Antiretroviral Rx: Resistance
Case 2: Discussion
Among HIV-infected persons who are not on antiretroviral therapy, HIV replicates at an extraordinarily high rate, typically with billions of virions produced daily. A key event in HIV life cycle takes place when HIV RNA is reversed transcribed into HIV DNA. The HIV DNA formed in the reverse transcription process is comprised of nucleotides organized into triplets (codons). The HIV DNA is subsequently transcribed into messenger RNA (mRNA), with each mRNA nucleotide triplet coding for an amino acid. The linking of amino acids results in the synthesis of a polypeptide, and large polypeptides are subsequently modified to form proteins (Figure 1). A high mutation rate occurs during the reverse transcription process, predominantly because HIV reverse transcriptase fails to correct erroneously incorporated nucleotides during the reverse transcription process. The altered nucleotide sequences can result in amino acids substitutions during translation, with the potential formation of a mutated protein. In the setting of selective pressure from an antiretroviral medication, the overall effect is the generation of virus that may have diminished susceptibility to the medication. These mutated HIV strains are referred to as resistant strains of HIV. If a mutation produces an HIV strain resistant to one or more antiretroviral agents currently used by an individual, the resistant strain typically has a selective advantage over the wild type virus, and it may eventually become one of the dominant circulating strains of HIV (Figure 2).
Two types of antiretroviral resistance assays are currently available to assist the clinician in assessing HIV resistance: genotypic assays and phenotypic assays[3,4]. In addition, a genotype test can be used to generate a predicted phenotype, referred to as a virtual phenotype. With the virtual phenotype, the viral sequence (genotype) is entered into a database consisting of paired genotypes and phenotypes in order to derive an estimated phenotype. More recently, some investigators have employed newer techniques, such as allele-specific PCR, single-genome, and ultra-deep sequencing, to assess the role of minority HIV variants that harbor drug resistance, but are not detectable by current standard genotypic or phenotypic assays.
The HIV genotype test involves multiple steps: (1) polymerase chain reaction (PCR)-amplification of HIV RNA circulating in the patient's plasma; (2) direct sequencing of these amplified regions, which averages the sequence, obscuring minor genotypes; (3) comparing the patient's gene sequences to known HIV "wild-type" gene sequences; and (4) determining the corresponding amino acid alterations that would result from the specific identified DNA mutations. Commercial laboratories that perform genotype testing typically sequence the polymerase gene (encodes HIV reverse transcriptase) and the protease gene. In addition, some laboratories now have the capacity to analyze regions of the envelope and integrase HIV genes. The genotype report contains specific information regarding the amino acids that deviate from those found in a "wild-type" control strain of HIV, typically with an interpretation of what the pattern of identified mutations means in terms of predicted resistance to currently available antiretroviral medications (Figure 3). The shorthand convention used in genotype reports lists the wild type amino acid, followed by the position of the amino acid in the protein, followed by the substituted amino acid that confers resistance. Thus, the M184V notation refers to replacement of the amino acid methionine (M) by valine (V) at amino acid position 184 in the protein. In addition, most genotype reports provide an interpretation of the test results for each medication based on the overall pattern of mutations present; the interpretation typically classifies drug susceptibility in one of three categories: no evidence of resistance, low-level resistance, or high-level resistance. The interpretation of the significance of the codon mutations is based on extensive research that has identified relevant HIV mutations. The interpretation of resistance mutations, however, has not been standardized, and different laboratories may interpret the same genotype differently. Charts of known mutations have been developed to help the clinician interpret the results of a genotypic assay. The International AIDS Society-USA provides updated charts for mutations related to nucleoside reverse transcriptase inhibitors (Figure 4), non-nucleoside reverse transcriptase inhibitors (Figure 5), protease inhibitors (Figure 6), and entry inhibitors (Figure 7).
