HIV+ - the big science issue

DEC. 1998/JAN. 1999

NUMBER TWO

WHAT LIES AHEAD

Charting the (Troubled) Waters
How useful are the new diagnostic tests?
By Emily Bass

Three years ago, protease inhibitors were in their infancy and viral-load tests were regarded with suspicion. Today, more people are living longer, healthier lives due to HIV combination therapy, and highly sensitive viral-load tests have become the gold standard of how well drugs can control the virus. They are used together with T-cell tests that provide a picture of how the immune system is responding to HIV infection. While early diagnostic tests looked at levels of surrogate markers-indirect measures of HIV infection-ultrasensitive viral-load tests turn the floodlights on HIV, searching through millions of healthy cells in the blood to spotlight individual virus copies inside infected cells.

These tests allow us to monitor an active infection, but what about a dormant or latent one? Right now viral-load tests can often find infectious virus particles (called provirus) within a small percentage of dormant or latent T-cells inside tissue or organs. But the tests aren't that sensitive, and researchers can only make a "best guesstimate" of the size of the latent proviral pool, also known as the third compartment. The first generation of new proviral tests are being developed now to measure this pool.

Viral-tests are also imperfect tools for accurately measuring HIV in protected sites, such as the brain and central nervous systems, that HIV drugs don't penetrate very well. How exactly do we measure any virus escaping from these protected compartments that might reignite an infection?

Finally, while viral-load tests can count the amount of HIV in cells and tissues, we can't effectively track the shape-shifting mutations of HIV that are creating drug-resistant strains of the virus. Some are multiply-drug resistant viruses, others are novel "recombinant" or hybrid viruses that carry genes from two parent viruses. For now, resistance tests are rough tools at best and can't identify multidrug-resistant strains of virus. A look at what they can and can't tell us underscores how, in many ways, HIV scientists are still working in the dark.

Positive results from several drug company-sponsored studies suggest that resistance tests can improve the likelihood of selecting an effective HIV drug regimen. Recently German researchers said that viral phenotype is a "striking predictive marker" for disease progression, suggesting a future role for such tests. Still, there are skeptics. "The relative benefit of these tests is not as much as a very good treatment history," says frontline New York AIDS doctor Gabriel Torres. "I don't think anything adds to what we get from CD4 T-cell counts and viral load," agrees Cliff Lane, clinical director of the National Institute of Allergy and Infectious Diseases. "Given that we're not eradicating the virus, the sensitivity of [viral-load tests] seems quite fine."

Of course, it pays to maintain a sense of perspective. Three years ago, viral-load tests were also regarded with healthy skepticism by many. Today, most advocates and researchers are taking a wait-and-see attitude about expensive HIV resistance tests (see Box). "Resistance tests are obviously going to be important," counters treatment activist John James. "The only question is how important are they right now?" Except in clear-cut cases of treatment failure, there's a limited understanding of how to use the information we have about HIV mutations, for example (see Box).

The same applies to the next generation of viral-load and proviral tests. We have little idea how to use them to predict or prevent disease.

Meanwhile, a number of immune-based tests are being developed to help us measure how well the immune system is recovering in people on therapy. Researchers are focusing on "activation markers"-certain proteins expressed by T-cells (like HLA-DR, CD38 and CD28). Preliminary studies suggest that immune activation increases during HIV infection and may fall as people go on therapy. A Swiss research team, for example, found that as viral-load levels fall in people on therapy, there's a decrease in CD8 T-cells expressing CD 28 and CD38 proteins. In the future, tests based on such immune markers might help measure how people control HIV over time.

So get ready.

New HIV Tests
Testing for HIV Infection

Getting tested for HIV just keeps getting easier. Up to now, standard HIV-antibody blood tests have been used to tell you whether you've been exposed to HIV; results typically take two weeks. New non-invasive HIV antibody tests are being used that reduce the risk of needle sticks. Sentinel (Seradyn) is a urine screening test; positive results are confirmed with Calypte's confirmatory urine test. Smith Kline Beecham's Orasure tests a swab of mucosa from inside your mouth. Murex offers a highly accurate FDA-approved antibody test that delivers results in a few minutes.

