Education > Clinical Updates

Clinical Update #2 - News from the 5th European HIV Drug Resistance Workshop

Cascais, Portugal
March 28-30, 2007


The 5th European HIV Drug Resistance Workshop featured 7 important scientific contributions by researchers at Virco Labs.


Influence of HIV-1 Resistance Testing on Individual Physician Treatment Decision-Making Diaz RS, Sucupira MCA, Allamasey T, Bacheler L, Staes M, Tuohy M, Winters B, Noble P, and the Brazilian Network Reference Physicians Working Group. Poster 70.
Ricardo Diaz and colleagues from Virco, on behalf of the Brazilian Network Reference Physicians Working Group, reported on how HIV resistance testing can influence physician selection of antiretroviral regimens. Ten Reference Physicians from the Brazilian Network for Antiretroviral Resistance reviewed case histories of 102 treatment-experienced patients (69% had >2 virologic failures) who had a recent failure and received a genotype. Reference Physicians made new regimen recommendations based on: 1) treatment and clinical history only, then 2) the genotype (ViroSeq 2.6) result, and finally 3) by the results from a virtual phenotype (vPT, virco®TYPE HIV-1 3.9.00). The Reference Physicians used a 4-point qualitative scale to evaluate the utility of each resistance interpretation system in helping to select a new drug regimen. The activity of the drugs selected was scored by a continuous phenotypic sensitivity score (cPSS) based on vircoTYPE analysis. Patients in this analysis had undergone a median of 3 treatment regimens over an average of 5.2 years of therapy. The average number of drugs used during this period was 5.7.


In 79% of cases, the Reference Physicians changed their empiric antiretroviral therapy choices after analyzing the genotypic report for a patient. Following results from the vPT, the majority (75%) of this group made further changes to their choice. The number of active drugs in the proposed regimen increased from 1.8 when selected based on treatment and clinical history alone to 2.2 after consulting the genotype report (P=0.0003). A further increase in the number of active drugs in the final proposed regimen from 2.2 to 2.8 (P<0.0001) was achieved through use of the vPT in addition to the genotype based prediction. Based on results from the Reference Physicians' assessment of the utility of the resistance interpretations, researchers reported that the testimony of the clinicians themselves lent credence to the value of both genotypic and particularly vPT resistance interpretation in helping select better ARV regimens. If you are interested in viewing the complete poster, please click here.


Correlation of Resistance Algorithms for Tipranavir Susceptibility With Response To Tipranavir-Containing Regimens in the RESIST Trials. Van der Borght K, Winters B, Van Craenenbroeck E, Braido V, Bacheler L, Kohlbrenner V, Hall D, Villacian J. Poster 71.
Koen Van der Borght and colleagues from Virco and Boehringer Ingelheim compared 3 genotypic algorithm-based scores to phenotypic resistance-based scores in predicting response to tipranivir (TPV)-containing regimens in 743 triple-class experienced patients who participated in the RESIST trials. The resistance call for TPV by 4 resistance interpretation systems [Stanford (version 4.2.6), TPV mutation score (developed by Boehringer Ingelheim), the US Product Label mutation list, and the virco®TYPE HIV-1 (version 4.0.01)] was correlated with clinical outcome data from the RESIST trials. Interpretation of resistance to the optimized background therapy used in combination with TPV was standardized in order to focus the comparison on the prediction of TPV activity. A responder was defined as a patient having reached >1 log10 reduction from baseline in HIV RNA level (or below the detection limit of the viral load test kit) at week 8 and/or 24.


The best accuracy and correlation with treatment response to TPV-containing regimens in triple-class experienced patients were achieved using the continuous phenotypic susceptibility score based on resistance interpretation by virco®TYPE HIV-1 (week 8 and 24). While virco®TYPE HIV-1 clinical cut-off values are based on week 8 outcome data, good correlation with virologic response was seen through week 24 (a more relevant time-point for predicting a sustained and durable treatment response). The authors surmise that the better performance of a predicted phenotype approach could be due to the ability of this method to factor complex mutational profiles in its resistance interpretation and more accurately predict expected activity of drugs against viruses with varying degrees of drug resistance. Among the genotypic algorithms evaluated, the Boehringer Ingelheim TPV mutation score provided the best performance. If you are interested in viewing the complete poster, please click here.


