Nonnucleoside reverse transcriptase inhibitors (NNRTI) are popular components of combination antiretroviral therapy. Despite its proven efficacy, the clinical use of first-generation NNRTI, as nevirapine (NVP) and efavirenz (EFV), has been limited by side effects, low barrier to resistance and broad cross-resistance. To try to overcome these limitations, a second-generation of NNRTI has been developed that includes etravirine (ETR) and rilpivirine (RPV), both of which were recently approved as therapy for HIV-1 infection.
The approval of RPV was based on the results from the ‘Rilpivirine versus efavirenz with tenofovir and emtricitabine in treatment-naive adults infected with HIV-1 (ECHO): a phase 3 randomised double-blind active-controlled trial’ and the ‘Rilpivirine versus efavirenz with two background nucleoside or nucleotide reverse transcriptase inhibitors in treatment-naive adults infected with HIV-1 (THRIVE): a phase 3, randomized, non-inferiority trial’, which assessed the efficacy and safety of the drug in nearly 1400 antiretroviral-naive patients [1–3]. Although the resistance profile for RPV has not been well defined yet, there is information suggesting that the susceptibility to the drug is not compromised or only marginally affected by the presence of single NNRTI resistance-associated mutations (RAMs). A total of 15 changes at the HIV-1 reverse transcriptase gene have been associated with a decreased susceptibility to RPV . By far, mutation E138K was the most frequently selected (45%) in antiretroviral-naive patients that failed on RPV therapy in ECHO and THRIVE studies. Interestingly, this change was generally seen along with M184I (34%), which confers lamivudine and emtricitabine resistance . The combination E138K/M184I confers a 6.7-fold reduced phenotypic susceptibility to RPV compared with a 2.8-fold reduction for E138K alone. Mutation K103N, which is associated with clinical resistance to EFV and NVP, does not reduce susceptibility to RPV.
There is scarce information about the rate of RPV RAMs in HIV-1-infected patients with prior history of NNRTI failure. Likewise, very few studies have examined the clinical outcome of patients harbouring NNRTI-resistant viruses that subsequently received RPV . Drug resistance interpretation systems for antiretroviral agents (i.e., Stanford, ANRS, and so on) have recently incorporated predictions of virological response to RPV based on the available information derived from the ECHO and THRIVE trials, from in-vitro studies and from expert opinion. The Drug Resistance Platform of the Spanish AIDS Research Network (http://www.retic-ris.net) has weighted NNRTI RAMs , and for considering resistance to RPV at least two RT mutations must be present. Changes with the greatest impact on RPV susceptibility are at four codons (K101E/P/T, E138A/G/K/R, Y181C/I/V and M230L), whereas changes at other nine positions display a lower impact (V90I, L100I, V106A/I, V108I, V179F/I/L, Y188I, G190E, H221Y and F227C/L). However, in the presence of M184I, only one of two changes (either E138K or K101E) is enough to produce high-level RPV resistance. This information is important for clinicians, particularly when simplification strategies using coformulations with RPV or rescue interventions in patients failing on NVP, EFV or ETR are being considered.
The aim of this study was to examine the rate of RPV RAMs and the proportion of estimated RPV resistance in HIV-1-infected patients who had failed other NNRTI-based regimens in a large network of HIV-1 clinics in Spain.
Patients and methods
The RT genotypes and clinical information from all HIV-1-infected patients on regular follow-up at 22 different HIV clinics in Spain who had failed NNRTI-based regimens were identified at the Spanish national resistance database (ResRIS) [6,7]. This is a large clinical database that records information from HIV-1 patients treated outside clinical trials. Data recorded includes drug resistance mutations, antiretroviral therapy, HIV clade, viral load and CD4 cell counts. Ultimately it produces back a virtual interpretation of the resistance mutation profile for all antiretroviral agents for a given sample, which is then send back to clinicians.
Drug resistance mutations and interpretation
The prevalence of RPV RAMs as well as the proportion of estimated RPV-resistant samples, as reported using the ResRis national algorithm (http://www.retic-ris.net), were assessed in the whole population of HIV-1 patients that had failed on NNRTI-based regimens. Drug resistance mutations were examined taking into account the updated mutation list from the IAS-USA panel (December 2011), also other recent changes that have been highlighted from the ECHO and THRIVE trials as well as in-vitro studies . Briefly, these changes are the following: V90I, L100I, K101E/P/T, V106A/I, V108I, E138A/G/K/Q/R, V179F/I/L, Y181C/I/V, Y188I, and M230L. All these changes are considered in the current Spanish resistance interpretation algorithm, which additionally provides a weighting impact for each mutation (Table 1).
