Resource Logo

Genotypic and Phenotypic Resistance Testing


The power of highly active antiretroviral therapy (HAART) to suppress HIV has revolutionized the clinical management of HIV disease in the developed world. The capacity for HIV to develop resistance to antiretroviral drugs, however, is a significant cause of the failure of HAART. (It should be noted that "failure" in this context may refer to the breakthrough of HIV RNA to a detectable level, and not necessarily clinical disease progression.) Due to the phenomenon of cross-resistance, in which loss of sensitivity to one antiretroviral drug may confer resistance to other drugs in the same class, the development of resistance to one or more drugs in a regimen may also limit future therapeutic options.

Changes in therapy due to viral load rebound (sometimes referred to as treatment failure) therefore have been guided by a person's antiretroviral history and the cross-resistance profile of drugs he or she has taken. Switching to as many new drugs as possible is a guiding principle.

Genotypic and phenotypic resistance tests have the potential to help identify which drugs in a regimen are failing and to guide the selection of drugs for new regimens. Arguably, such testing could improve clinical care by making management of antiretrovirals more precise-and less based on educated guesswork. To date, however, the role of resistance testing in routine HIV care remains unclear. This article will review the mechanisms for the development of drug resistance, characterize genotypic and phenotypic resistance assays (tests), and outline current thinking on the best use of HIV resistance testing in clinical care.

Mechanisms of Resistance

HIV is an RNA-based, rapidly mutating virus. Unlike DNA-based viruses and other organisms, HIV lacks the ability to check for and correct genetic mutations that can occur during replication. In chronic HIV disease, ten billion new viral particles can be generated each day; given that an average of one nucleotide (an RNA molecule, or building block) is likely to mutate during each replication, it is easy to see how genetic variation in HIV is quick to occur. In just one day, all of the potential mutations coding for single-drug resistance could easily take place. This is noteworthy because the currently approved non-nucleoside reverse transcriptase inhibitors (NNRTIs) and the nucleoside analog 3TC (lamivudine, Epivir) need only one mutation to develop high-level resistance. This explains why resistance to these agents develops so rapidly when they are used as monotherapy and underscores the need to use these drugs only in the context of a maximally suppressive combination regimen.

While single mutations are common during the HIV replication process, double and triple mutations are much more rare-yet they are responsible for most cases of drug resistance. Certain drugs, such as AZT (zidovudine, Retrovir) and some of the protease inhibitors, lose their effectiveness only in the presence of multiple mutations. The specific multiple mutations are less likely to be generated by chance, but occur instead in the presence of selective drug pressure.

Selective drug pressure refers to the ability of antiretrovirals to suppress only the strains (or "quasispecies") of HIV that are sensitive to the drugs being taken. Those quasispecies that have mutated and are more resistant to the drug remain unaffected and therefore remain free to continue the viral life cycle, i.e., to replicate. The resulting viral strain is resistant to one or more of the drugs being taken. When combination HAART regimens are maximally suppressive, it is unlikely, at least in theory, that resistance would develop quickly-there is not enough viral replication occurring to generate new mutated viral particles. In reality, though, several factors can lead to inadequate suppression of HIV: heavy prior treatment with antiretrovirals and low levels of drugs in the blood due to either lack of adherence or problems with drug absorption or metabolism. In the presence of some drug exposure, these situations can lead to the development of resistance.

People who have already tried numerous HIV drugs and not achieved viral suppression, especially those who may have been on sequential monotherapy, may have developed resistance to many, if not all, of the currently available treatments. For them, constructing an effective regimen (i.e., one that will not have some degree of cross-resistance to agents they have already tried) is very difficult.

Whether or not someone with HIV has been heavily pretreated or is starting a regimen for the first time, the issue of adherence is critical. People with HIV and their providers are increasingly aware of the difficulties of adhering to complex HIV regimens. When adherence is "less than excellent," and doses are missed or are not taken in accordance with food restrictions, the blood levels of the antiretroviral drugs can drop below the level needed to suppress the virus. And when this occurs, viral replication in the presence of sub-therapeutic (low) drug concentrations can lead to the development of resistance to those drugs.

