Background: HIV-1 is characterized by increased genetic heterogeneity which tends to hinder the reliability of detection and accuracy of HIV-1 RNA quantitation assays.
Methods: In this study, the Abbott RealTime HIV-1 (Abbott RealTime) assay was compared to the Roche Cobas TaqMan HIV-1 (Cobas TaqMan) and the Siemens Versant HIV-1 RNA 3.0 (bDNA 3.0) assays, using clinical samples of various viral load levels and subtypes from Greece, where the recent epidemiology of HIV-1 infection has been characterized by increasing genetic diversity and a marked increase in subtype A genetic strains among newly diagnosed infections.
Results: A high correlation was observed between the quantitative results obtained by the Abbott RealTime and the Cobas TaqMan assays. Viral load values quantified by the Abbott RealTime were on average lower than those obtained by the Cobas TaqMan, with a mean (SD) difference of -0.206 (0.298) log(10) copies/ml. The mean differences according to HIV-1 subtypes between the two techniques for samples of subtype A, B, and non-A/non-B were 0.089, -0.262, and -0.298 log(10) copies/ml, respectively. Overall, differences were less than 0.5 log(10) for 85% of the samples, and >1 log(10) in only one subtype B sample. Similarly, Abbott RealTime and bDNA 3.0 assays yielded a very good correlation of quantitative results, whereas viral load values assessed by the Abbott RealTime were on average higher (mean (SD) difference: 0.160 (0.287) log(10) copies/ml). The mean differences according to HIV-1 subtypes between the two techniques for subtype A, B and non-A/non-B samples were 0.438, 0.105 and 0.191 log(10) copies/ml, respectively. Overall, the majority of samples (86%) differed by less than 0.5 log(10), while none of the samples showed a deviation of more than 1.0 log(10).
Conclusions: In an area of changing HIV-1 subtype pattern, the Abbott RealTime assay showed a high correlation and good agreement of results when compared both to the Cobas TaqMan and bDNA 3.0 assays, for all HIV-1 subtypes tested. All three assays could determine viral load from samples of different HIV-1 subtypes adequately. However, assay variation should be taken into account when viral load monitoring of the same individual is assessed by different systems.