Antiviral Therapy 6: 21–40
Laboratory guidelines for the practical use of HIV drug resistance tests in patient follow-up A-M Vandamme1*, F Houyez2, D Bànhegyi3, B Clotet4, G De Schrijver5, KAL De Smet6, WW Hall7, R Harrigan8, N Hellmann9, K Hertogs10, C Holtzer11, B Larder10, D Pillay12, E Race13, J-C Schmit14, R Schuurman15, E Shulse16, A Sönnerborg17 and V Miller18 1
AIDS Reference Laboratory, Rega Institute and University Hospitals, KU Leuven, Leuven, Belgium Vaincre Le Sida, Paris, France 3 Szent Lâszlo Korhaz, Budapest, Hungary 4 Fundacio IRSI-Caixa, Badalona, Spain 5 Lage Heirweg 21, 9830 Sint Martens Latem, Belgium 6 Innogenetics, Gent, Belgium 7 Virus Reference Laboratory, University College, Dublin, Ireland 8 BC Center for Excellence in HIV/AIDS, Vancouver, Canada (currently at Virco, UK) 9 ViroLogic, South San Francisco, USA 10 Virco, Mechelen, Belgium and Virco, UK 11 Visible Genetics, Paris, France 12 PHLS Antiviral Susceptibility Reference Unit, University of Birmingham Medical School, UK 13 Hôpital Bichat-Claude Bernard, France (currently at VIRalliance, France) 14 Centre Hospitalier de Luxembourg, Luxembourg 15 Eijkman-Winkler Institute, Utrecht University, Utrecht, The Netherlands 16 Applied Biosystems, Foster City, Calif., USA 17 Karolinska Institutet, Huddinge, Sweden 18 Klinikum der JW Goethe Universität, Frankfurt, Germany 2
*Corresponding author: Tel: +32 16 332 160; Fax: +32 16 332 131; E-mail:
[email protected]
HIV drug resistance is one of the major limitations in the successful treatment of HIV-infected patients using currently available antiretroviral combination therapies. When appropriate, drug susceptibility profiles should be taken into consideration in the choice of a specific combination therapy. Guidelines recommending resistance testing in certain circumstances have been issued. Many clinicians have access to resistance testing and will increasingly use these results in their treatment decisions. In this document, we comment on the different methodologies available, and the relevant issues relating to the clinical application of these tests. Specifically the following recommendations can be made: (i) genotypic and phenotypic HIV-1 drug resistance analyses can yield complementary information for the clinician. However, insufficient information currently exists as to which approach is preferable in any particular clinical setting; (ii) when HIV-1 drug resistance testing is required, it is
recommended that testing be performed on plasma samples obtained before starting, stopping or changing therapy, on samples that have a viral load above the detection limit of the resistance test; (iii) the panel recommends that genotypic and phenotypic HIV-1 drug resistance testing for clinical purposes be performed in a certified laboratory under strict quality control and quality assurance standards; and (iv) the panel recommends that resistance testing laboratories provide clinicians with resistance reports that include a list of drug-related resistance mutations (genotype) and/or a list of drug-related fold resistance values (phenotype), with interpretations of each by an experienced virologist. The interpretation of genotypic and phenotypic analysis is a complex and developing science, and in order to understand HIV-1 drug resistance reports, communication between the requesting clinician and the expert that interpreted the resistance report is recommended.
Introduction The rapid development of drug resistance is one of the principle reasons for current recommendations to employ combination antiviral therapy in the treatment of HIV infection. Although these more potent drug ©2001 International Medical Press 1359-6535/01/$17.00
combination therapies may be able to delay the emergence of virus drug resistance, there are no data supporting the view that they will indefinitely prevent resistance from arising. One way to manage drug resis21
A-M Vandamme et al.