Even though lack of adherence could certainly cause someone's HIV viral load to increase while on therapy, it is important to consider other factors as well. These may include problems with drug absorption, which may be common in people with gastrointestinal complications, or drug clearance (breakdown), which can also influence blood levels of antiretroviral drugs. Detection and correction of these problems may help the drugs to maximally suppress the virus before resistance has a chance to develop.

While selective drug pressure and the real-world problems associated with maintaining maximum viral suppression account for the rise of drug resistance, the recurrence of resistance on a molecular level involves an examination of the genetic code of HIV. The nucleotide sequence of a gene is responsible for coding various amino acids that are the building blocks of the specific polypeptides that characterize the structure and function of a given organism. A codon is a specific region in RNA of three adjacent nucleotides that code for a particular amino acid. During replication, changes in nucleotides (known as mutations) may occur at specific codons so that an amino acid different from the one normally found there is produced. An example of this would be the D30N mutation associated with resistance to nelfinavir (Viracept). The number, in this case 30, designates the position of the mutation-codon 30. The first letter represents the amino acid normally found at this codon; the second letter represents the amino acid that has been produced as a result of the mutation. In this mutation, aspartic acid ("D") has been replaced with asparagine ("N") at codon 30 (refer to the chart below). When amino acid replacements like this occur, the protease gene is changed, conferring a structural change to the enzyme produced. The drugs that inhibit these enzymes therefore may no longer work as effectively. Resistance tests that quantify these genetic changes and the resulting enzyme activity may provide a clearer view into this process and help physicians develop optimal, individualized antiretroviral regimens.

Amino Acids and Their Letter Codes
Letter Code :
Amino Acid
Aspartic acid
Glutamic acid

Resistance Test Types

There are two main types of resistance tests: genotypic and phenotypic. Genotype refers to genetic constitution, reflecting the sequencing of genes of an individual organism. Phenotype is the functional manifestation of one or more genotypes. Genotypic resistance testing therefore looks for mutations in the genetic code of HIV that would be associated with resistance to specific antiretroviral drugs, while phenotypic resistance testing looks at the ability of HIV to replicate in the presence of a drug. There are advantages and disadvantages to each type of assay.

Since both genotypic and phenotypic resistance tests have relative merits and shortcomings, it is important to understand how they can help guide clinical decision-making (at least for those who can afford them). Reaching that level of understanding has been difficult so far; many clinicians are not sure how useful the tests may be in influencing treatment decisions. While current ideas regarding the use of antiretroviral resistance testing in patient care can be outlined, it should be kept in mind that there is not a clear consensus (though one is forming) at this time.

However, at the Second International Workshop on Salvage Therapies for HIV Infection, held May 19-21, 1999 in Toronto, conference co-chair Julio Montaner, MD, noted that a cost analysis of resistance testing at the BC Centre for Excellence in HIV/AIDS in Vancouver, Canada, had neutral results. That is, the analysis indicated that implementing expensive resistance tests resulted in neither added savings nor added expenses. The savings that offset the cost of the tests came from not using drugs that were determined through phenotypic testing to be of no benefit-expensive drugs that, in the absence of resistance testing, might have been used for a period of time.

Genotypic Resistance Testing

Genotypic assays look for mutations in the reverse transcriptase or protease genes that are associated with resistance to one or more drugs. Several techniques are available to do this. All of them start by taking the HIV RNA or DNA present in a sample and using polymerase chain reaction (PCR) technology to amplify the genetic material to a sufficient level to allow for genotypic testing. Some genotypic assays sequence nucleotides by using a fluorescent dye that causes each nucleotide to produce a different color. The sequences can then be read and analyzed by either a person or a computer. Other techniques, such as the Genechip assay produced by Affymetrix, use computer microchips to look for mutations and generate reports of the results. The line probe assay is a genotyping technique that compares a segment of a virus from an individual to a segment of a virus with a known sequence to determine mutations. Unlike the other assays, which can look more comprehensively at the range of codons in a virus, the line probe assay looks at only specific codons. Thus, the line probe assay is more useful when there is some idea which mutations are being sought.

Regardless of the type of genotypic assay used, the HIV viral load usually needs to be greater than 1,000 copies/mL for the test to be reliable and informative. In addition, quasispecies (strains) can be detected only if they represent 20% or more of the sample. If, for example, resistance mutations developed during past antiretroviral therapy but have receded to a low level due to a switch in therapy, they are unlikely to be detected.

