University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Assessment of the low-cost Cavidi ExaVir Load assay for monitoring HIV viral load in pediatric and adult patients
Global Health Sciences Literature Digest
Published January 14, 2010
Journal Article

Greengrass V, Lohman B, Morris L, et al. Assessment of the low-cost Cavidi ExaVir Load assay for monitoring HIV viral load in pediatric and adult patients. Acquir Immune Defic Syndr. 2009 Nov 1;52(3):387-90.

In Context

Monitoring HIV viral load (VL) is critical for evaluating progression of disease and response to antiretroviral treatment. Unfortunately, most resource-constrained settings are unable to perform VL testing because the standard HIV RNA assays are expensive. Recently, assays based on measuring virus-associated reverse transcriptase (RT) have become available that are less costly, are easier to perform, and require less sophisticated equipment. One question is whether these assays can be used with the small sample volumes often obtained from pediatric patients.


To determine the sensitivity of recent HIV RT assays compared with the standard, ultrasensitive HIV RNA assay; to determine the ability of the RT assay to quantify VL using blood samples of less than the recommended volume of 1 mL


Clinical research laboratory, Melbourne, Australia

Study Design

Comparative laboratory assay


Plasma samples had been collected previously from 201 HIV-seropositive adults attending an infectious disease clinic in Melbourne, Australia and from 27 pediatric HIV-positive patients <18 years old from Kenyatta National Hospital, Nairobi.


Samples were tested retrospectively with versions 2 and 3 of the RT assay (Cavidi ExaVir Load Assay, Uppsala, Sweden) and with HIV RNA assay (Roche COBAS Amplicor v1.5, ultrasenstive preparation, New Jersey). To determine the validity of using smaller plasma volumes, adult samples of 0.5 mL and 0.25 mL were compared with 1 mL samples using HIV RT version 2 assay (n=35) and version 3 (n=27). Six pediatric samples were tested comparing 1-mL volumes with 0.25-mL (n=2) and 0.5-mL (n=4) samples; 21 of the 27 original pediatric samples were <1 mL, so only smaller sample volumes were compared. When serum volumes of <1 mL were used in the RT assay, the volume was adjusted to 1 mL with HIV-seronegative human plasma.


Results from adult small-volume samples tested with HIV RT assays were not significantly different from results obtained using 1-ml samples, both in version 2 (median difference 0.03 log10 copies/mL) and version 3 (median difference 0.05 log10 copies/mL) of the RT assay. Differences between the RT assay and RNA assay for matched small-volume samples were small (-0.14 log10 and 0.06 log10, version 2 and version 3, respectively). The overall correlation between the RT and the RNA assay for 116 samples in the detectable range was high (r=0.91, P<0.001). The sensitivity of the version 2 assay was 94% for samples with RNA copies >1000 copies/mL. Sensitivity was much less with low viral loads <1000 copies/mL (55%-71%).


The newer versions of this RT assay perform well using the low plasma volumes often obtained from pediatric patients. For samples with low viral loads (<800 copies/mL) and small volumes, however, the RT assay may be compromised. The newer version of the RT assay (version 3) is an improvement over earlier versions with respect to level of detection (200 copies/mL) and correlation with RNA assays.

Quality Rating

This study was of good quality. More details regarding methods and extrapolation of results would have been helpful, but their inclusion may have been constrained as the article was a brief report.

Programmatic Implications

This RT assay could be considered as a low-cost alternative to RNA tests for monitoring HIV VL, particularly in resource-constrained settings. The authors indicated that the cost, including labor and consumables, was one fifth the price of the COBAS Amplicor RNA test. Programs may want to evaluate the use of these assays in their settings, although version 3 is not yet commercially available. The RT assay detectable limit of VL still is not as low as that of RNA assays (50 copies/mL), but nevertheless could be useful in monitoring response to antiretroviral therapy.