tance is to include resistance testing in current patient management and follow-up. The EuroGuidelines group consists of representative academic experts from each European country (clinicians and virologists), representatives from industry involved in the development of assays for resistance testing or of new antiviral agents, as well as HIVcommunity representatives. This panel gathered on several occasions in 1999 and 2000, working in four groups, to produce documents discussing the current best use of drug-resistance assays for patient follow-up. The panel mission and goal was to develop guidelines for the use of resistance testing in the clinical setting for the European HIV population (workgroup 1); to generate recommendations related to methodological aspects, quality assurance in the generation and reporting of resistance results in Europe (workgroup 2); to assess the cost-effectiveness of resistance testing (workgroup 3); and to recommend steps to facilitate the implementation of the guidelines as standard of care for all European patients (workgroup 4). The guidelines are to be adapted as new information becomes available. The document presented here is the consensus of workgroup 2, reviewing the currently available resistance assays, and formulating some practical recommendations for their use. A summary of the recommendations Table 1. Concise guidelines (i) Which resistance test to use? Genotypic and phenotypic HIV-1 drug resistance analyses can yield complementary information for the clinician. However, insufficient information currently exists as to which approach is preferable in any particular clinical setting. (ii) Which sample to take? When HIV-1 drug resistance testing is required, the panel recommends that testing be performed on plasma samples obtained before starting, stopping or changing therapy and which have a viral load above the detection limit of the resistance test. (iii) Where to test the sample? The panel recommends that genotypic and phenotypic HIV-1 drug resistance testing for clinical purposes be performed in a certified laboratory under strict quality control and quality assurance standards. (iv) How to interpret the results? The panel recommends that resistance testing laboratories provide clinicians with resistance reports that include a list of drug-related resistance mutations (genotype) and/or a list of drug-related fold resistance values (phenotype), with proper interpretations by an expert virologist for both. The interpretation of genotypic and phenotypic analysis is a complex and developing science. For proper understanding of an HIV-1 drug resistance report, communication between the requesting clinician and the expert that interpreted the resistance report is recommended. 22
of this workgroup is presented in Table 1. These recommendations are mainly addressed to laboratories engaging in HIV drug resistance testing, but may also be of interest to clinicians ordering drug-resistance evaluations, since knowledge of the laboratory considerations will improve their appreciation of the resistance report. This document is intended as a support document for the ‘Clinical and laboratory guidelines for the use of HIV-1 drug resistance testing as part of treatment management – recommendations for the European setting’, which are published separately [1]. Other documents of the panel include a study on the cost-effectiveness of resistance testing [2] and on the need for specific European guidelines [3]. Updated information can be found on the EuroGuidelines website (http://www.euroguidelines.org).
(i) Genotypic and phenotypic HIV-1 drug resistance analyses can yield complementary information for the clinician. However, insufficient information currently exists as to which approach is preferable in any particular clinical setting For many patients, optimizing anti-HIV treatment will require resistance testing. Tables 2 and 3 summarize the currently available assays and their characteristics. Two basic approaches of assessing resistance are used; genotypic and phenotypic tests. The resistance genotype is assessed by identifying genotypic mutations resulting in amino acid changes that have been reported to correlate with reduced susceptibility to a particular drug. The phenotype is a direct measure of the drug susceptibility, identifying the drug concentration necessary to inhibit virus replication, usually by 50% (IC50). Genotypic assays measure the presence of resistance-related mutations either by direct PCR-sequencing, by point mutation assays such as selective PCR, by hybridization assays or by single nucleotide sequencing [4–8] (Tables 2 and 3). Firstly, all genotypic assays require the extraction of the virus genome (viral RNA or proviral DNA). Viral RNA has to be reverse transcribed in cDNA, subsequently the DNA is amplified in a single or nested PCR (Table 3). A complete genotype of the protease and reverse transcriptase, obtained by sequencing, detects resistance-related mutations for all available drugs simultaneously. In selective PCR, primers that selectively amplify either the wild-type (WT) or mutant (MT) variant at a particular position are used, a technique that is often referred to as the Amplification Refractory Mutation System (ARMS). Hybridization assays are based on the selective detection of PCR amplified protease (PRO) or reverse transcriptase (RT) ©2001 International Medical Press
Guidelines for use of HIV drug resistance testing
Table 2. Resistance tests presently available in Europe Company (contact details)
Name of the test
Type of test
Technology used
RT genotype
Probe hybridization
PRO genotype PRO + RT genotype PRO + RT genotype PRO + RT genotype
Probe hybridization Probe hybridization (Gene Chip) Sequencing Sequencing
VirtualPhenotype
PRO + RT genotype
