[Note: On November 9, 2001, the U.S. Centers for Disease Control and Prevention issued two revised guidelines encouraging health care providers to routinely offer HIV testing more often. The goal is to increase the number of people who know their HIV status. These guidelines are available at: http://www.cdc.gov/hiv/pubs/guidelines.htm, or by calling 1-800-458-5231.
[We had asked previously asked Bruce Mirken to write an introduction to HIV testing, including reliability of the tests today, oral HIV tests, rapid HIV tests, anonymous testing, the home test kit (which makes anonymous testing available in all states), and viral load testing to detect HIV in the "window period" before the immune system has produced antibodies, which standard HIV tests detect. JSJ]
HIV antibody testing has been with us since 1985. Testing technology has evolved considerably over the years, with a variety of new and improved tests coming into use, both in research and daily practice. Since determining one’s HIV status is the first step in treatment decisions, it is important to understand the tests being used today, including their limitations.
"The primary purpose of the test in 1985 was to screen the blood supply," recalls Steve Morin, director of the University of California San Francisco’s AIDS Policy Research Center. At the time there were no treatments available for HIV infection, and no one knew how likely it was that an HIV-infected person would get AIDS or how quickly it might happen. The only medical interventions available dealt with the opportunistic diseases that appeared once the immune system was weakened, so there was no pressing need for people to find out their HIV status.
Still, it was clear that some who believed themselves at risk would want to know.
And that, Morin notes, led to "a fear that people who wanted to know whether they were infected or not would go to donate blood in order to find out." To head off this possible threat to the blood supply, federal and state governments set up alternative test sites, where people could be tested without giving blood. Back then, the main public health value of HIV testing was "as a prevention tool," Morin says, "to counsel people… about not transmitting the virus to anyone else."
That changed with the advent of antiretroviral therapy and prophylaxis (preventive treatment) aimed at preventing opportunistic infections. Over time, HIV testing became a gateway to treatment, as well as a prevention tool.
Although there have been technological changes, HIV testing in the U.S. still follows the same basic testing procedure as in 1985: HIV infection is only considered confirmed after two tests have been done, a screening test and a confirmatory test. In a recent article for the University of California San Francisco’s HIV InSite (http://hivinsite.ucsf.edu), University of Maryland researcher Niel Constantine explains that "screening tests possess a high degree of sensitivity, whereas confirmatory assays have a high specificity. Tests with high sensitivity produce few false-negative results, whereas tests with high specificity produce few false-positive results." Because the screening tests can produce false positives, a second screening test is typically run on the same sample – in duplicate – with the confirmatory tests only run on samples that are repeatedly positive ("reactive" in lab parlance).
The combination of the two types of tests produces results that are "highly accurate," Constantine notes, but technical errors are possible, and biological factors can occasionally produce problems.
The most common screening test is the enzyme-linked immunosorbent assay (ELISA), sometimes called enzyme immunoassay (EIA). The most often used confirmatory test is the Western blot. Identical technology is used in tests for numerous illnesses, including Lyme disease, Constantine explains. Indeed, the immunological methods underlying these tests are so fundamental that Sally Liska, head of the city of San Francisco’s Public Health Laboratory, calls it "serology 101."
The ELISA is used for initial screening because of both its high sensitivity and its practical advantages, Liska adds: "It’s a lot easier to do many specimens on an ELISA. It’s smaller volume, it’s less handling, it’s more automated."
Over 40 different ELISA HIV test kits are available from various manufacturers, though only a fraction of these are licensed by the FDA – a requirement if they are to be used in the U.S. (a few tests are approved for research only). These tests use artificial HIV proteins that are able to capture antibodies to the virus. Once those antibodies are caught, Constantine explains, they "can be detected using other reagents that are usually coupled to an indicator such as a dye or enzyme that can produce color." The change in color is read by a machine.
The Western blot is somewhat similar, but uses an electrical field that separates out the various components by their molecular weight. This allows identification of antibodies to specific viral antigens, which show up as identifiable "bands" on a strip of test paper.
The Western blot, Liska says, "is a little more complicated to do… It’s more hands-on." Because it is less sensitive, she adds, it "should never be run by itself."
Although the Western blot is the most common confirmatory test, others are sometime used, including the indirect fluorescent antibody assay (IFA) and the radioimmunoprecipitation assay (RIPA). "If performed and interpreted correctly, these extremely specific tests should not produce biologic false-positive results," Constantine writes.
One major drawback of antibody tests is the "window" period: the time it takes the body to produce antibodies after infection has begun. The standard tests for HIV do not detect the virus itself, but the antibodies that the body produces in response. During the period before the antibodies are produced, a person can be infected with HIV and can infect others, but still test negative on the HIV antibody test.
For the first tests licensed, this window period ranged from six to 12 weeks, but improved technology has allowed the detection of lower levels of antibodies, making it possible to identify them earlier. "Currently used tests can detect HIV infection between three and five weeks in most individuals," Constantine says. "This is true of just about all of the ELISAs and the rapid tests [discussed further below]. Some tests are a little more consistent in detecting at the three week period, but in general they are all equivalent." To some degree, he explains, "it also depends on the individual (who may not produce antibodies as fast as another)."
All of the early tests were done on blood. More recently developed tests look for antibodies in oral fluid or urine.
The oral test, which follows the same screening/confirmatory protocol as blood tests, has the advantage of being a noninvasive procedure that can be done in settings where blood draws would be impractical or unsafe. Presently just one brand of oral test, called OraSure, is FDA-approved. It is not a saliva test, but instead uses a small pad to draw fluid from within the gums. These fluids are in fact derived from blood, Constantine explains. "Therefore they include the same fluid (plasma) that is used for testing with serum-based tests."
"We’re not taking fluid that’s already available" in the mouth, Liska notes. "It’s not saliva or spit."
The pad used to draw the fluid is attached to a small handle resembling a toothbrush. It is placed against the gum, where it must remain for at least two full minutes to collect a proper sample.
Because the saliva present in the mouth dilutes the antibodies obtained, oral tests must be able to detect weaker concentrations of antibodies than blood tests. In general oral tests have been found to be just as accurate, but Liska believes "the oral fluid test may not be as sensitive for early seroconverters as the blood test."
Urine tests exist as well, but have not been as popular as the oral fluid test. This may be because the FDA has not yet approved a confirmatory urine test, so anyone with a positive urine test must return for a confirmatory test .
Another innovation has been the development of rapid tests. In conventional tests the sample is collected and sent to a central laboratory for processing, a procedure that usually requires the patient to return a week or more later for the results. Rapid tests (which come in both blood and oral versions) are done on-site and give a reading within half an hour. As with the ELISA, a positive reading on a rapid test requires a second, confirmatory assay such as a Western blot.
An advantage of rapid tests is that they eliminate the problem of testers who don’t return for their results. Non-return rates are fairly low in doctors’ offices and anonymous test sites, but can be quite high in other settings, including STD clinics, whereas many as one third of patients never come back to get their results.
According to Constantine, rapid tests have "proved to be as accurate as the ELISA when performed carefully by experienced personnel. Technical errors are common with these assays, however, because users become careless with these simple procedures." To deal with this problem, many rapid tests now include a built-in control that indicates whether or not the test was done properly. At present, the FDA has licensed one rapid test, made by Murex Diagnostics.
[Part II will include viral load testing for acute (primary) HIV infection, the "detuned ELISA," accuracy of HIV tests today and answers to "denialist" claims they are unreliable, anonymous testing, home testing anonymously, informed consent for testing, and counseling.]