A major barrier to developing HIV cure interventions is the lack of validated assays that reliably quantify HIV in plasma below the limit of detection of current clinical assays. We sought to objectively assess the performance characteristics of newer, more sensitive assays to quantify plasma viremia.
The HIV Reservoir Assay Validation and Evaluation Network (RAVEN) project was created to assess the sensitivity, specificity, limits of detection and quantitative ranges of blood-based assays of HIV persistence. To accomplish this for HIV detection in plasma, blinded 50-sample panels containing subtype B and C HIV including duplicate, virus-spiked analytic standards (2 subtype-specific 5-step, 3-fold dilution series with highest concentrations ranging from 45 to 6 copies/mL to lowest concentrations ranging from 0.33 to 0.07 copies/mL), clinical samples with expected low-level viremia and negative controls were distributed to 8 laboratories for HIV quantification using 9 assays. Four assays included centrifugation to concentrate virus prior to HIV RNA nucleic acid extraction and PCR amplification; two included replicate testing using diagnostic HIV RNA assay platforms; while two utilized ultrasensitive detection of p24 Ag without virus enrichment. Results were analyzed for sensitivity, specificity, reproducibility, and ability to accurately quantify HIV in standards.
Data from five laboratories using 6 assays were included in this analysis. Four of five RNA-amplification based assays detected virus in the standards down to ~1 copy/mL in at least 1 of the 2 replicates; negative controls were all negative. All RNA-amplification-based assays had strong correlations between replicates across the standards (p<0.05, rho>0.8). Four assays quantified standards with little bias (mean recovery 69-218% of nominal HIV [RNA]) whereas one assay overestimated copies/mL by >300%. Ultrasensitive p24 Ag assays were not able to quantitatively measure HIV in the diluted standards. One p24 assay detected HIV protein in virus-spiked samples, with observed specificity of 90%.
p24 assays can detect virus in seronegative plasma without virus enrichment, but dynamic sensitivity was lacking at <45 copies/mL. Ultrasensitive RNA-amplification assays following virus enrichment or with replicate testing can quantitatively measure HIV RNA down to ~1 copy/mL, which is necessary to assess the impact of experimental curative interventions on residual viremia.