Prospective Laboratory Evaluation of the cobas Plasma Separation Card for HIV and Treponema pallidum Antibody Analysis

Abstract

Background The cobas Plasma Separation Card (PSC; Roche Diagnostics) was developed for HIV viral load testing. This study evaluates the performance of HIV and Treponema pallidum (Tp) antibody (Ab) detection on PSCs as an alternative to dried blood spots (DBSs). Methods EDTA whole blood samples were collected from HIV-positive (n = 100), HIV-negative (n = 50), Tp-positive (n = 100), and Tp-negative patients (n = 50) and spotted on DBS and PSC. Antibody detection performance was evaluated for HIV Ab using the Genscreen ULTRA HIV Ag-Ab test (Bio-Rad) and for Tp Ab using the Syphilis Total Ab test (Bio-Rad). Plasma was used as a reference specimen. Results Receiver operating characteristic curve analysis for DBS and PSC generated areas under the curve (AUC + 95% confidence interval) of 0.985 (0.960–1.000) and 0.987 (0.973–1.000) for HIV Ab and 1.000 (1.000–1.000) and 0.996 (0.983–1.000) for Tp Ab, respectively. Receiver operating characteristic areas under the curve were not significantly different between DBS and PSC for HIV or TP Ab. At selected cutoff values rendering at least 99% sensitivity for HIV Ab detection, the specificity was 96% on DBS and 68% on PSC. For Tp Ab detection at 90% sensitivity, 100% specificity is reached on both DBS and PSC (exceeding the required 95%). However, the median quantitative HIV and Tp Ab signal of positive samples significantly decreased in PSC compared with DBS and plasma. Conclusions Although receiver operating characteristic analysis does not seem to indicate significant differences in performance between DBS and PSC, the significant reduction in quantitative Ab detection signal dictates card composition optimization before its use for HIV and Tp Ab detection can be advised.