The implications of variability

Are there differences in transmission?

It has been observed that certain subtypes/CRFs are predominantly associated with specific modes of transmission. In particular, subtype B is spread mostly by homosexual contact and intravenous drug use (essentially via blood), while subtype C and CRF A/E tend to fuel heterosexual epidemics (via a mucosal route).

Whether there are biological causes for the observed differences in transmission routes remains the subject of debate. Some scientists, such as Dr Max Essex of Harvard, believe such causes do exist. Among their claims are that subtype C and CRF A/E are transmitted much more efficiently during heterosexual sex than subtype B.5,6 However, this theory has not been conclusively proven.

More recent studies have looked for variation between subtypes in rates of mother-to-child transmission. It has been claimed that such transmission is more common with subtype D than subtype A9, and that subtype C is more often transmitted than either D or A10.

Is it possible to be infected more than once?

Until about 1994, it was generally thought that individuals do not become infected with multiple distinct HIV-1 strains. Since then, many cases of people coinfected with two or more strains have been documented.

All cases of coinfection were once assumed to be the result of people being exposed to the different strains more or less simultaneously, before their immune systems had had a chance to react. However, it is now thought that "superinfection" is also occurring. In these cases, the second infection occured several months after the first. It would appear that the body's immune response to the first virus is sometimes not enough to prevent infection with a second strain, especially with a virus belonging to a different subtype. It is not yet known how commonly superinfection occurs, or whether it can take place only in special circumstances.11,12

Do HIV antibody tests detect all types, groups and subtypes?

Initial tests for HIV are usually conducted using the EIA (or ELISA) antibody test or a rapid antibody test.

EIA tests which can detect either one or both types of HIV have been available for a number of years. According to the US Centers for Disease Control and Prevention, current HIV-1 EIAs "can accurately identify infections with nearly all non-B subtypes and many infections with group O HIV subtypes."13 However, because HIV-2 and group O infections are extremely rare in most countries, routine screening programs might not be designed to test for them. Anyone who believes they may have contracted HIV-2, HIV-1 group O or one of the rarer subtypes of group M should seek expert advice.

Rapid tests - which can produce a result in less than an hour - are becoming increasingly popular. Most modern rapid HIV-1 tests are capable of detecting all the major subtypes of group M.14 Rapid tests which can detect HIV-2 are also now available.15

What are the treatment implications?

Most current HIV-1 antiretroviral drug regimens were designed for use against subtype B, and so hypothetically might not be equally effective in Africa or Asia where other strains are more common. At present, there is no compelling evidence that subtypes differ in their sensitivity to antiretroviral drugs. However, some subtypes may occasionally be more likely to develop resistance to certain drugs. In some situations, the types of mutations associated with resistance may vary. This is an important subject for future research.

The effectiveness of HIV-1 treatment is monitored using viral load tests. It has been demonstrated that some such tests are sensitive only to subtype B and can produce a significant underestimate of viral load if used to process other strains. The latest tests do claim to produce accurate results for most Group M subtypes, though not necessarily for Group O. It is important that health workers and patients are aware of the subtype/CRF they are testing for and of the limitations of the test they are applying.

Not all of the drugs used to treat HIV-1 infection are as effective against HIV-2. In particular, HIV-2 has a natural resistance to NNRTI antiretroviral drugs and they are therefore not recommended. As yet there is no FDA-licensed viral load test for HIV-2 and those designed for HIV-1 are not reliable for monitoring the other type. Instead, response to treatment may be monitored by following CD4+ T-cell counts and indicators of immune system deterioration. More research and clinical experience is needed to determine the most effective treatment for HIV-2.16

What are the implications for an AIDS vaccine?

The development of an AIDS vaccine is affected by the range of virus subtypes as well as by the wide variety of human populations who need protection and who differ, for example, in their genetic make-up and their routes of exposure to HIV. In particular, the occurance of superinfection indicates that an immune response triggered by a vaccine to prevent infection by one strain of HIV may not protect against all other strains. The effectiveness of a vaccine is likely to vary in different populations unless some innovative method is developed which guards against many virus strains.

Inevitably, different types of candidate vaccines will have to be tested against various viral strains in multiple vaccine trials, conducted in both high-income and developing countries. To learn more about efforts to develop a vaccine for AIDS, visit the website of the International AIDS Vaccine Initiative.