Comparison of the Abbott RealTime HBV assay with the Roche Cobas AmpliPrep/Cobas TaqMan HBV assay for HBV DNA detection and quantification.

Research paper by Ming-Lun ML Yeh, Chung-Feng CF Huang, Ming-Yen MY Hsieh, Jee-Fu JF Huang, Chia-Yen CY Dai, Ming-Lung ML Yu, Wan-Long WL Chuang

Indexed on: 09 May '14Published on: 09 May '14Published in: Journal of Clinical Virology


Measurement of hepatitis B virus (HBV) DNA levels is essential in the clinical management of patients with chronic HBV infection. Performance and accuracy of quantitation for HBV DNA are therefore critical in clinical practice.We aimed to compare and evaluate the performance characteristics of two HBV DNA quantitative assays: Abbott RealTime HBV (RealTime assay) and Cobas AmpliPrep/Cobas TaqMan HBV assays 2.0 (TaqMan assay).Serum samples from 220 HBV-infected patients were collected. Performance characteristics of the HBV DNA quantitative assays, including sensitivity, linearity, and reproducibility were measured. The assays were compared based on the viral status, including HBeAg, genotype, core promoter and pre-core region mutations.The RealTime assay had a sensitivity and specificity of 98.2% and 100%, respectively. The intra-assay coefficients of variation of serum samples ranged from 0.00% to 11.25% for the RealTime assay and 1.22% to 8.22% for TaqMan assay. Paired quantitative results showed excellent correlation by linear regression analysis (R(2)=0.961), good level of agreement with a mean difference of 0.31log10IU/mL, and limits of agreement of -0.62 to 1.24log10IU/mL, irrespective of HBeAg, genotype, core promoter and pre-core region mutation-specific differences. In this study, a difference of ≥1log10IU/mL between the two assays was observed in 8.6% of the samples. Genotype B and average HBV DNA levels of <3log10IU/mL were significant associated factors of this discordance.The Abbott assay delivered high performance for HBV DNA quantification and correlated extremely well with the TaqMan assay, irrespective of viral status.

More like this: