Using antibody tests to diagnose HIV

The basic requirement for access to HIV treatment is a confirmed HIV diagnosis, usually carried out using antibody tests.

Early antibody testing strategies used in US laboratories relied on a screening ELISA test followed by a Western Blot test to confirm those results. This has been replaced in many settings by strategies that use two or more ELISA-based tests, one to screen and one (or two) to confirm. WHO has taken this one step further, with strategies based on rapid test kits, of which a number have been developed and are widely available, covering HIV-1 and HIV-2.

The advantages of using rapid test kits for this purpose include the expansion of treatment to areas with minimal laboratory facilities and technicians with very basic training.

When rapid test kits are used both for initial testing and confirmation, it is also possible to provide people with their test results before they leave the clinic. Thus, fewer people need to travel long distances to have access to treatment, and those who still need to travel have one less journey to make. In practice, this can make the difference between people receiving test results and getting access to care and treatment and not doing so. This is why some clinics now run two different tests on each sample in parallel, to further speed the process, even though this is more costly than using one test as an initial screen and a second test only when the first is positive.

In Zimbabwe, for example, nurse-counsellors have been trained to conduct rapid tests at voluntary counselling and testing centres, while maintaining the highest standards of accuracy. Initial testing is with two different rapid test kits; if these give the same result, the client is told. Otherwise, the rapid tests are both repeated and if they still give different results, a sample is taken and sent to a laboratory for an ELISA test to be performed on it (Osewe, Ziyambi).

The main disadvantage of rapid test kits is that they are more expensive, per test, than using laboratory-based ELISA tests.

There is more on HIV antibody tests on aidsmap.comhere.

Diagnosing HIV in babies

Babies first antibodies are inherited from their mothers so antibodies in a babys blood do not indicate the babys HIV status, at least not for the first 18 months of the baby's life. An antibody test between 18 months and 2 years, and after any breast-feeding has ceased, is still considered appropriate to establish that a baby born to an HIV positive mother is definitely HIV negative.

There is an enormous benefit to families and to clinics providing services from knowing the baby's true HIV status much earlier than this.

Firstly, for the majority of babies who are HIV negative, many months of extra anxiety for the parents may be avoided.

Secondly, for those who are HIV positive, clinical resources can be focussed on providing co-trimoxazole and other support and monitoring their need for ARV treatment when this is available.

Diagnosis can also inform infant feeding decisions. It may sometimes be possible to advise and support a mother to breast-feed, even in settings where this is otherwise not advised, to avoid the extra risks to an HIV positive baby from formula feed or other breast milk substitutes. Alternatively, where (exclusive) breast feeding is generally encouraged, the choice might be between early weaning for babies that are still HIV negative at 6 months and extended breast feeding for those who are positive.

In the absence of definitive diagnosis, it is still possible to base treatment decisions on the mothers antibody status, combined with information on the CD4/CD8 ratio in the babys blood and, of course, clinical examination of the baby.

With babies, viral culture is not a reliable method for HIV diagnosis as it normally requires excessively large blood samples.

The standard way to identify babies that are truly HIV-infected is with tests such as PCR (Polymerase Chain Reaction), which amplify HIV viral genetic material and make it detectable. PCR testing is extremely sensitive, requiring very careful handling of specimens and high standards of laboratory practice. It requires special reagents and equipment which takes up a great deal of lab space, it needs highly trained technicians and well-managed quality assurance.

In Uganda, proviral DNA tests have been used successfully at the Joint Clinical Research Centre to diagnose babies with HIV, using small blood samples taken at birth, 4 and 6 weeks, 3 and 6 months, with consistent results. The costs - including transport and handling - are not stated, however (Enzamba).

In Cote d'Ivoire, researchers working for the French ANRS developed low-cost rapid PCR tests for HIV-1 and HIV-2 and were able to show - at least for HIV-1 - that these gave the same results as commercial branched DNA tests, but more rapidly and cheaply (Rouet). In the case of HIV-2, mother to child transmission is very rare and none of the 25 children born to HIV-2 positive mothers were infected.