All results are expressed as absolute numbers and percentages. The prevalence of RPV RAMs in patients who had failed on EFV, NVP or ETR was compared using χ2 tests. Significant differences were only considered for P values below 0.05. All statistical analyses were performed using SPSS v15.0 (SPSS Inc., North Chicago, Illinois, USA).
From a total of 8200 RT genotypes derived from 5873 different HIV-1 individuals recorded at ResRIS, 1064 belonged to HIV-1-infected patients that had failed NNRTI-based regimens. Overall, 27.1% (n = 288) of specimens did not harbour any NNRTI RAMs. Among the 1064 genotypes examined, 580 (54.5%) had failed on EFV, 426 (40%) on NVP and 58 (5.5%) on ETR. Up to 45.9% (n = 488) were on their first NNRTI treatment, 39.8% (n = 424) had previously been exposed to another NNRTI and 14.3% (n = 152) had received two NNRTI.
Figure 1 records the prevalence of distinct RPV RAMs in the study population. The most prevalent mutations were Y181C (21.8%), V108I (10.2%), K101E (9.1%), V90I (7.9%) and V179I (6.1%). All other RPV RAMs examined were present at rates below 5%, being mutations E138R and Y188I absent in our study population. Only three patients (0.3%) harboured mutation E138K, and two of them had failed on ETR. The NRTI resistance mutation M184I was present in 3.4% of the whole genotypes, whereas M184V was seen in 36.2%. The combination K101E/M184I was seen in 1% of specimens, being absent E138K/M184I.
Mutations L100I and V108I were significantly more frequent in patients failing on EFV than NVP (7.9 vs. 0.2 and 12.2 vs. 7.3%, respectively). Conversely, Y181C, Y181I, V106A, H221Y and F227L were significantly more prevalent in patients failing on NVP than EFV. Interestingly, the lamivudine/emtricitabine RAM M184V was more frequent in patients failing on NVP than EFV (43.7 vs. 32.1%; P < 0.001). Finally, changes at positions V90I, E138K, V179I and Y181C were more common in patients failing on ETR than EFV (P < 0.05).
Based on the virtual reports produced by the Spanish drug resistance interpretation system, a total of 206 genotypes (19.3%) from patients failing NNRTIs should be considered as RPV-resistant. They corresponded to failures on NVP (51.5%), EFV (40.8%) or ETR (7.8%). When the proportion of RPV resistance was considered for distinct NNRTI failures, figures were as follows: 14.5% for EFV, 25% for NVP and 27.6% for ETR. Of note, cross-resistance between RPV and ETR (27.6% in 58 ETR failures) was reported mainly as result of changes Y181C (81.3%), V179I (43.8%), V90I (31.3%) and V108I (18.8%).
RPV is the latest approved NNRTI for the treatment of HIV-1 infection . Virologic responses as well as selection of drug resistance up to week 96 have recently been reported for ECHO and THRIVE trials . The combination E138K + M184I was the most frequently selected failing on RPV (44.2%) in this population .
In ResRis, the prevalence of codon 138 mutants was very rare (1%) in patients that had failed NNRTI-based regimens other than RPV. Moreover, the combination E138K + M184I was absent in this population. These results are in agreement with those recently reported by German authors that tested a group of antiretroviral-experienced patients and found a very low rate (0.5%) of E138K . Altogether, these results support that this mutation is not selected in patients failing on NVP and EFV and should be considered as specifically selected by ETR . Taking into account that 39.8% of specimens tested come from individuals on their second NNRTI and 14.3% on their third NNRTI, the three individuals we found with viruses harbouring E138K were retrospectively reassessed, and two of them were found to have recently received ETR.
It must be noted that RPV and ETR largely share their respective resistance profile, and that E138K has recently been added to the ETR genotypic score . In our study, 27.6% of 58 ETR failures harboured mutant viruses interpreted as RPV-resistant. Interestingly, loss of RPV susceptibility was mainly interpreted as a result of selecting changes other than E138K, as Y181C, V179I, V90I and/or V108I.
