Published in the Bulletin of Experimental Treatments for AIDS 1999 Year-End issue, by the San Francisco AIDS Foundation.

Main Page

HAART and Experimental Approaches to Treatment

Anti-HIV Drug Resistance Testing

Anti-HIV Therapies Early in the Pipeline

Anti-HIV Therapy Trials: New Studies and Follow-Up of Existing Trials

New Anti-HIV Drug Interactions, Toxicities, and Dosing Options

HIV Transmission

OIs and Cancers

Year-End 1999 Table of Contents

Main Page

beta@sfaf.org

Conference Notes -- Anti-HIV Drug Resistance Testing

Harvey S. Bartnof, MD

AIDS researcher proposes when and how to use anti-HIV drug resistance tests 

Charles A.B. Boucher, MD, from the University Hospital in Utrecht, The Netherlands, has proposed an algorithm regarding usage of new, experimental drug resistance tests. In a presentation at the 39th ICAAC, Dr. Boucher indicated there were five situations in which a resistance test would be indicated, including pregnancy, primary or acute HIV infection (first weeks after infection), viral load rebound, inability to suppress viral load to undetectability, and postexposure prophylaxis (prevention). He recommends a genotypic resistance test be done first (this determines the presence of HIV gene mutations in the blood that previously have been associated with resistance to anti-HIV drugs). Genotypic tests involve a quicker turnaround time and less expense than phenotypic tests. Interpretation of genetic resistance tests requires specific expertise that not all experienced AIDS physicians have-and the field is changing rapidly. If results of the genotypic test were inconclusive or ambiguous according to HIV virologists, then a phenotypic test would be ordered. (A phenotypic resistance test measures growth of a particular HIV strain in the presence of each of the anti-HIV drugs approved by the Food and Drug Administration [FDA].) Phenotypic test results would indicate which drugs not to use and would suggest which drugs are more likely to be effective. The current turnaround time for phenotypic tests is two to three weeks, and they can cost several hundred dollars. Dr. Boucher's proposal represents the first time an HIV/AIDS researcher has proposed a mechanism as to how the tests should be used.

Many AIDS physicians are using these new tests, which are not yet FDA-approved. There are now two studies-reported at the HIV Drug Resistance Workshop last June (see later report)-indicating the benefits of genotypic resistance tests. Until these tests are approved, the issue of payment arises. Third-party payers, including insurance companies, health maintenance organizations (HMOs) and federal Medicare/Medicaid programs will be reluctant to reimburse anyone for these tests until they are sanctioned by the FDA. A proposal as to how the tests are best used and in what order represents the first step in seeking third-party reimbursement.

One of the more sobering presentations at the 3rd International Workshop on HIV Drug Resistance & Treatment Strategies was authored by Robert Schuurman, MD, from Utrecht University Hospital, The Netherlands. He presented the results of ENVA-2, a survey that evaluated the ability of 60 laboratories worldwide to identify accurately the genetic makeup of blood samples infected with HIV. Five different blood plasma samples composed of known mixtures of two HIV strains were sent to "experienced" [HIV] genotype laboratories in Europe, Canada, Australia, and the U.S. One of the strains had mutations in HIV's protease gene, while the other strain had mutations in the reverse transcriptase gene.

Among the different laboratories, there was considerable variability in gene resistance test results using the same samples. As a group, they did quite well with the samples having no mutant HIV strains. When the strains had five mutations in the protease gene, however, the labs were correct only 71% of the time. Worse yet, when the strains had five mutations in the reverse transcriptase gene, the labs correctly identified them only 66% of the time. Moreover, the percentage with accurate results using a mixture of half-mutated and half-nonmutated enzymes fell to less than 50%. The lab reports generally underestimated the presence of genotypic mutations in the HIV strains tested.

HIV viral loads decrease more when genotypic resistance tests are used 

Final reports from two prospective studies that arrived at similar conclusions when using genotypic tests to guide therapy decisions were first presented at the 3rd International Workshop on HIV Drug Resistance & Treatment Strategies. The first was a European study called Viradapt, results of which were presented by Philippe Clevenbergh, MD, from Nice University Hospital in France. A total of 108 HIV positive persons were enrolled. All experienced HIV RNA viral load rebound while taking a PI-based, triple-drug combination for more than three months. The control arm of 43 people received "standard of care" in terms of changing their antiretroviral regimen, which meant relying on the individual's anti-HIV drug history, viral load, and CD4 cell count measurements. The other arm with 65 people had genotypic antiretroviral testing (GART) while still taking the drugs that led to viral rebound. Results of GART and specific drug combination recommendations (from an AIDS drug resistance specialist) were given to the primary physician.

Interim results after six months of the Viradapt study showed that the GART group had a significantly greater viral load reduction (median decrease of 1.2 log copies/mL) than the arm without GART (median decrease of 0.7 log copies/mL). This was an on-treatment analysis, since the 17% who did not remain in the study six months were not included in the data. Also, a significantly greater percentage in the GART group had an undetectable viral load (32% had less than 200 copies/mL) when compared with the group lacking GART (14%). After the first six months, persons in the group without GART were offered open-label GART. Two-thirds of them accepted. At the end of twelve months, there was little difference between the two groups. The 12-month GART group had a mean viral load reduction of 1.2 log copies/mL, while the group that rolled into GART after six months had a mean reduction of 1.0 log copies/mL. The percentage of viral load undetectability was identical (30%) in each group.

The second study to use GART was CPCRA 046. The lead author was John Baxter, MD, from Cooper Hospital-Robert Wood Johnson Medical School in Camden, NJ. This trial was similar to the Viradapt study above; however, results after only twelve weeks were presented. A total of 153 HIV positive persons were enrolled who experienced HIV viral load rebound after 16 weeks or longer of PI-based, triple-drug therapy. GART results and AIDS resistance expert recommendations were made available to 78, while the other 75 received standard of care without GART results. The average 4- and 8-week viral load reductions for those given GART and those without GART, respectively, were a decrease of 1.2 log and 0.6 log copies/mL. The results remained significant after 12 weeks, at which point the viral load reduction difference between the two groups was 0.4 log copies/mL. Interestingly, the GART group was prescribed and taking, on average, one more drug than the other group. GART offered benefits whether the number of previous PIs was one or more than one. The study was limited in several ways, including the fact that not all primary physicians of people in the GART group followed the expert recommendations.

Even though the CPCRA study was much shorter than the European Viradapt study, similar trends occurred. Notwithstanding the results from the ENVA-2 study above, these two studies seem to show some benefits of GART with expert recommendations. However, the follow-up-particularly in the CPCRA study-was rather short. Differences may not remain after longer follow-up (e.g., up to two years).

In a late-breaker report at the 39th ICAAC, Michael Saag, MD, from the University of Alabama at Birmingham presented the results of another study indicating benefits of phenotypic resistance tests. In this trial, HIV positive people experiencing viral rebound from a current regimen underwent a PhenoSense test (PhenoSense is manufactured by Virologic); their new anti-HIV drug regimen was determined using the test results. The data presented by Dr. Saag represented a follow-up evaluation of 71 people after eight months. The only significant predictor of a sustained virologic response to the new anti-HIV regimen was phenotypic sensitivity of the new drugs in that regimen. Anti-HIV drug history was a less significant predictor of virologic success. These results were derived using a multivariate (involving multiple variables analyzed separately) analysis; they add to the weight of increasing evidence regarding benefits of anti-HIV drug resistance tests, especially in the setting of so-called salvage therapy.

Page last updated 8 January 2000