icon-folder.gif   Conference Reports for NATAP  
  12th International HIV Drug Resistance Workshop
June 10-14, 2003, Los Cabos, Mexico
Back grey_arrow_rt.gif
Genotype Testing Did Not Detect Low Level NNRTI Mutations and Tenofovir Resistance, Single-Drug Treatment Interruption, 3TC Efficacy Persists Despite Resistance
Reported by Jules Levin
  The annual Resistance Workshop is one of the very few key HIV conferences each year. It is the conference where a number of basic research findings and key pilot clinical studies exploring key basic potentially important concepts are reported and very importantly discussed interactively by peer thought leaders. Often the findings and discussions at this conference results in further studies of important ideas and presentations at the Retrovirus Conference. NATAP covers and reports proceedings from this conference every year. This year reporting is being provided by several writers and key researchers. The reports from this year's conference are posted on the NATAP website at either www.natap.org or www.natap.org/2003/resistance/ndxResistance.htm New reports are produced on an ongoing basis and regularly emailed out to our listserve and posted on the website.
The subject of this immediate report is the John Mellors oral presentation at the Resistance Workshop on his findings that commercially available genotypic resistance testing did not detect NNRTI mutations in NNRTI-treatment naive and experienced patients who experienced viral failure. Although resistance was detected in these patients by novel sensitive genotypic and phenotypic resistance tests. These findings suggest the possibility of other reasons for treatment failure in some patients besides non-adherence.
The findings also suggest a further confirmation that pre-existing drug resistance due to transmission of HIV drug resistance is a concern and may impair response to therapy. HIV drug resistance may not be detectable by standard commercially available resistance testing. The study findings reported by Mellors also confirm how complicated treating HIV has become. It is increasingly difficult for doctors and care providers to remain up to date on the lasted research and treatment. It is important, however, for doctors, care providers, and patients to be well informed about the many issues affecting HIV treatment. Before launching into the details of his findings here are several important subjects of discussion at this year's Resistance Workshop. Perhaps these are the key stories out of this meeting.
There are several important messages from this Summer's Resistance Workshop which I'll highlight now. As you know, from the very beginning of the availability of protease inhibitors and before with AZT/3TC understanding the activity and how to use HIV antiretroviral drugs is very much tied to understanding the resistance associated with these anti-HIV drugs. Since protease inhibitors and NNRTIs became available much research has been conducted to understand how resistance for these drugs develops and interacts with other HIV drugs in the same class. We have also focused on NRTI resistance similarly.
Tenofovir (Viread) was made commercially available about 1.5 years ago. The K65R mutation is associated with tenofovir. At this Resistance Workshop for the first time resistance researchers started in-depth discussions about the significance of the K65R mutation. Several studies were presented regarding K65R prevalence, the potential for cross-resistance with other NRTIs, and the clinical implications of having or developing the K65R mutation. I think it was a beginning in trying to understand the clinical implications. But clearly, much more research and discussion needs to be conducted in order to have a better grasp. For example, if you use tenofovir in first line therapy in combination with AZT or 3TC what mutations will emerge upon failure of the regimen. If you combine tenofovir with abacavir or ddI what mutations will emerge. And how does this effect sequencing drugs and future treatment options. Leading resistance researchers and thought leaders started to have these discussions at the annual Resistance Workshop this Summer for the first time. For the first time there were several abstracts and studies on this subject discussed at the conference. This discussion follows the release by Gilead at the November 2002 Glasgow HIV Conference of tenofovir resistance data from the 903 study in treatment-naive patients comparing a tenofovir regimen to a d4T regimen with both regimens also including efavirenz and 3TC. Studies at the conference suggest that AZT+tenofovir may delay the development of the K65R, while other NRTI combinations may not; the prevalence of K65R appears to be increasing in the last two years from <2% to up to 4%; the presence of the K65R mutation in patients previously untreated by tenofovir may interfere with response to future tenofovir therapy; the presence of the K65R mutation may lead to cross-resistance to other NRTIs but we don't have a good understanding yet of how often this might occur. A more detailed report is forthcoming on the studies from the Resistance Workshop on tenofovir resistance. I think you will find the information interesting.
