Determination of the mean duration of recent infection and false recency rate for the HIV triplex multiplex bead assay

Abstract

Background: We developed the HIV Triplex multiplex bead assay to identify and serotype HIV infection with high sensitivity and specificity; and distinguish recent from long-term HIV-1 infections. It can facilitate accurate incidence estimation, while reducing the number of tests and blood collected, which is highly desirable for use in future studies and surveys. Using previously collected, treatment-naive longitudinal seroconversion HIV-1 positive panels and specimens from individuals infected for >12 months, we determined the assay's mean duration of recent infection (MDRI) and false-recency rate (FRR) respectively, at various mean fluorescent intensity (MFI) cutoffs.

Methods: We tested seroconversion specimens (N = 814) from 142 individuals infected with HIV-1 subtypes B, C, or AE, and 1341 cross-sectional specimens from individuals infected >12 months. The MFI cutoffs of 1000 to 2000 were evaluated for recency classification, including an MFI of 1250 corresponding to the limiting antigen avidity enzyme immunoassay (LAg-EIA) cutoff of 1.5 normalized optical density for MDRI and FRR. We used four statistical methods: Methods 1 and 2 used the empirically balanced observation time approach. Method 2 MFI values were raised to power = 1.33, based on a repeated measures model to linearize the relationship between MFI and time points, whereas Method 1 was not linearized. Methods 3 and 4 employed quadratic and linear interpolations for each seroconversion panel. FRR was calculated by dividing the number of specimens misclassified as recent by the total number of specimens tested.

Results: MDRI values ranged from 135-146 days at MFI = 1000 to 229-279 days at MFI = 2000 by the 4 methods. FRR varied from 0.15%-1.27% with increasing MFI cutoff. At MFI = 1250, the average MDRI of 4 methods was 169 days and ranged from 159-183 with overlapping 95% CIs and FRR = 0.52%.

Conclusion: The HIV Triplex assay demonstrates a longer dynamic range compared to current HIV recency assays with a low FRR for cutoffs examined. With a longer dynamic range and low FRR, the MDRI for recent infection can be extended as appropriate to detect more recent infections, increasing the value of incidence assays benefiting public health surveillance and future surveys.

Fig 1. Antibody avidity measurements over time.

Development of antibody avidity after seroconversion, as measured by the limiting concentration of rIDR-M (bead region 13) on the HIV Triplex assay for all 142 subjects; MFI is plotted versus days post seroconversion. Each line represents an individual subject.

Fig 2. MDRI for three different HIV-1 subtypes at MFI cutoff of 1250 intensity units.

The MDRI of HIV-1 subtypes B, C, and AE are shown for the samples tested. There were 218 subtype B samples, 128 subtype C samples and 68 subtype AE samples tested.