GENcheq Antivirogram GeneSeq HIV PhenoSense HIV Guidelines report PGL regular test PGL SO-LO test (low copy number test)
PRO + RT genotype PRO + RT phenotype PRO + RT genotype PRO + RT phenotype PRO + RT genotype PRO + RT genotype
Sequencing linked to database interpretation Sequencing Recombinant virus assay Sequencing Recombinant virus assay Sequencing Sequencing
In house ± commercial kit
PRO and/or RT genotype
Companies providing HIV drug resistance testing kits Innogenetics (www.innogenetics.com)
Affymetrix (www.affymetrix.com) Visible Genetics (www.visgen.com) Applied Biosystems (www.perkin-elmer.com)
Inno-LiPA HIV key nucleosides; Inno-LiPA HIV key non-nucleosides; Inno-LiPA HIV key MDR Inno-LiPA HIV key protease HIV PRT-440 TruGene HIV-1 Genotyping Kit ViroSeq Genotyping System
Commercial laboratories providing HIV drug resistance testing services Virco (www.vircolab.com)
ViroLogic Inc. (www.virologic.com) ACTGene/Visible Genetics (
[email protected]) Professional Genetic Laboratory (PGL)
Local AIDS reference laboratories National agencies
In house
genes using WT- or MT-specific probes on membranes (LiPA) or on an array of probes coated on chips after which the hybridization pattern is interpreted by specific software (Gene-Chip). Comparisons between different genotyping methods including the two commercial sequencing kits and two commercial reverse-hybridization assays show a relatively good concordance, with discrepancies mainly resulting from different scoring of mixtures [9–11]. The resistance phenotype for a particular drug can be measured using replication-based or enzyme-based assays [2,12,13]. The phenotype can be monitored with the EC50 or IC50, which is the concentration of drug required to reduce virus replication or enzymatic activity by 50%. Fold-resistance is generally calculated as the ratio of EC50 for the resistant patient isolate to the EC50 for a sensitive laboratory strain. In clinical practice, different variants of the recombinant virus assay are currently used [2,14–19] (Table 1 and 2). In this assay, the patient’s virus derived gene under investigation is amplified starting with viral RNA (as it is for genotypic assays, Table 3) and recombined through cloning and/or co-transfection with a laboratory viral clone lacking the gene under investigation. In this way the resistance-related characteristics of the patient’s virus derived gene are transferred to the laboratory strain that is more easily propagated and tested in Antiviral Therapy 6:1
Sequencing and/or point mutation assays PRO and/or RT phenotype Recombinant virus assay and/or PBMC assay
laboratory cell lines than the patient isolate. Using the recombinant virus, susceptibility towards different drugs is tested separately. The variability of these assays is typically in the range of two- to fivefold. Direct comparisons between results generated by the different variant assays based on the recombinant virus principle have not been extensively made, and their fold resistance values may differ. However, a preliminary comparison of the Antivirogram and the PhenoSense HIV seems to indicate that at the lower range of the assay, the difference in fold resistance is rather small [20]. HIV replication assays based on inhibition of the replication of patient derived isolates in cultured donor peripheral blood mononuclear cells (PBMCs), are more cumbersome, less reproducible and regularly fail to provide results due to the difficulties in propagating the patient derived strains in vitro. Thus, although this assay more closely represents the in vivo situation in terms of virus and cells used, it is of limited use in routine clinical practice. As outlined in Table 3, the recombinant virus assay is technically very demanding requiring a containment level 3 facility and can only be done in specialized laboratories; thus it is less accessible for use in routine clinical laboratories. Genotypic tests are easier to perform and commercial kits already exist, although experience in PCR and nucleotide sequencing is still 23
A-M Vandamme et al.
Table 3. Flow chart of genotypic and phenotypic resistance assays For both genotyping and phenotyping the clinical sample is generally plasma, although PBMC may be appropriate in certain situations where information about latent virus populations is needed. The lowest viral load at which resistance assays can be reliably performed is currently around 1000 copies/ml. The ability to analyse samples with lower viral loads will depend on: " Correct sample preparation / anticoagulant " Proper transport conditions and storage (limited freeze/thaw)
Clinical sample
RNA (or DNA) extraction
RT-PCR ± nested PCR
" The choice of primers and PCR conditions
Primer choice will also determine the ability to detect non-B subtypes.
Genotyping Either – Direct sequencing Or – Point mutation assays " Selective PCR " Hybridization assays " Single nucleotide sequencing
The preliminary steps of viral RNA extraction and PCR amplification are similar for all genotypic and phenotypic assays.
Amplified protease and /or RT encoding DNA
Certain regions of the genome which may play a role in drug resistance (e.g. gag, 3’ regions of RT) are not included in some assays
Phenotyping Recombinant virus produced by cloning and/or co-transfecting patient-derived HIV genes in a backbone of a laboratory HIV strain is tested for susceptibility to inhibitors during one or more cycles of replication in a continuous cell line.
Turnaround Relatively quick and simple enough to be performed ‘in house’ by experienced clinical laboratories.
Slower than genotyping and more complicated to perform, requiring a containment level 3 facility.
Detection of minor variants Point mutation assays can reveal the presence of minor resistant variants down to
