Analytical and clinical evaluation of the cobas Epstein–Barr virus test at a tertiary care cancer hospital

Abstract

BackgroundEpstein–Barr Virus (EBV) viral loads in hematopoietic stem cell transplant (HSCT) recipients are typically monitored using quantitative molecular assays. The Cobas EBV test (Roche Molecular, Pleasanton, CA) has recently been FDA-cleared for the monitoring of EBV viral loads in plasma samples of transplant patients. In this study, we compared the viral loads obtained by a laboratory-developed test (EBV LDT) using Altona Analyte specific reagents (ASR) to those obtained on the Cobas EBV test. MethodsThe analytical performance of the assay was established using the EBV verification panel from Exact Diagnostics and the EBV ATCC strain B95-8. The clinical evaluation was performed using 343 plasma samples initially tested on the EBV LDT. ResultsThe analytical sensitivity (<18.8 IU/mL), precision (SD < 0.17 log) and linear range (35.0 IU/mL to 1E + 08 IU/mL) of the Cobas EBV assay established by the manufacturers were confirmed. The strength of the qualitative agreement was substantial between the cobas EBV and the EBV LDT (85.6 %; κ = 0.71) and almost perfect when discordant results were resolved (96.4 %; κ = 0.93). The quantitative agreement was moderate (82.9 %; κ = 0.53) with the viral load obtained on the Cobas EBV test being lower across the linear range of the two tests (mean log difference of 1.0). While the absolute values of the viral loads were markedly different, the overall trends observed in patients with multiple consecutive results were similar between the two tests. ConclusionsThe Cobas EBV test provides an accurate and valid, in vitro diagnostic (IVD) option for monitoring of EBV viral loads in transplant patients and should provide an opportunity for increased standardization and commutability of tests results across laboratories.