A phenotypic resistance assay attempts to assess the susceptibility of a patient's virus to various antiretroviral agents by directly measuring the viability of the predominant strain of HIV in the presence of each antiretroviral medication. The reverse transcriptase, protease, and possibly the envelope genes from a patient's dominant circulating strain of HIV are amplified with PCR, and the genes are then inserted into a laboratory HIV strain from which these genes have been deleted, generating large numbers of recombinant HIV clones. These clones are then tested for drug susceptibility to antiretroviral agents using automated assays. For each antiretroviral agent tested, the report provides an IC50 (or IC90) value; this value represents the drug concentration required to inhibit the replication of HIV by 50% (or 90%). The IC50 of the patient's sample is divided by a reference IC50 value from wild-type virus to generate a "fold change" value that represents relative resistance of the patient's HIV to the antiretroviral medication tested (Figure 8). Based on the "fold change" observed when testing the patient's HIV against an antiretroviral medication, the HIV can be considered either susceptible (Figure 9), resistant (Figure 10), or hypersusceptible (Figure 11) to the medication tested. The final report shows the fold change for each of the antiretroviral medications tested and provides an interpretation for the clinician (Figure 12).
Virtual Phenotypic Assays
The virtual phenotype is a test that attempts to combine the advantages of the genotype and the phenotype assays. This assay relies on a computerized database of more than 10,000 patient samples that include matched genotype and phenotype data. With the virtual phenotype, a conventional genotypic assay is first performed on the patient's sample of HIV. The resulting genotype is then pattern-matched to other patient samples in the database. The corresponding phenotypes for the matching samples are then extracted from the database and a predicted, or virtual phenotype result is synthesized. The final report includes both the genotype and the virtual phenotype result.
Advantages and Disadvantages of Genotypic Assays
Advantages of genotypic assays over the phenotypic assay include a faster turn-around time (roughly 2 to 3 weeks, compared with 4 to 6 weeks for a phenotypic assay), lower cost (roughly half that of a phenotypic assay), and more evidence of clinical utility from randomized controlled trials. In addition, genotypic assays are more sensitive for detecting emerging resistance, since mutations that are present at low levels may not affect overall phenotypic susceptibility, but may eventually lead to drug resistance as they are selected by ongoing drug pressure. The most significant drawback of the genotypic assay is the complexity of interpreting the results. Indeed, interpreting the genotypic resistance assay can be challenging even for expert HIV clinicians. Furthermore, the assay identifies mutations without assessing the net effect of these mutations on resistance to antiretroviral agents, and a simple summation of each mutation's expected effect may not accurately reflect the net resistance effect in the individual patient. For example, M184V, a common mutation that confers resistance to lamivudine (Epivir), is also known to partially reverse the effects of mutations that would otherwise confer significant resistance to zidovudine (Retrovir) and stavudine (Zerit). In addition, the utility of genotype tests requires a complete understanding of the genotypic correlates of resistance, which may be lacking, especially for newer antiretroviral agents. Finally, the sensitivity of either the genotype or phenotype for detecting minority HIV populations is limited.
Advantages and Disadvantages of Phenotypic Assays
The phenotype assay has several advantages when compared with the genotype assay. First and foremost, a phenotypic assay report is easier for the clinician to interpret. Second, by directly measuring the viability of a recombinant strain of HIV in the presence of various agents, it avoids the complexity of trying to gauge the net effect of multiple mutations, some of which interact with each other, which complicates the interpretation of a genotypic result. In addition, phenotype assays will accurately assess resistance even when the genotypic correlates of resistance are not well understood, a particular advantage for newer agents. The phenotypic assay, however, has its limitations as well, such as the correlation of the clinical cutoffs in relation to the drug level and how the drug is used. For example, the cut-off for atazanavir (Reyataz) would be different than if atazanavir is used in combination with ritonavir (Norvir), since the ritonavir would substantially boost the atazanavir levels. Other drawbacks of the phenotypic assay include higher cost (roughly double that of a genotypic assay), longer turn-around time (about 4 to 6 weeks, as compared with 2 to 3 weeks for a genotypic assay), lower sensitivity for emerging strains of resistant virus, and fewer clinical trials demonstrating utility compared with the genotypic assay. In addition, the virus tested in a phenotypic assay is not the patient's actual virus, but is a synthetic analog that incorporates those regions of the patient's HIV genome (reverse transcriptase, protease, and in some cases envelope), believed to be relevant for drug resistance. Whether the in vitro performance of this synthetic analog will exactly mirror the in vivo activity of the patient's strain of HIV has not been clearly established.