Genotypic testing may be able to detect the development of resistance earlier than phenotypic testing, since it looks for genetic mutations that may occur prior to the emergence of phenotypic resistance. Genotypic testing also is more readily available than phenotypic testing, takes less time to return results (usually a few days to a few weeks), and is comparatively less expensive. However, genotypic testing is not cheap. For example, at St. Francis Memorial Hospital in San Francisco, $356 is the fee for complete results. (At other places, charges may go as high as $800.) Since genotypic testing is not approved by the Food and Drug Administration (FDA) for use in the clinical setting, it may not be covered by insurers or reimbursement programs.

Understanding how mutations that appear in a genotypic test result can influence therapeutic options is a complex matter, since the reports generated are often difficult to comprehend. Expert interpretation is therefore required to make clinical decisions using the results of genotypic testing.

Phenotypic Resistance Testing

Phenotypic testing shares some features with genotypic testing. Like genotyping, phenotyping requires an HIV viral load of at least 1,000 copies/mL. Phenotyping is also able to detect only those mutations that comprise 20% or more of the viral population. However, there are some fundamental differences between phenotypic testing and genotyping. Rather than looking for gene mutations that would indicate the likelihood of resistance, phenotyping involves culturing a person's virus in the presence of different concentrations of antiretroviral drugs. The result is compared to a strain that is known to be susceptible to the drugs. Phenotypic testing is similar to the technique used to determine antibiotic resistance to bacteria, called a "culture and sensitivity" test, which is probably more familiar to many people.

Phenotypes that are resistant to antiretrovirals can be detected by measuring the amount of drug necessary to inhibit viral growth in vitro. The standard measures are called IC50 and IC90. These designations refer to the inhibitory concentration of drug needed to inhibit the growth of HIV by 50% and 90%, respectively. Typically, a four-fold increase in IC50, compared to a laboratory sample of wild type (nonmutated) virus, would be the minimum change that could be detected.

Like genotyping, phenotypic testing is not licensed for clinical use, so reimbursement can be refused. Phenotypic testing is more expensive than genotyping (the tests typically cost $900), and results can take up to six weeks to be returned, reflecting the labor-intensive nature of the technique. Another shortcoming of phenotypic resistance testing is that HIV growth is determined by using only one drug at a time. Results are reported for each drug in monotherapy. Yet this does not occur in clinical practice, since combinations of three or more drugs are used.

On June 23, 1999, ViroLogic, Inc., a biotechnology company located in South San Francisco, announced the commercial availability of their rapid phenotypic resistance assay, called PhenoSense HIV. Their test provides results in two weeks, either in a "comprehensive panel" format (includes all FDA-approved drugs and costs $1,075) or the less expensive "select panel" format (12 drugs for $775). Health-care providers may call ViroLogic for information about ordering the assay at 800-777-0177. PhenoSense is a "home-brew" assay, which means that it is not packaged in a kit or sold through interstate commerce, and therefore does not require FDA approval at this time. As a clinical reference laboratory, ViroLogic has met and passed the requirements of the applicable federal regulation, the Clinical Laboratory Improvement Act (CLIA). The test is currently available to people in states that mandate CLIA, and in New York and California. Approval is pending in other, CLIA-exempt states including Florida and Rhode Island.

The Role of Resistance Testing in Clinical Management

The use of resistance testing in a clinical setting involves many issues. When should the tests be used? How are the results interpreted? Can the tests actually help people with HIV and clinicians make better decisions? Will utilizing the tests increase survival for persons with HIV? Will the cost of resistance testing be balanced by eliminating the cost of drugs not used because of resistance testing results? (See "Research Notes" in this issue of BETA.) While there is no universal agreement on the answers to these questions, information is available at least to address them.