As with viral load tests used to monitor treatment (discussed later), the best answer to the problem now seems to be, to carry out tests on blood spot samples, dried on filter paper (or a proprietary medium, designed for the purpose, though some laboratories are finding this to be an unnecessary expense). Once dry, these samples can be packed safely and, if necessary, sent by ordinary post to a processing lab. At the lab, economies of scale, access to highly trained technicians and laboratory scientists, technical support to maintain the equipment, and rigorous quality assurance can all be delivered.

The cost has already been driven below US $10 a test in some African services, especially where services can develop some of their own reagents. The South African National Health Laboratory Service is now providing a service not only across South Africa but also to some neighbouring countries. Similar arrangements seem likely to emerge in West Africa, South East Asia, and the Caribbean.

As an alternative to PCR tests, it is possible to use p24 antigen tests. The p24 protein is the main HIV core protein, made by infected cells in larger quantities than the other HIV proteins and therefore the easiest to detect. Current p24 tests are no more complex than ELISA antibody tests, so they can be done in a relatively small laboratory space, using equipment which is relatively easy to maintain. Qualilty assurance is still, of course, necessary.

A test for p24 is included in some screening tests for blood transfusions, because the protein can be detected before antibodies appear in the blood. However, once antibodies do appear in the blood, they mask the p24 and make it harder to detect. Similarly, p24 in a babys blood can be masked by antibodies from the babys mother. Therefore, those screening tests are of no use for diagnosing babies.

The problem of p24 masking by antibodies is overcome by heating the sample to separate the p24 from any antibodies it contains. A commercial test based on this principle is now available from PerkinElmer Life Sciences.

The results for diagnosing babies at one month old are closely comparable to those from PCR tests, with no false positives, although there is still a need for further clinical studies to establish the best time(s) after the baby is born at which to take samples for testing (Withum).

There has been some concern that the antibodies used to recognise the p24 were based on subtype B viruses, but this has been alleviated at least for subtype C viruses by work carried out in South Africa (Singh). The same study also evaluated an alternative strategy based on specific tests for IgA antibodies as opposed to IgG antibodies. The theory is that the antibodies transferred from mother to infant are of the IgG type, so IgA antibodies which target HIV are likely to mean that the baby is living with the virus. Unfortunately, IgA antibody tests were not nearly as sensitive or specific as p24 tests - there were a significant number of false positives and false negatives, compared to either p24 or PCR testing.

There is some question as to the cost of p24 tests, since economies of scale would be difficult to achieve and the technology has to be provided in a standard (i.e. commercial) version. At the time of writing, it seems most likely that nucleic acid based tests applied to dried blood spots will win this particular competition.

References

Enzamba R et al. Pro-viral HIV-1 DNA PCR test for HIV diagnosis in infants. 13th ICASA, Nairobi, abstract 105886, 2003.

Osewe et al. Accuracy of rapid HIV testing algorithm within new start VCT in Zimbabwe. 13th ICASA, Nairobi, abstract 788408, 2003.

Rouet F et al. Real-time PCR technology developed for the detection of HIV-1 RNA and HIV-2 DNA permits an inexpensive and early diagnosis of HIV infection in African neonates from Cô´¥ d'Ivoire. Antiviral Therapy 8 (Suppl. 1): abstract 277. 2003.

Singh M et al. Simple and reliable tools for the early diagnosis of perinatal HIV. Antiviral Therapy 8 (Suppl. 1): abstract 1030. 2003.

Withum DG et al. Evaluation of an ultrasensitive p24 antigen assay (UPTA) for use in the diagnosis of pediatric HIV-1 infection. XIV International AIDS Conference, Barcelona, abstract ThPeB7227, 2002.

Ziyambi Z et al. Evaluation of the performance of non-laboratory staff in the use of simple rapid HIV antibody assays at New Start voluntary counseling and testing (VCT) centers. XIV International AIDS Conference, Barcelona, abstract MoPeB3110, 2002.