Accuracy and Correlation of Response Predictions Based on 4 Methodologies to Interpret Darunavir Resistance in Patients Receiving Darunavir-Containing Regimens in the POWER Studies. Winters B, Van der Borght K, Van Craenenbroeck E, Braido V, Villacian J, Bacheler L. Poster 72.
Using the same methodology outlined above in Poster 71, Bart Winters and colleagues from Virco correlated the resistance call for darunavir (DRV) by four interpretative systems (Stanford [version 4.2.6], Tibotec DRV mutation score, the US Product Label mutation list, and the virco®TYPE HIV-1 [version 4.0.01]) with clinical outcome data from the POWER trials. The best accuracy and correlation with treatment response to DRV-containing regimens in triple-class experienced patients were achieved using the continuous phenotypic susceptibility score based on resistance interpretation by virco®TYPE HIV-1 (week 8 and 24). Among the genotypic algorithms evaluated, the Tibotec DRV mutation score provided the best performance. If you are interested in viewing the complete poster, please click here.


Prevalence of Darunavir Resistance-Associated Mutations in Samples Received for Routine Resistance Testing. Rinehart A, Picchio G, de Béthune M-P, Bacheler L, Pattery T, Wasikowski B, Falcon R. Poster 44.
Alex Rinehart and colleagues from Tibotec Therapeutics, Tibotec, Virco, and xLeo reported on the prevalence of darunavir resistance-associated mutations (DRV RAM) among a large number of samples submitted to Virco for routine clinical resistance testing between July 1998 and June 2006. DRV RAM had been previously defined as V11I, V32I, L33F, I47V, I50V, I54L or M, G73S, L76V, I84V and L89V Among 207,910 isolates, 91,932 (44%) had evidence of reduced protease inhibitor (PI) susceptibility, defined by having a predicted fold change (FC) in IC50 for any PI greater than the respective vircoTYPE lower clinical cut-off (CCO1, virco®TYPE HIV-1 [version 4.0.01]).


Among the samples with reduced PI susceptibility, 62.4% had no DRV RAM while 21%, 9.9%, and 6.7% harbored 1, 2, and >3 DRV RAMs, respectively. Overall, 7 of the 11 DRV RAMs occurred in <5% of isolates with reduced PI susceptibility. The most frequent double darunavir mutations were G73S + I84V (2.6%) and L33F + I84V (1.5%). Three triple mutations appeared in 0.4% of isolates: L33F + G73S + I84V, L33F + I54 + I84V, and V11L + G73S + I84V. The most frequent major PI mutations not on the DRV list were L90M (46.2%), V82A (23.2%), and M46I (22.3%). The authors note that their findings are consistent with a recent report of the low prevalence of DRV RAMS in 532 patients who were on failing PI-based regimens. If you are interested in viewing the complete poster, please click here.


Deriving Clinically Relevant Phenotypic Cut-Off Values for the virco®TYPE HIV-1 Resistance Analysis. Bacheler L, Winters B, Van Craenenbroeck E, and the Virco Clinical Cut-Offs Collaborative Group. Poster 62.
Lee Bacheler and colleagues from Virco reported on the development of linear regression models (LM) to predict the drug susceptibility phenotype from genotype for atazanavir (ATV), darunavir (DRV), and amprenavir (APV). Starting with clinical isolates for ATV (n=16,046), DRV (n=10,770), and APV (n=43,193) where both viral genotypes and measured phenotypes (by the Antivirogram® assay) were available, LMs were developed which identify individual mutations and mutation interactions that influenced in vitro susceptibility to these 3 drugs. The LMs also provided resistance weight factors (RWFs) for these mutations which could be used to calculate the respective predicted phenotypic fold-change (FC) values from any nucleotide sequence of the protease gene. The resulting FC values predicted by virtualPhenotype™-LM provided an accurate prediction of the FC measured by the Antivirogram® phenotypic assay. Additionally, the authors demonstrated that FC values for APV predicted by virtualPhenotype™-LM were strongly correlated with FC measured in the PhenosenseTM (Monogram Biosciences) phenotypic assay.


