resistance profile, as well as a higher genetic barrier . This translates into lack of drug resistance at first-line failure with DTG-based therapy and possible rescue of RAL or EVG failure with DTG salvage therapy .
The three most commonly used genotype interpretation systems, namely the stanford HIV resistance database (HIVdb), the french agence nationale de recherche sur le sida (ANRS) and the belgian rega institute (REGA) systems, agree to consider the R263K mutation and the Q148/H/K/R and G140A/S and/or E138K combination as the major pathways leading to DTG resistance. However, as seen with other HIV-1 drug targets, a number of natural polymorphisms and minor drug-selected mutations can modulate the INSTI resistance pattern and result in unpredicted phenotype, particularly for the most recently introduced DTG. Indeed, a recent case report  showed failure to first-line RAL and second-line DTG therapy in a patient harbouring virus with a naturally occurring E157Q as the only INSTI-related mutation. In the same study, the patient-derived E157Q integrase was shown to confer approximately nine-fold resistance to DTG in a biochemical assay. Anstett et al. later characterized the E157Q site-directed mutant derived from the reference NL4-3 strain and reported that E157Q was not able to decrease susceptibility to DTG per se in a cell-free assay; however, E157Q was able to restore the infectiousness of the low fit R263K mutant, as well as to increase R263K-based resistance to DTG by 10-fold compared with R263K alone. Quite confusingly, E157Q is scored with potential low-level resistance to RAL and EVG, but not to DTG, by HIVdb; resistance to RAL and EVG and possible resistance to DTG by ANRS; and susceptible to all the three INSTIs by REGA.
To clarify the clinical role of E157Q when occurring alone, we retrieved six clinical samples from our routine drug resistance activity where HIV-1 integrase had only E157Q as an INSTI-related mutation based on the comprehensive HIVdb INSTI resistance mutation list. Access to these samples for research use was regulated by patient informed consent, as approved by the South-East Tuscany Ethical Committee. Only one of the six patients had previous INSTI experience. NL4-3 derived recombinant viruses harbouring the clinical integrase were constructed and used to measure phenotypic susceptibility to RAL and DTG through a previously described replication competent assay shown to correlate well with the reference Phenosense assay . The integrase sequences of the recombinant viruses have been submitted to GenBank with accession numbers MF445001 through MF445006. Table 1 indicates that RAL and DTG susceptibility fold changes were above two-fold in four cases for DTG but in no cases for RAL. Indeed, fold changes were significantly larger for DTG than for RAL (mean ± SD values of 0.94 ± 0.33 for RAL and 2.10 ± 0.87 for DTG; P = 0.013, paired t-test). However, based on the Phenosense assay susceptibility cut-off values (1.5 biological cut-off for RAL, four and 13 lower and upper clinical cut-offs for DTG, respectively), one recombinant virus would be labelled with borderline resistance to RAL (fold changes 1.5), whereas all the viruses would be considered as fully susceptible to DTG. Notably, the E157Q site-directed NL4-3 mutant showed virtually no change in susceptibility to RAL or DTG, supporting previous biochemical data . Attempts to detect signature polymorphisms possibly explaining the DTG-resistant phenotype described by Danion et al. upon alignment of all the E157Q sequences failed because of a low sample size in the context of extensive integrase variability.
According to the HIV Stanford database (https://hivdb.stanford.edu), the integrase E157Q variant is prevalent as a natural polymorphism in 1.4% to 7.4% of isolates, depending on subtype. In the most widely INSTI-treated subtype B, its prevalence is significantly enriched from 2.5% in INSTI-naïve to 6.5% in INSTI-treated patients (P < 0.0001). Of note, phenotypic resistance data obtained by the Phenosense assay on isolates harbouring E157Q alone show mean ± SD fold change values of 1.15 ± 0.31, 1.71 ± 0.54, and 1.12 ± 0.50 for the 17, 18 and only four cases available for RAL, EVG, and DTG, respectively. In addition, successful first-line treatment in study participants harbouring virus with E157Q alone in the integrase region has been specifically documented .
Overall, both in-vitro and in-vivo data indicate that the natural integrase polymorphism E157Q has minimal, if any, impact on susceptibility to INSTIs. The case of RAL and DTG resistance reported earlier by Danion et al. likely reflects a particular and rare combination of E157Q with other as yet unidentified polymorphisms. Nevertheless, it also suggests that genotypic prediction of INSTI susceptibility may occasionally fail and supports the use of phenotypic testing to investigate unexpected INSTI failures. It must be noted that in our dataset, five of six cases were derived from INSTI-naïve patients. INSTI treatment emergent E157Q may have a more relevant role, at least as a compensatory mutation in the context of recognized INSTI resistance mutations . Comparing integrase genotypes obtained at baseline and INSTI failure can be helpful to further understand the potential contribution to INSTI resistance for E157Q, as well as for other accessory mutations.
Dolutegravir was kindly provided by ViiV Healthcare. Raltegravir and the NL4-3 reference viruses were obtained through the AIDS Reagent Program. We thank our collaborators at HIV clinics for providing patients’ treatment data. This work has been presented at the 15th European Meeting on HIV and Hepatitis, Rome, 7–9 June 2017 and at the 9th Italian Conference on AIDS and Antiviral Research, Siena, 12–14 June 2017; Study design: F.S. and M.Z. Generation of recombinant viruses: D.T. and A.B. DNA sequencing: I.V. Phenotypic drug resistance assays: F.S. and A.G. First manuscript draft: F.S.
Conflicts of interest
There are no conflicts of interest.
2. Anstett K, Brenner B, Mesplede T, Wainberg MA. HIV drug resistance against strand transfer integrase inhibitors
3. Brenner BG, Wainberg MA. Clinical benefit of dolutegravir in HIV-1 management related to the high genetic barrier to drug resistance
. Virus Res
4. Danion F, Belissa E, Peytavin G, Thierry E, Lanternier F, Scemla A, et al. Non-virological response to a dolutegravir-containing regimen in a patient harbouring a E157Q-mutated virus in the integrase region
. J Antimicrob Chemother
5. Anstett K, Cutillas V, Fusco R, Mesplède T, Wainberg MA. Polymorphic substitution E157Q in HIV-1 integrase increases R263K-mediated dolutegravir resistance and decreases DNA binding activity
. J Antimicrob Chemother
6. Saladini F, Giannini A, Boccuto A, Vicenti I, Zazzi M. Agreement between an in-house replication competent and a reference replication defective recombinant virus assay for measuring phenotypic resistance to HIV-1 protease, reverse transcriptase, and integrase inhibitors
. J Clin Lab Anal
2017; [Epub ahead of print].
7. Ambrosioni J, Nicolás D, Manzardo C, Agüero F, Blanco JL, Mosquera MM, et al. Integrase strand-transfer inhibitor polymorphic and accessory resistance substitutions in patients with acute/recent HIV infection
. J Antimicrob Chemother
8. Marcelin AG, Grude M, Charpentier C, Bellcave P, Rodallec A, Pallier C, Delaugerre C. French national survey of resistance to integrase inhibitors shows high differences of resistance selection rate in case of virological failure in a context of routine hospital care (ANRS-AC11 virology network)
. J Int AIDS Soc
2016; 19 (suppl 7):17.