Several other rapid tests are being developed, including a "dot blot" or "immunoblot" test that produces a colored dot on a solid surface. There are also dried-blood tests that are useful for testing newborns or when blood samples can't be refrigerated. Two-minute, finger-prick tests that cost as little as $5 are expected in 1999.

Anonymous home-test kits are increasingly used, but only one is FDA-approved: the Home Access Health Corp.'s HIV-1 Test System. Users prick a finger and smudge the blood against a special card that's identified by numerical code and sent to a laboratory. Results are given over a hot line by a trained counselor. A word to the wise: Positive results from home kits and rapid results must be confirmed with a second laboratory antibody test. Also, while HIV advocates suggest in-person counseling, home-test kits only provide phone counseling.

Viral-Load Tests
Viral-load tests, which measure the number of copies of HIV RNA (viral genes) per milliliter in the blood, are the most reliable predictor of HIV disease progression, opportunistic infection, and viral resistance. Standard viral-load tests measure down to 400 or 500 copies-so called undetectable levels. Remember though, that doesn't mean there's no virus left-just that it falls below the limit of detection by the test. Newer ultrasensitive viral-load tests that measure down to 50 viral copies include Chiron's Quantiplex bDNA (branched DNA) 340 and Roche Molecular Systems' Amplicor HIV-1 Monitor UltraSensitive test. The NucliSense HIV 1 QT made by Organon Teknika goes down to 40 copies.

The main difference among the various tests is the technique used to count the viral copies. Once you've chosen a specific test, it's crucial to stick with it every time you measure viral load because results can vary between manufacturers. Nadir-the absolute lowest point-is the hot new watchword for test results. The lower your nadir, the longer and better your drug regimen is likely to work. Ideally the nadir will fall below the limit of viral detectability.

Resistance Tests
Genotypic and phenotypic resistance tests are newcomers to the diagnostics arena, and while they're gaining popularity among researchers, it's not clear how useful they are, especially because of their price (more than $800). These tests are designed to determine whether a given drug regimen is failing because the virus has mutated and become resistant to it. HIV mutates once every time it copies itself-up to one billion times a day. Some of these mutations help the virus survive, others cripple it, and still others have no effect at all. Scientists have identified several common "point mutations" that make HIV less responsive, or "sensitive" (read: resistant), to different drugs.

Genotypic tests search viral populations for point mutations-substitutions of amino acids in HIV's genetic code-a twisted ticker tape of proteins. These mutations are identified by a number and a pair of letters (A, C, T, or G, for example) like coordinates on a map. One type of genotypic test reads, or "sequences," entire sections of the HIV genome using an amplification technique called polymerase chain reaction (PCR). Another line-probe assay (LIPA) uses a synthetic probe that acts as a homing device to bind to specific mutations. Unfortunately, most genotypic tests only read certain sections that are hot spots for resistance and may pass over other regions where potent mutations occur.

In July, LabCorp and Virco joined forces to release HIV VircoGEN, the first commercially available genotypic test. San Francisco-based ViroLogic plans to have a contender on the market in 1999, as does Chiron Diagnostics.

Phenotypic tests demonstrate what genotypic tests try to predict: how well a virus grows in the presence of a given drug. Test results will tell you the amount of drug required to reduce the amount of virus by 50 percent or even 90 percent-values called IC50 and IC90, respectively. While phenotyping is easier to interpret than genotyping, scientists haven't established what IC50 and IC90 values really mean when it comes to switching a drug regimen.

In general, resistance tests can't predict which drugs will work-only the ones that don't-which is why they're mostly recommended to help people whose regimens are failing. As with viral-load tests, it's critical to be consistent about which tests you use. For now, experts say, the decision to change regimens should be based primarily on increases in viral load.

Currently the International AIDS Society recommends resistance testing for people considering therapy only if there is "an increased prevalence of viral resistance in a particular population." To find out what mutations are linked to HIV drugs, get your hands on the Academy of Continuing Education's pocket GeneChart (on the Web at aceponline.com); and for more on resistance, visit these Web addresses: http://www.viral-resistance.com and http://hiv-web.lanl.gov.

-EB

HIV PLUS