Although RPV has so far been approved as first-line treatment for HIV-1 infection, the drug is currently being considered in other clinical scenarios, such as in simplification strategies or in rescue interventions , given its good tolerability and easy to take coformulation as a single-tablet regimen. In-vitro data per se are not enough to predict clinical response, but they could support that RPV would be active following EFV failure, acknowledging minor overlap in selected drug resistance mutations; however, clinical data proving this assumption are scarce . To validate the chances of any clinical benefit of RPV based on drug resistance genotyping following NNRTI failure, it would be worth collecting more clinical data in this specific scenario. However, our study is the first to support this hypothesis in a relatively large number of patients examined outside clinical trials. Only 14.5% of 580 EFV failures in our study were considered as RPV-resistant. In contrast, this figure was 25% for 426 NVP failures. Anyway, these rates are not negligible, and, therefore, drug resistance testing should be recommended before considering RPV therapy in patients that had failed on other NNRTIs. Prospective studies evaluating the efficacy of RPV in patients who have failed on EFV, NPV or ETR should be conducted.
Although phenotypic drug resistance testing could be useful to determine the susceptibility of recently approved antiretrovirals, for which the genetic correlates of clinically relevant drug resistance have not yet been well characterized, the situation is distinct for NNRTI. So far, the phenotypic susceptibility data have poorly predicted the efficacy of most NNRTI, as generally these drugs behave as on-off, being active or not, with no room for clinically relevant partial activity. In this situation, genotypic tests perform the best and facilitate the interpretation of mutations in drug resistance algorithms. However, the situation may be different for RPV. There is still scarce information about the distinct weight of mutations leading to resistance and the initial list of RPV RAMs must be refined.
In summary, the rate of E138 mutants is very rare in individuals failing NNRTI-based regimens other than RPV in ResRIS. Almost 20% of patients failing NNRTIs should be considered as RPV resistant, as a result of selecting changes at other positions. The extent of cross-resistance seems to be higher for ETR and NVP in comparison with EFV, but prospective clinical trials should confirm the clinical value of this observation. The sequential use of RPV in patients that had failed other NNRTIs should not be done in the absence of drug resistance testing excluding cross resistance. Nevertheless, analyses of virological responses in patients treated with RPV following failures to NNRTI are needed and will provide more robust evidence about the impact of distinct resistance changes on RPV activity.
The study design and data analysis was made by L.A., C.d.M., V.S. and J.M.L. The manuscript draft was written by L.A., C.d.M. and V.S. Critical reading was contributed by J.M.L. (H. Germans Trias i Pujol, Badalona), E.P. (H. Carlos III, Madrid), J.L.B. (H. Clinic, Barcelona), M.Á. (H. Clínico San Cecilio, Granada), M.J.P.E. (H. Ramón y Cajal, Madrid), A.A. (H. Conxo, Santiago de Compostela) and E.C. (H. Vall d’Hebrón, Barcelona).
Members of the Drug Resistance Platform of the Spanish AIDS Research Network (Red de Investigación en SIDA, RIS): JA Iribarren, Hospital de Donostia, San Sebastián; JL Blanco and JM Gatell, Hospital Clinic, Barcelona; E Caballero and E Ribera, Hospital Vall d’Hebron, Barcelona; JM Llibre, J Martínez-Picado and B Clotet. ICREA, Fundación IrsiCaixa, Hospital Germans Trias I Pujol, Badalona; A Jaén and D Dalmau, Hospital Universitari Mútua Terrassa, Terrassa; J Peraire and F Vidal, Hospital Joan XXIII, Tarragona; C Vidal and M Riera, Hospital Son Espases, Palma de Mallorca; J Córdoba and J López-Aldeguer, Hospital La Fe, Valencia; MJ Galindo, Hospital Clínico Universitario, Valencia; C Robledano and F Gutiérrez. Hospital General, Elche; M Álvarez and F García, Hospital Clínico San Cecilio, Granada; I Viciana and J Santos, Hospital Vírgen de la Victoria, Málaga; P Pérez-Romero and M Leal, Hospital Vírgen del Rocío, Sevilla; JA Pineda, Hospital de Valme, Sevilla; F Fernández-Cuenca, Hospital Vírgen Macarena, Sevilla; C Rodríguez and J del Romero, Centro Sanitario Sandoval, Madrid; L Menéndez-Arias, Centro de Biología Molecular Severo Ochoa CSIC-UAM, Madrid; MJ Pérez-Elías, C Gutiérrez and S Moreno, Hospital Ramón y Cajal, Madrid; M Pérez-Olmeda and J Alcamí, Instituto de Salud Carlos III, Madrid; A Cañizares and J Pedreira, Hospital Juan Canalejo, La Coruña; C Miralles and A Ocampo, Hospital Xeral-Cíes, Vigo; L Morano, Hospital Meixoeiro, Vigo; JJ Rodríguez-Calviño and A Aguilera, Hospital de Conxo-CHUS, Santiago de Compostela; JL Gómez-Sirvent, Hospital Universitario de Canarias, Santa Cruz de Tenerife; L Anta, E Poveda, V Soriano and C de Mendoza, Hospital Carlos III, Madrid, Spain.