Treatment interruption with a single HIV drug was the subject of a presentation at the Workshop. This is a concept just beginning to receive attention and is an experimental idea. In actuality, this idea was first suggested in a poster at the Resistance Workshop two years ago and was the subject of a talk by Steve Deeks at the Retrovirus Conference in February 2002. At this Resistance Workshop Deeks provided follow-up resistance data from his study and a researcher from the NIH presented a study on discontinuing d4T or efavirenz in a HAART regimen which also included 3TC for 5 patients who were failing therapy with >5000 copies/ml. The studies find that discontinuing efavirenz from a regimen when viral load is detectable may lead to stable viral load for a prolonged period of time, but discontinuing d4T led to increased viral load. I will examine these study results in more detail and explore the potentially important clinical implications in a forthcoming report.
A third interesting story out of the Resistance Workshop is the benefit of developing 3TC resistance. Two studies report confirmatory evidence that there is virologic benefit despite the development of 3TC resistance reflected by the presence of the M184 3TC resistance mutation. We have thought this since the advent of 3TC years ago but these studies provide clinical data supporting this concept. After patients in the study who were taking 3TC and had the 3TC M184 mutation discontinued 3TC viral load increased, showing the efficacy of 3TC even when resistance is detected. But resistance testing reports to not consider this possibility. The information from these studies will also be the subject of a forthcoming report.
I think that the Mellor's study and the others highlighted immediately below raise concerns about the current commercially available genotypic resistance tests, and how well do they inform us in making treatment decisions. Both genotypic and phenotypic resistance tests do not clearly identify cross-resistance between NRTIs. As well, the 3TC studies will show you when I report the details that despite the presence of 3TC genotypic resistance 3TC viral efficacy persists. The genotypic tests do not dectect this efficacy.
Standard Genotype Testing Did Not Detect NNRTI Mutations But Sensitive Genotype Test Did
"Low Frequency NNRTI-Resistant Variants Contribute to failure of Efavirenz-Containing Regimens in NNRTI-Experienced Patients" (Mellors et al-abstract 134)
At the Resistance Workshop John Mellors (a noted HIV drug resistance researcher from the University of Pittsburgh and the AIDS Clinical Trials Group) reported results from a study that found standard genotypic resistance testing did not detect NNRTI mutations, which were detected at a low level by novel sensitive genotypic and phenotypic resistance tests. The presence of these mutations appear to have contributed to viral failure for the patients. There have been previous studies showing that low level resistance may not be detected by standard genotypic and phenotypic resistance tests. This study addressed the implications of this in patients. Mellors and colleagues report on patients patients who had these undetected NNRTI mutations and who experienced viral failure.
In ACTG study 398 John Mellors and the ACTG study team (including D Havir, s Palmer, L Demeter, S Hammer, J Coffin) found 48 patients enrolled in this study who were NNRTI experienced but for whom they could not find NNRTI mutations before starting the study using standard genotype testing. The study team hypothesized that that there was a low frequency (the NNRTI mutations were present at too low a level to be detected) of these mutations that developed by prior exposure to NNRTIs and they emerge with efavirenz treatment.
Interestingly, Mellors reports in this study that standard genotype testing did not detect NNRTI mutations before the study in NNRTI-treatment-experienced in a lower percentage of treatment-naive patients, but a novel sensitive genotype testing did find NNRTI mutations. Further, after viral failure standard genotype testing detected the same mutations found by the sensitive test before the study started. In this study Mellors reports using a novel phenotype test, which also had similar findings: it detected phenotypic NNRTI resistance in patients before starting the 398 study where standard genotype testing did not find resistance.
Here is a brief description of ACTG Study 398, which was a study to treat patients who had failed PI therapy and was started several years ago. Final study results were reported by Scott Hammer et al in JAMA 2002. 481 patients were enrolled in the study. HIV viral load was detectable (>1000 copies/ml) on a protease inhibitor regimen. The patients had no prior abacavir, amprenavir, and efavirenz experience. 56% of the patients were NNRTI-naive and 44% were NNRTI-experienced (>7 days experience). Standard baseline genotyping was performed on 452 patients (92%) by the ABI ViroSeq v2.0 test. All patients received efavirenz 600 mg once daily abacavir 300mg twice daily, adefovir 60mg once daily, and amprenavir. In addition patients were randomized to receive 1 of 4 PI regimens: saquinavir 1600 mg twice daily, indinavir 1200mg twice daily, nelfinavir 1250 mg twice daily, or PI placebo. The study results after 48 weeks were: 90% of NNRTI experienced patients experienced viral failure and 71% of NNRTI-naive patients experienced viral failure. At week 24, 58% on NNRTI-naive experienced viral failure and 83% of NNRTI-experienced had viral failure. Of 246 NNRTI-naive patients in the study 237 had no NNRTI mutation but interestingly 9 patients had NNRTI mutations. Of 206 NNRTI-experienced patients in the study 48 patients had no NNRTI mutations by standard genotype testing.