When to Use Resistance Testing

Resistance testing is likely to be useful in several situations. During very early infection, resistance tests may help to assess the presence of drug-resistant viral strains. Once relatively rare, transmission of resistant virus is increasing. There are recent reports of people, including one in San Francisco, who acquired HIV that was resistant to both protease inhibitors and nucleoside analogs; such reports garnered much attention at last year's World AIDS Conference in Geneva. If used early enough, resistance testing could provide information that would help select an initial regimen for people who may have been exposed to resistant viral strains. (Once too much time has passed, the mutated strain may become a minority quasispecies and resistance testing would be less likely to detect the strain. Further, this quasispecies might not reemerge until being selected out later when antiretroviral drugs are started.)

Resistance testing is also likely to be useful prior to switching to a new regimen, when HIV viral rebound has occurred due to drug resistance. If a person's viral load is below the limit of detection and therapy is switched for other reasons, such as unmanageable side effects or dosing schedules, resistance testing would not be useful since no resistance has developed. Moreover, the laboratory would be unable to generate results with a viral load below 1,000 copies/mL. When a loss of sensitivity to drugs has occurred and viral load is increasing, resistance testing can suggest which drugs in the regimen have lost their effectiveness. In the case of genotypic testing, information about which mutations are present can give clues as to which other drugs are likely to pose resistance problems. Drugs that are not likely to be cross-resistant can then be selected. A substudy of ACTG 343, reported at the 6th Conference on Retroviruses and Opportunistic Infections (CROI) held in Chicago in February 1999, suggests that when viral rebound occurs on HAART, resistance may sometimes occur to only one of the drugs in the regimen . Current treatment guidelines support changing all of the drugs, if possible, when viral breakthrough occurs. It is possible in theory that in certain situations, resistance testing may allow for the replacement of the "failing" drug alone, while conserving other drugs and therapeutic options for the future. In practice, however, genotypic testing is still primarily used only after viral load rebound has occurred.

Interpretation of Results

Resistance tests have their limitations-genotypic mutations do not always correlate with phenotypic resistance, and phenotypic testing occurs not in the body, where multiple reactions and metabolic changes take place, but in isolated culture, or in vitro. Therefore, the results of resistance tests are probably best evaluated when antiretroviral history, the CD4 cell count, HIV viral load, and clinical status are taken into account; that is, the results must be considered in context. Evaluating CD4 cell count and viral load is the basis for determining when antiretroviral therapy should start and when it should change. Basing the decision to change therapy solely on the results of resistance testing may cause a regimen to be abandoned prematurely.

Thus understanding and interpreting the results of resistance tests is challenging, particularly for genotypic testing, since correlating different gene mutations with various antiretroviral drugs is not always a simple, linear process. Certain mutations, such as the K103N mutation that is associated with NNRTI resistance, are sufficient to cause high-level resistance by themselves. Some mutations do not trigger high-level resistance alone but can cause resistance when accumulated with others (see the three sidebar charts entitled Mutations Associated with Resistance..." ). The development of resistance can be a complex process; an interpretation of resistance test results should therefore consist of an expert evaluation that considers both the test result and an individual's medication history and clinical status.

Resistance Testing and Clinical Decision-Making: the Evidence

There is little doubt that the technology-intensive genotypic and phenotypic resistance tests provide useful information. The question is whether people who receive these tests (and their results) fare better than those who do not. Recently, researchers who attempted to answer this question have concluded that the tests may indeed play a role in HIV clinical care.

At the February CROI, the Terry Beirn Community Programs for Clinical Research on AIDS (CPCRA) reported results from the CPCRA 046 study . In this Genotypic Antiretroviral Resistance Testing (GART) study, 153 participants who experienced significant HIV viral rebound on a triple combination regimen including two nucleoside analogs and a protease inhibitor were randomized to receive either genotypic testing with or without results being interpreted by experts and communicated. That is, all participants received genotypic testing but not all had test results interpreted by experts and explained to them, for use in HIV treatment decision-making. Both groups received standard of care treatment. Physicians of the people in the group receiving GART were able to use the results in concert with their medication history and other information to make decisions about the new drug combination regimen. The group receiving no GART had to make the same decision based on medication history and lab values alone.