Advantages and Disadvantages of Virtual Phenotype
The virtual phenotype has some potential advantages compared with conventional genotypic and phenotypic assays. The turn-around time is similar to that of the conventional genotypic assay, and the cost is only slightly higher, but the resulting phenotype report provides additional and easily understandable information. Furthermore, each report has valuable genotypic information in addition to the virtual phenotype result. Nevertheless, the virtual phenotype has potential drawbacks. For HIV isolates with unusual or rare genotypes, especially those selected by recently available antiretrovirals, the amount of corresponding phenotypic data in the database would be minimal, thus potentially compromising the accuracy of the resulting virtual phenotypic report. In addition, the virtual phenotype is also a new and relatively untested technology and few clinical trials have been performed to assess its clinical benefit.
Clinical Impact of Resistance Testing
A number of randomized controlled trials have assessed the clinical utility of antiretroviral resistance assays to help guide the choice of salvage regimens in the setting of virologic failure[4,7,8,9,10,11,12,13]. The endpoints in these trials have generally consisted of the percentage of patients able to achieve an undetectable viral load on their new regimen, designed with or without the benefit of a resistance assay, and the decrease in HIV RNA levels among study patients on their new regimen. Although the various studies differ in some respects, such as the length of follow-up, extent of prior experience with antiretroviral agents, and inclusion or exclusion of expert advice, these trials generally show a moderate virologic benefit with the use of genotypic resistance assays[7,8,9,10,11]. Less data are available regarding the utility of the phenotypic assay. Results from two randomized controlled trials of the phenotypic assay have shown conflicting results, with one having found a significant benefit of phenotype testing compared with the standard of care whereas the other study did not show a benefit from phenotypic testing. Retrospective analyses of patients on failing HAART regimens have found that phenotypic testing predicted response to salvage therapy, suggesting potential clinical utility for this assay[14,15]. The relative benefit of resistance testing depends on the degree of the patient's antiretroviral experience--those with relatively less antiretroviral experience generally obtain greater benefit from resistance testing because they have more options to choose from when constructing a salvage antiretroviral regimen. Other factors, such as optimizing drug levels, can also improve virologic responses in salvage therapy.
Limitations of Resistance Assays
The evidence supporting the clinical utility of resistance assays has led the DHHS and the International AIDS Society-USA to recommend the use of resistance assays in specific scenarios, as reviewed in Case 1 of this same section (Antiretroviral Therapy Rx: Resistance). Nevertheless, resistance testing does have significant limitations. For example, both genotypic and phenotypic resistance assays may fail to detect mutant strains that constitute a minority of the patient's HIV population. The failure to detect minority populations can be problematic in several scenarios: (1) as mutant populations begin to emerge, they may exist in low numbers and not be evident on the resistant test; (2) if a patient discontinues antiretroviral therapy, wild-type virus tends to outgrow resistant virus and the resistant virus may diminish to a non-detectable level on the resistance assay; (3) if a patient switches antiretroviral therapy, the selective drug pressure changes and resistant populations may diminish to a level not detected on the resistance assay; and (4) in the setting of absence or altered drug pressure, previously generated resistance mutations may exist archived at low levels[16,17], but will re-emerge to confer resistance if the previously used antiretroviral agent is re-introduced. Thus, resistance tests provide the most accurate information when performed while the patient is on therapy and they most accurately reflect resistance to the medications currently being taken. The tests less reliably detect resistance to drugs taken in the past. Clinicians should regard the patient's antiretroviral resistance as the cumulative resistance that has developed in the patient's past, recognizing that a recent resistance assay may not detect all of these resistance mutations. Resistance results must be well documented for future consideration. If resistance tests were not performed at the time of failure of previous regimens, then it becomes especially important for the clinician to consider the patient's complete antiretroviral history, including virologic responses to past regimens, when designing a salvage regimen.
Assistance in Interpreting Resistance Tests
Data on the use and interpretation of resistance testing are constantly evolving, and neither phenotypic nor genotypic resistance testing has been standardized. Mutation resistance guides, algorithms and sensitivity cutoffs need to be updated frequently to keep pace with the latest research findings. Given the complexity and uncertainties associated with the interpretation of resistance assays, expert clinical consultation is advised for clinicians who do not have significant experience in interpreting resistance tests. The following are excellent resources for help with resistance issues.
National HIV Clinicians' Consultation Center
International AIDS Society-USA: Drug Resistance Mutation Figures
Stanford University's Resistance Database Website
Regional AIDS Education Training Centers
Copyright © 2004-2013 University of Washington