Cheap viral load test is as reliable as standard: Botswana places large order

A new viral load test designed for resource-poor settings is as reliable as a standard assay, according to a report in the Journal of Acquired Immunodeficiency Syndromes. The test, the ExaVir Load assay designed by the Swedish firm Cavidi, produces reliable results with samples as small as 0.25 millilitres, implying that it could be used for paediatric use. However it only produced consistent results for people with viral loads over 1000, the researchers note.

Cavidi has been producing low-cost viral load assays since 2002, and the researchers tested versions 2 and 3 of the ExaVir Load assay. The researchers found that the average cost of the assay including all equipment was one-fifth of the test they compared it with, the Roche COBAS Amplicor assay.

The test is cheaper and simpler than standard viral load assays because it does not detect the amount of the end product of viral replication – viral RNA – but rather the degree of viral genetic transcription, as measured by the length of strands of DNA generated by the viral reverse transcriptase enzyme. The production of these strands is measured by the number of monoclonal antibodies that form bonds with newly generated DNA.

ExaVir is a purely quantitative assay that measures the amount of reverse transcription and therefore cannot detect what kind of viral RNA is being produced. Because of this it cannot detect resistance mutations or HIV subtype – but should be able to detect novel subtypes which may be missed by RNA assays.

The researchers took 221 samples from HIV-positive adults attending the infectious diseases unit of Alfred hospital in Melbourne, Australia. They also took 27 paediatric samples from HIV-positive children under 18 months of age who were part of a mother-to-child transmission study in Nairobi, Kenya.

They found that the overall sensitivity of the ExaVir assay was lower than that of the COBAS assay. However it approached the sensitivity of the COBAS assay for viral loads over 1000 and matched it for viral loads over 10,000. It detected 94% of all samples with viral loads over 1000 copies per millilitre but only 55% of samples with viral loads between 50 and 400. This compares with a sensitivity of 95% for viral loads over 53 copies/ml for the latest automated version of the COBAS Amplicor assay (Perrin).

The ExaVir test is designed to work with a volume of one millilitre of blood. However this is not always easy to obtain in paediatric cases and in some resource-poor field settings.

The researchers found that there was little difference in the reliability of results using samples of 0.5 and 0.25 millilitres instead. The average difference in viral load reported using smaller samples was minus 0.14 logs (-30%) using the version 2 assay and a plus 0.06 logs difference (+15%) using the version 3 assay. The main difference between the version 2 and 3 assays is faster throughput – version 3 takes about 90 minutes to produce a result compared with three hours for version 2.

The researchers found that the ExaVir assay was one-fifth as expensive as the COBAS assay. A previous study (Stewart) found that a single ExaVir test cost $15 to $20. Standard viral load tests are advertised as costing $40 to $50 per test, but a study of the COBAS Amplicor test found that when all laboratory expenses were taken into account the per-test cost was $127 to $148 (Germer). Standard viral load tests also require the use of a dedicated viral load testing room to avoid contamination; this is less of a concern for the ExaVir assay.

Twenty dollars is still a considerable sum for resource-poor settings; a twice-yearly viral load test at this cost would add between 10-20% to the cost of a year’s worth of ARV drugs, depending on the regimen. Nonetheless the government of Botswana recently ordered 22,020 ExaVir tests, or one for every five people currently on antiretroviral therapy in the country.

For more reports on the ExaVir test and the use of viral loads in resource poor settings,see this report.