While both groups started with viral loads of approximately 28,000/mL, viral loads in the GART group dropped to a median of 815 copies/mL while the no-GART (no-recommendation) group only achieved a median of 7,950 copies/mL after 12 weeks. In addition, 51% of the GART group had an undetectable viral load after eight weeks, compared to 25% of the no-GART group. These results suggest that, at least with expert interpretation, the use of genotypic testing can improve the selection of appropriate drugs when switching therapy due to HIV viral load rebound. When the GART group started therapy, they received on average one more drug than the no-GART group: regimens taken by the GART group primarily consisted of four drugs while those taken by the no-GART group primarily consisted of three drugs. It should be noted that the study did not include an arm in which participants received expert opinion but no GART-i.e., there was no group that received expert advice based solely on lab values and medication history. Evaluation of this information by an expert in addition to the regular physician may have resulted in an outcome similar to those receiving expert evaluation as well as GART. Similar benefits from GART after viral rebound have also been reported from researchers in Europe .

In a study looking at phenotypic analysis, Stephen Deeks, MD, and colleagues from San Francisco General Hospital (SFGH) enrolled people whose viral loads rebounded after an indinavir (Crixivan) combination was failing; they were then switched to a "salvage protocol" using abacavir (Ziagen), nelfinavir (Viracept), and saquinavir (Fortovase), plus an NNRTI. Phenotypic testing revealed that participants who were sensitive to none or only one of the new drugs at the time of their switch in regimens had a viral load decrease of only 0.14 log copies/mL after four months, whereas people whose phenotypic results revealed sensitivity to two or three of the new drugs had a viral load decrease of 2.25 log copies/mL. These outcomes suggest that phenotypic testing prior to switching therapy can aid in selecting a drug combination that will be as potent as possible for an individual.


While research into the use of genotypic and phenotypic antiretroviral resistance testing in HIV care is still ongoing, the current evidence suggests that resistance testing has a role in clinical decision-making. Currently, resistance testing is expensive, and the results, particularly for genotyping, are challenging to interpret. If further studies show that resistance testing increases the likelihood of selecting maximally effective regimens, thereby preserving treatment options for the future, then the tests may become an important part of routine HIV care. If, as most experts expect, the incidence of drug-resistance in people taking HAART and the acquisition of drug-resistant virus in those newly infected continues to increase, the need for meaningful, affordable, and understandable assays to detect resistance will increase as well.

Greg Szekeres is a member of the research team at the Community Consortium, part of the UCSF Positive Health Program.

Selected Sources

Baxter, J. and others. A pilot study of the short-term effects of antiretroviral management based on plasma genotypic antiretroviral resistance testing (GART) in patients failing antiretroviral therapy. 6th Conference on Retroviruses and Opportunistic Infections. Chicago. January 31-February 4, 1999. Abstract LB8.

Deeks, S. and others. Correlation of baseline phenotypic drug susceptibility with 16-week virologic response in a pilot combination therapy study in HIV-infected patients who failed indinavir therapy. Antiviral Therapy 3(3; s.1): 36. 1998.

Erbelding, E. Resistance testing: a primer for clinicians, The Hopkins HIV Report 11(3). May 1999.

Gallant, J.E. Antiretroviral therapy in the experienced patient.

Havlir, D. and others. Viral rebound in the presence of indinavir without protease inhibitor resistance. 6th CROI. Abstract LB12.

Hecht, F. and others. Transmission of protease inhibitor resistant HIV-1 to a recently infected antiretroviral-naive man: the UCSF options primary HIV project. 12th World AIDS Conference. Geneva. June 28-July 3, 1998. Abstract 32288.

Hirsch, M. and others. Antiretroviral drug resistance testing in adults with HIV infection: implications for clinical management. Journal of the American Medical Association 279(24): 1984-1991. June 24, 1998.

Wegner, S. and others. High frequency of antiretroviral drug resistance in HIV-1 from recently infected therapy-naive individuals. 6th CROI. Abstract LB9


Copyright © 1999 -BETA, Publisher. All rights reserved to the San Francisco AIDS Foundation. Reproduced by permission. Reproduction of this article (other than one copy for personal reference) must be cleared through BETA: PO Box 426182, San Francisco, CA 94142-6182. Tel: 415 487 8060 Fax: 415 487 8069 San Francisco AIDS Foundation, Mail SFAF..

Information in this article was accurate in July 10, 1999. The state of the art may have changed since the publication date. This material is designed to support, not replace, the relationship that exists between you and your doctor. Always discuss treatment options with a doctor who specializes in treating HIV.