The investigators also established clinical cut-off (CCO) values for ritonavir-boosted ATV, DRV and fAPV using separate clinical trial and cohort data from patients whose regimens included these agents. The response model which predicts HIV RNA reduction at week 8 as a function of baseline factors including viral resistance was used to establish clinical cut-off values for each drug. The lower clinical cut-off value (CCO1) is the FC value corresponding to a 20% loss of response compared to a wild type virus. The upper clinical cut-off value (CCO2) is the FC associated with 80% loss of response compared to a wild type virus. These new clinical cut-offs can be used by HIV clinicians to interpret the FC values reported on the virco®TYPE HIV-1 analysis in the context of how patients with similar FC values responded in the clinical setting. If you are interested in viewing the complete poster, please click here.


New Quantitative and Qualitative Platform To Test Co-Receptor Usage of HIV-1 Strains. Rondelez E, Van Eygen V, Van Baelen K, Vandenbroucke I, Vermeiren H, Stuyver LJ. Poster 58.
Evelien Rondelez and colleagues from Virco described development of a tropism testing platform based on a 1-step reverse transcriptase PCR amplification of the gp120 region that spans variable loop V1 through V4 (termed NH2-V4). This platform allows both qualitative, population level V3 sequencing and env phenotyping (NH2-V4), as well as quantitative clonal NH2-V4 genotyping and phenotyping. The population V3 and clonal NH2-V4 sequencing demonstrated 94 % and 84% success rate, respectively, with population sequencing results being confirmed by the clonal sequence logos. Population NH2-V4 phenotyping performed on 105 clinical samples confirmed population genotypic predictions for most R5 and X4 predicted clones. Clonal NH2-V4 sequencing and clonal NH2-V4 phenotyping could each demonstrate a broad variation in tropism in each of the samples analyzed. Based on these results, the investigators proposed a cascade for tropism testing using this platform. If you are interested in viewing the complete poster, please click here.


Sensitivity of a Clonal Genotypic and Population-Based Phenotypic HIV-1 Tropism Testing Platform Towards Minority Populations. Van Eygen V, Rondelez E, Van Baelen K, Vandenbroucke I, Stuyver LJ. Poster 78.
Using the previously described clonal genotypic and population-based phenotypic tropism assay (see Poster 58), Veerle Van Eygen and colleagues from Virco tested the sensitivity of this platform using mixtures of two HIV-1 strains at different viral loads. R5-tropic recJRCSF represented the majority strain and recNL4.3 represented the X4-tropic minority variant. The following viral mixtures were created and submitted to the clonal genotypic and population-based phenotypic tropism assay.


  • Mixture 1: 2.5% and 1% X-4 tropic with 97.5% and 99% R-5 tropic, respectively, at 5 log IU/mL.
  • Mixture 2: 20% and 8% X-4 tropic with 80% and 92% R-5 tropic, respectively, at 4 log IU/mL.
  • Mixture 3: 20% X-4 tropic with 80% R-5 tropic at 3 log IU/mL.

In the clonal genotypic assay, a 1% minority variant could be detected in the high viral load sample (5 log IU/mL, 33.3% success rate for X-4). At 4 log IU/mL, the minority X-4 variant present at either 8% or 20% of the population was detected in 100% of the tests. At 3 log IU/mL, the 20% minority variant was detected in three out of five experiments; however the proportion of X-4 tropic clones deviated substantially from the expected value. Similarly, the population-based phenotypic assay detected a 1% X4-tropic minority species at 5 log IU/mL. At 4 log IU/mL, the minority species were detected in all experiments; at 3 log IU/mL, the 20% X-4 minority variant was detected in two out of three experiments. Researchers concluded that it might be unrealistic to expect detection of minority variants below 1% at 5 log IU/ml, and that sensitivity for minority X-4 strains decreased at lower viral loads but that population-based phenotypic assay could provide a fast alternative for clonal genotypic analysis as results from both assays correlated with each other. If you are interested in viewing the complete poster, please click here.


PDF documents require Adobe® Reader® to view.
Click here to download Adobe Reader for free.