The study objective was to compare the frequency of minor NNRTI-resistant viruses among NNRTI-naive and experienced patients for whom standard genotype testing did not find for NNRTI mutations. The methods to do this were they randomly selected baseline patient plasma (blood) samples that were negative for NNRTI mutations using the ABIv2.0 genotype test and who experienced viral failure. There were 12 NNRTI-naive and 11 NNRTI-experienced patients selected for this study from ACTG Study 398.
TEST METHODS USED IN THIS STUDY TO IDENTIFY LOW FREQUENCY RESISTANCE (below the radar of commercially available tests).
Researchers used a sensitive genotypic test called Single Genome Sequencing (SGS) to detect mutations present at low frequency. This test was the subject of another abstract at this conference by Mary Kearney from the HIV Drug resistance program at the Natl Cancer Institute of the NIH, and the abstract report of her findings are reported at the end of this report. In addition to the SGS test method the study team used another method for detecting resistance. The Ty1-HIV RT Hybrid Retrotransposon System detects the percent of efavirenz-resistant yeast colonies. The test was the subject of an abstract at this conference by DV Nissley (NCI and SAIC-Frederick, Frederick, Md) and Mellors. They reported that this test is a sensitive phenotypic test that detects HIV drug resistant viruses present at less than 1%. In their abstract they reported that in their testing the K103N efavirenz and NNRTI mutation was accurately detected when present at 2 and 0.4% of the total virus population. They also reported that previously unidentified efavirenz resistant mutations were also detected, and the viruses containing these resistance mutations have robust RT activity. They report that this system could be used to identify novel drug resistant mutations.
NNRTI-NAiVE. In the 12 patients who were NNRTI-naive before entering the 398 study there were no NNRTI mutations found by standard genotype testing, but they found NNRTI mutations in 2 of 9 patients using the SGS method. The samples from 3 of these 12 patients were nonamplifiable. In one patient SGS testing found 103N (a key efavirenz and NNRTI mutation that confers resistance to NNRTIs). They found this mutant variant in 1 out of 41 genomes looked at. In a second patient they found two mutations: 100I and 225L. They found this mutant variant in 2 out of 33 genomes looked at. After viral failure the standard genotype test was used and then detected the 103N in the first patient, no doubt brought out by viral failure and replication. And after failure in the second patient the mutation found was a 190 mutation.
NNRTI-EXPERIENCED. By standard genotype testing 0 NNRTI mutations were found in the 11 NNRTI-experienced patients before the study. But using the SGS method mutations were found in 6 out of 10 patients before entering the study. The 103N mutation was found alone in two patients. These were found in 1 out of 33 sequences performed and in 1 out of 34 sequences performed, respectively. The 108I mutation was found alone in one patient, and was found in 3 of 19 sequences performed in this patient. A variant virus with 181C+190A was found in one patient, and found in 5 out of 15 sequences performed in this patient. And in another patient the 181C alone was found and was found in 3 of 22 sequences performed in this patient. Again after viral failure standard genotype testing was performed and was now able to detect NNRTI mutations in these patients where they could not detect them before the study. In fact, the mutations detected by standard genotype testing after viral failure were the same ones detected by the SGS method before the study. And in a few cases additional key mutations evolved, no doubt due to viral failure and increased viral replication.
Ty1-RT Hybrid in NNRTI-Naive and Experienced
This phenotypic test found similar results. Where standard genotype testing found 0 of 12 NNRT-naive patients had NNRTI resistant mutations, 2 of 10 patients had efavirenz-resistant colonies (2 patients had non-amplifiable samples). In one patient they found 0.6% (2 of 328 samples examined) efavirenz resistant colonies and the 188H/C mutation. Before starting this study, as stated above, standard genotype testing found no NNRTI mutations, but after viral failure standard testing they found NNRTI mutations (103N, 190A) in this patient. In the second patient 1 of 271 samples had efavirenz-resistant colonies (0.4%) and they found the 108I mutation. Again before the study standard genotyping found no NNRTI mutation in this patient, but after viral failure mutations were found by standard testing (100I, 103N).
