Analytical sensitivity of three real-time PCR assays for measuring subtype B HIV-1 RNA

Lack of HIV RNA during antiretroviral therapy (ART) is regarded as a desirable outcome. Commercial assays of HIV virus load now need to detect virus RNA concentrations below 50 c/ml and several of them have claimed a limit of detection (LOD) of 20-45 c/ml.

We have compared the performances of three commercial assays of HIV RNA, the Abbott RealTime HIV-1, the Qiagen Artus RG HIV-1 and the Roche Cobas Ampliprep Cobas TaqMan (CAPCTM) HIV-1 vs 2.0 using replicate of specimens with HIV-1 subtype B RNA concentrations of 20-200 c/ml.

Despite fair-to-moderate agreement between the three assays, probit analysis showed that their LODs differed; they were 81, 65 and 18c/ml respectively. The CAPCTM HIV-1 vs 2.0 values were higher than those of the other two; the maximum difference was 0.26 log c/ml. By testing 20 replicate of each concentration, coefficients of variation were between 0.6% and 9.2% (Abbott RealTime HIV-1), 10.3% and 38% (Qiagen Artus RG HIV-1) and 5.2% and 13.1% (Roche CAPCTM HIV-1 vs 2.0). The three assays also differed in their reproducibility and linearity for virus loads of 50-200 c/ml.

The analytical performances of commercial virus load assays differ. Direct comparisons of widely used commercial assays in clinical studies could help to identify the residual viremia that is clinically relevant for effective long term therapy.