Improved testing of recent HIV-1 infections with the BioRad avidity assay compared to the limiting antigen avidity assay and BED Capture enzyme immunoassay: evaluation using reference sample panels from the German Seroconverter Cohort

Abstract

Background: The variety and limitations of current laboratory methods for estimating HIV-incidence has driven attempts to improve and standardize the performance of serological 'Tests for Recent HIV-Infections' (TRI). Primary and follow-up HIV-1 positive plasma samples from individuals with well-defined dates of infection collected as part of the German Seroconverter Cohort provided specimens highly suitable for use in comparing the performance of three TRIs: the AWARE™ BED™ EIA HIV-1 Incidence test (BED-CEIA), Genetic systems HIV-1/HIV-2 Plus O EIA antibody avidity-based assay (BioRad Avidity) and Sedia™ HIV-1 LAg Avidity EIA (LAg Avidity).

Methods: The evaluation panel included 180 specimens: 44 from antiretroviral (ARV)-naïve individuals with recently acquired HIV-infection (≤ 130 days; 25 B and 19 non-B subtypes) and 136 from long-term (>12 months) infected individuals [101 ARV-naïve subtype B, 16 non-B subtypes, 14 ARV-treated individuals, 5 slow progressors (SLP)].

Results: For long-term infected, ARV-naïve individuals the false recent rates (FRR) of both the BioRad and LAg Avidity assays were 2% (2/101 for subtype B) and 6% (1/16 for subtype 'non-B'), while the FRR of the BED-CEIA was 7% (7/101 for subtype B) and 25% (4/16 for subtype 'non-B') (all p>0.05). Misclassification of ARV-treated individuals and SLP was rare by LAg (1/14, 0/5) and BioRad Avidity assays (2/14, 1/5) but more frequent by BED-CEIA (5/14, 3/5). Among recently-infected individuals (subtype B), 60% (15/25) were correctly classified by BED-CEIA, 88% (22/25) by BioRad Avidity and significantly fewer by LAg (48%, 12/25) compared to BioRad Avidity (p = 0.005) with a higher true-recency rate among non-B infections for all assays.

Conclusions: This study using well-characterized specimens demonstrated lower FRRs for both avidity methods than with the BED-CEIA. For recently infected individuals the BioRad Avidity assay was shown to give the most accurate results.

Figure 1. BED-CEIA, Bio-Rad Avidity and Lag-Avidity results according to duration of infection.

Samples from the evaluation panel were analyzed using the BED capture enzyme immunoassay (A), Bio-Rad Avidity assay (B) and LAg-Avidity enzyme immunoassay (C). Respective assay cut-offs are indicated by the black line. Data are shown for subsets of the evaluation panel.

Figure 2. Comparison of BED-CEIA, Bio-Rad and LAg-Avidity assay results according to subtype, clinical stage and disease progression.

True recent ratios within the ‘recent infections set’ subtype B and ‘non-B’ (A). False recent ratios among long-term subtype B and ‘non-B’ infected, ARV-naïve individuals (B). Misclassification in the ‘challenge specimens’ set (C). Only significant p-values (<0.05) of pairwise comparisons were indicated.

Figure 3. Number of samples in the set of recent (n = 44) and long-term infections (incl. ‘challenge specimens’; n = 136) misclassified by two incidence assays.

Misclassified samples (A) by the BED-CEIA, BioRad Avidity and by both assays (B) by the BED-CEIA, LAg Avidity and by both assays and (C) by BioRad Avidity, LAg Avidity and by both assays.