In the 11 NNRTI-experienced patients, standard genotyping testing found 0 NNRTI mutations, but 7 of 11 patients had efavirenz-resistant colonies. They found that 0.6% to 7.2% of the colonies looked at in the patients had efavrienz-resistant colonies. The NNRTI mutations found in the colonies included 108I, 101E, 101E+190E, 106M+108A, 103N, 188H+190E, 181C+190A. Again after viral failure standard genotype testing detected NNRTI mutations including 103N and others while the testing did not detect them before the study.
Mellors concluded that low frequency NNRTI-resistant viruses are:
--missed by standard genotyping but can be detected by SGS and the Ty1-HIV RT assay
--are more common in NNRTI-experienced patients bur can be found in NNRTI-naive patients
--what we don't know is how prevalent this occurrence is
In order to confirm these findings they used phylogenetic analysis to look at the genetic linkage between minor NNRTI-resistant variants detected at baseline and the dominant species at viral failure. Mellors found linkage to the dominant virus population at viral failure and thus the mutations are more likely to have contributed to failure of the efavrienz-containg regimen. And often additional efavirenz mutations were found.
What are the clinical implications?
Mellors said prior drug experience matters and "what you can't see can hurt you", and he postulated "is this true for other classes of drugs?"
Comparison of single-genome sequencing with standard genotype analysis for detection of HIV-1 drug resistance mutations
abstract 86. M Kearney 1 , S Palmer 1 , F Maldarelli 1 , C Bixby 4 , H Bazmi 4 , D Rock 2 , J Falloon 3 , R Davey 3 , R Dewar 2 , J Metcalf 2 , J Mellors 4 and J Coffin 1 1 HIV Drug Resistance Program, NCI, NIH; 2 NIAID/CCMD Clinic, NIH; 3 Laboratory of Immunoregulation, NIAID, NIH; and 4 Division of Infectious Diseases, University of Pittsburgh, Pittsburgh, Pa., USA
BACKGROUND: More sensitive analyses of HIV-1 populations are needed to understand viral diversity in drug-experienced patients. To investigate the extent to which drug-resistance mutations are missed by conventional methods, we performed single-genome sequencing and standard composite genotype analysis on the same plasma samples from patients with known multidrug-resistant HIV-1.
METHODS: Plasma samples were obtained from 24 patients failing antiretroviral therapy or known to be infected with multidrug-resistant HIV-1. Composite genotypes (standard testing) were obtained by RT-PCR and sequencing of bulk PCR product. For single-genome RT-PCR sequence (SGS) analysis, cDNA derived from plasma RNA was serially diluted to a single copy, and a region encompassing p6, protease and the first 300 codons of reverse transcriptase (RT) was amplified and sequenced. Sequences of 15-46 single viral genomes were obtained from each sample. Mutations included in the analysis were those defined as conferring resistance by the Stanford database.
RESULTS: All mutations present in composite genotypes were also detected by SGS. By contrast, not all mutations in single genomes were detected in compos-ite genotypes of the same plasma sample. With one exception, all drug resistance mutations present in less than 10% of single genomes were not detected in the composite genotype, whereas mutations in 10-35% of single genomes were detected in only 25% of cases. For example, in one patient, 10 mutations conferring resistance to protease inhibitors (PIs), nucleoside RT inhibitors (NRTIs) and non-NRTIs (NNRTIs) escaped detection in the composite genotype. Each of these mutations was present in 5-20% of the 20 genomes analysed. 15% of the genomes in this sample con tained linked PI mutations (L10V, M46I, I84V, L90M, I93L), none of which was detected in the composite sequence. In a second sample example, 33% of genomes contained five linked RT mutations associated with NNRTI resistance (A98S, K101E, Y181C, G190A, T215Y), none of which were detected in the composite sequence. A third sample contained two NRTI resistance mutations: one (D67N) present in 30% of the genomes was not detected in the composite sequence; and the other (V118I) present in 21% was detected in the composite. All plasma samples tested contained at least one drug resistance mutation that was identified by SGS but not in the composite genotype.
CONCLUSIONS: These results illustrate the inadequacy of composite genotype analysis for detecting drug resistance mutations present in less than 35% of the plasma virus population. Mutations present in <10% of the virus population were almost never detected and those present in 10-35% were inconsistently detected. In addition to its greater sensitivity, SGS permits detection of linked mutations that confer high-level drug resistance. Such linkage cannot be detected in composite genotypes.