The Community Programs for Clinical Research on AIDS (CPCRA) has designed a protocol that will for the first time use viral load measurements in making treatment determinations. "Other studies have looked at the role of viral burden, but this would be the first to make treatment changes based on viral burden," said the director of the AIDS Research Consortium of Atlanta and chairperson of the CPCRA study, Melanie Thompson, of the two-year trial that is scheduled to begin this month. Viral load testing has not been approved for commercial use and its benefits have not been confirmed, but some doctors and their patients are using a diagnostic tool for "research" purposes. Ultimately, viral-load testing could increase the ability to predict the clinical outcome of drugs. For now, however, viral load's connection to clinical outcome is mostly speculative. "There is a glimmer that maybe reducing viral load improves clinical outcomes," said the National Institute of Allergies and Infectious Diseases' Lawrence Deyton, who feels that Thompson's trial may provide some of the answers needed to reach a conclusion.
Until now, the CD4 cell count--more commonly known as the T- cell count--has been the best marker for HIV disease progression. There are, however, two types of markers: immune and viral. In an HIV-infected person, the immune system (the host) and the virus (the pathogen) exist in a balance. Immune markers are host- related markers, the most important of which is the T-cell count. Others include HIV-1 antibody, cutaneous anergy or skin testing, cell count. Now, studies have shown that HIV-1 viral RNA levels are highly correlated with response to therapy and can predict progression to AIDS. Viral RNA molecules are actually tiny pieces of the virus's genetic material, and viral load tests actually count the number of these RNA pieces. This is a more accurate, direct way to measure the virus, as opposed to T-cell counts, which measure the host response to the virus. Of course, viral RNA may not replace T-cell counts overnight--or even in the future--but most clinical researchers believe that RNA will become the marker of choice.
"Viral load" and "viral burden" are essentially the same thing. A viral load test measures the amount of a virus present in the blood. RNA is the genetic material of HIV. An HIV viral load test measures the amount of HIV-specific RNA. Measuring the genetic material of the virus is the most direct way to measure the virus. Another test, p-24 antigen, measures a protein that is found on the outside of the virus or that has been shed by the virus.
In the past 15 years, scientists have learned more about HIV than about any other virus in human history. However, from the scientists point of view, the measurements of HIV are quite crude. Recently, medical science has gotten more sensitive and quantitative in determining what's going on at the site of action of the virus. Clinical research has shown that therapy-induced changes in RNA levels are significantly associated with clinical outcome. This means that if a doctor measures the RNA level of HIV and then gives an antiretroviral drug, he or she can better predict and monitor the drug's effect. Now, doctors can see the immediate effects of a drug--in contrast to the pat where antiretroviral drugs were given, essentially, in a blind manner.
In general, doctors should try to individualize antiretroviral therapies, with the goal of improving the time to AIDS or the time a patient lives. The best way to use a viral load test is to determine a baseline level of RNA, assess levels over time, and then determine response after a drug is given. If the RNA level goes down and stays down, the drug is working by inhibiting replication of the virus. Unfortunately, with viral resistance, the level goes down but may come back up. If the level comes up to baseline or higher, a doctor should switch drugs or switch to combination therapy. This is the general model of how to use viral load tests. Of course, the specifics depend on the individual drug, how the patient is doing over time, what drugs are available, and what side effects the patient may experience.
The challenge for doctors is to interpret what the results of a viral load test mean for a specific patient in making drug therapy decisions, doctors must integrate their knowledge of current clinical research with their patients' individual history and wishes.
Two main viral load tests are available commercially. It's important to keep them separate because their units of measurement are different. Unfortunately, the results of the two tests cannot be compared. One type of viral load test uses the polymerase chain reaction (PCR) to measure the quantity of viral RNA. The commercial test is called the Roche Molecular System Amplicor Monitor assay. The results of this PCR test are reported as copies of HIV RNA per milliliter of plasma (copies-per-ml). The test can detect 200 RNA copies at the lower end and at the upper end 1,000,000. This is a broad range, and the test is highly sensitive.
The other assay is the Chiron branched-chain DNA assay, also called bDNA. This test measures HIV-related RNA in a slightly different way. The reported units are viral equivalents per milliliter of plasma (eq-per-ml). The results range from a low of 10,000 eq-per-ml to 1,600,000 eq-per-ml on the high end. The bDNA test is not as sensitive as PCR, but it is more precise (repeatable) over time. Therefore, bDNA is better for following the same patient over the long term.
It is crucial for people to understand that they should not switch between the tests. At this point, we don't clearly know the relationship between the two assays. They correlate somewhat, but there is not a one-to-one linear relationship.
Another important point for doctors is not to over-interpret minor changes in the test results. There is a 10-30 percent variation in an individual test. Also, there is about a three- to four-fold variation in viral HIV levels overall. For example, a level of 100,000 copies per ml is not significantly different from a level of 130,000 copies per ml. Doctors and patients should keep this in mind--or they will be over-interpreting results of the test.
Viral load correlates strongly with stage of disease. At seroconversion, viral load is highest--between several million and 20 million copies-per-ml. Then the host makes antibodies and contains the virus within the immune system. The patient then undergoes a long clinically quite period where the viral level is low: between 1,000 and 20,000 copies-per-ml. Later in infection, the level creeps up again, to 400,000 or 500,000 copies per ml-- even to 1 or 2 million copies-per-ml. However, the correlation between viral load and stage of disease can show marked variation in any single person, as each patient is different. One particular number, like 20,000 or 70,000 copies-per-ml, is not significant. It's important to follow changes from baseline and changes over time, together with how a patient is doing clinically. More research will help up understand what specific levels mean and how to use them with different drugs and drug combinations. But for now there is no simple answer.
These tests are not widely available and are not yet FDA- approved. They are now available only at commercial reference labs. At this point, the bDNA test is only available from the Corning Nichols Institute. Results are available in three days at a cost of $198. To find out more about the HIV bDNA test, call the Corning Nichols Institute at (800) 642-4657. The PCR test is available from MetPath, Inc. at (800) 631-1390. Results are available in one week at a cost of $195. Many health insurance companies reimburse for viral load tests.
Go to the SiliconValley GeoPage