Résumé
•We compared four commercial viral load tests with non-B HIV-1 subtypes vs m2000sp/rt (reference), CTM, EQ and GEN.•Correlation and concordance between techniques ranged from bias=−0.00185 and r2=0.98 for m2000sp/rt vs CTM, to bias=−0.135 and r2=0.90 for EQ vs GEN.•Discrepancies were observed for 37 samples with 33 related to EQ platform that under quantified 9/20 CRF02_AG samples.•Sequence analysis did not show any remarkable differences between CRF02_AG queries and references.•There was good correlation and concordance between techniques; however, the EQ test under estimated level of CRF02_AG load.
The aim of this study was to compare four HIV-1 viral quantitation platforms, Nuclisens EasyQ v2.0® (EQ), COBAS AmpliPreP/Cobas Taqman® HIV-1 test v 2.0 (CTM), GENERIC HIV CHARGE VIRALE® (GEN), with Abbott Real Time HIV-1® (m2000sp/rt) as reference technique.
The study had first evaluated m2000sp/rt performances and then compared quantitation between techniques. Discordant samples were genotyped on gag and pol gene and sequences were analyzed using Sequence locator and SeqPublish to detect eventual mismatches.
Performance analysis of m2000sp/rt showed good results with coefficients of variation values (CV) of 1.35%, 0.65%, and 0.54% for repeatability testing of low, intermediate and high concentrations, respectively. Reproducibility tests showed low CV values with 2.36% and 1.42% for low and high concentration levels, respectively and contamination test was very low value with 0.94%.
Correlation and concordance between techniques ranged from r2=0.98 and bias=−0.00185 (for m2000sp/rt vs CTM) to r2=0.90 and bias=−0.135 (for EQ vs GEN). Discrepancies were observed on 37 samples mostly CRF02_AG but despite some mismatches, sequence analysis (26/37) did not show any remarkable differences between CRF02_AG queries and references.
This study showed good correlation and good concordance between techniques. However, EQ yielded under-quantitation of CRF02_AG.