This work was supported in part by grants from Red de Investigación en SIDA (RIS) RD06/0006, European Commission project CHAIN (FP7/2007–2013; grant 223131), and Fundación Investigación y Educación en SIDA (F-IES).
Conflicts of interest
There are no conflicts of interest.
1. Molina JM, Cahn P, Grinsztejn B, Lazarin A, Mills A, Saag M, et al. Rilpivirine versus efavirenz with tenofovir and emtricitabine in treatment-naive adults infected with HIV-1 (ECHO): a phase 3 randomized double-blind active-controlled trial
2. Cohen CJ, Andrade-Villanueva J, Clotet B, Fourie J, Johnson M, Ruxrungtham K, et al. Rilpivirine versus efavirenz with two background nucleoside or nucleotide reverse transcriptase inhibitors in treatment-naive adults infected with HIV-1 (THRIVE): a phase 3, randomised, noninferiority trial
3. Fernández-Montero JV, Vispo E, Anta L, de Mendoza C, Soriano V. Rilpivirine: a next-generation nonnucleoside analogue for the treatment of HIV infection
. Expert Opin Pharmacother
4. Johnson V, Calvez V, Günthard H, Paredes R, Pillay D, Shafer R, et al. Update of the drug resistance mutations in HIV-1: December 2011
. Top HIV Med
5. Rimsky L, Vingerhoets J, Van Eygen V, Eron J, Clotet B, Hoogstoel A, et al. Genotypic and phenotypic characterization of HIV-1 isolates obtained from patients on rilpivirine therapy experiencing virologic failure in the phase 3 ECHO and THRIVE studies: 48-week analysis
. J Acquir Immune Defic Syndr
6. de Mendoza C, Anta L, García F, Pérez-Elías MJ, Gutiérrez F, Llibre JM, et al. HIV-1 genotypic drug resistance interpretation rules: 2009 Spanish guidelines
. AIDS Rev
7. Anta L, Blanco JL, Pérez-Elías MJ, Garcia F, Leal M, Ribera E, et al.Applications of a national HIV drug resistance database: surveillance and development of genotypic resistance algorithms.
7th European HIV Drug Resistance
2009 [Abstract 16].
8. Azijn H, Tirry I, Vingerhoets J, de Béthune MP, Kraus G, Boven K, et al. TMC278, a next-generation nonnucleoside reverse transcriptase inhibitor (NNRTI), active against wild-type and NNRTI-resistant HIV-1
. Antimicrob Agents Chemother
9. Rimsky L, Van Eygen V, Vingerhoets J, Hoogstoel A, Stevens M, Boven K, et al.Week 96 resistance analysis of the rilpivirine (RPV, TMC278) phase III trials in treatment-naive HIV-infected adults
[abstract 708]. In: Program and Abstracts of the 19th Conference on Retroviruses and Opportunistic Infections;
5–8 March 2012; Seattle. Washington, USA.
10. Reinheimer C, Doerr H, Stürmer M. Prevalence of TMC278 (rilpivirine) associated mutations in the Frankfurt Resistance Database
. J Clin Virol
11. Tambuyzer L, Nijs S, Daems B, Picchio G, Vingerhoets J. Effect of mutations at position E138 in HIV-1 reverse transcriptase on phenotypic susceptibility and virologic response to etravirine
. J Acquir Immune Defic Syndr
12. Vingerhoets J, Tambuyzer L, Azijn H, Hoogstoel A, Nijs S, Peeters M, et al. Resistance profile of etravirine: combined analysis of baseline genotypic and phenotypic data from the randomized, controlled Phase III clinical studies
13. Bunupuradah T, Ananworanich J, Chetchotisakd P, Kantipong P, Jirajariyavej S, Sirivichayakul S, et al. Etravirine and rilpivirine resistance in HIV-1 subtype CRF01_AE-infected adults failing nonnucleoside reverse transcriptase inhibitor-based regimens
. Antivir Ther
14. Cohen C, Molina J, Cahn P, Clotet B, Fourie J, Grinsztejn B, et al. Efficacy and safety of rilpivirine (TMC278) versus efavirenz at 48 weeks in treatment-naïve, HIV-1-infected patients: Pooled results from the phase 3 double-blind, randomized ECHO and THRIVE trials
. J Acquir Immune Defic Syndr