Background: Among patients infected with hepatitis B virus (HBV), quantification of hepatitis B e antigen (HBeAg) is accruing substantial clinical relevance as a marker for HBeAg-loss during treatment. No direct comparison has been made between assays that quantify HBeAg.
Objectives: To compare the performance of HBeAg quantification (qHBeAg) between Architect and Elecsys HBeAg assays among 183 patients with chronic HBV infection (94 treatment-naïve HBV-monoinfected and 89 antiretroviral-experienced HIV-HBV co-infected).
Study design: qHBeAg was determined in Paul Erlich Institute Units (PEIU)/mL using previously designed protocols. Values were compared with correlation and linear regression. Bland-Altman analysis was used to compare mean differences (d¯) between Elecsys and Architect assays and limits of agreement (LOA) (±2 standard deviations [SD]).
Results: Between-assay correlation was significant overall (r = 0.970), yet stronger for qHBeAg < 1000 (n = 131) versus > 1000PEIU/mL (n = 52) as determined by the Elecsys assay (r = 0.969 vs. 0.880, respectively). On average, the Elecsys assay reported qHBeAg at 13.3PEIU/mL lower than the Architect assay (LOA: -415.9, 389.3), while LOA between assays were much wider at higher levels (< 1000: -198.2, 147.9; ≥ 1000PEIU/mL: -688.4, 721.5). Further analysis indicated that d¯ did not change substantially with respect to HBV genotype, precore mutation, and CD4+ cell count, regardless of HBeAg-level. Nevertheless, seven of eight patients with highly divergent between-assay results had HBV-DNA> 2000IU/mL.
Conclusions: Elecsys and Architect assays report similar qHBeAg units with high correlation. Since qHBeAg was performed using an in-house approach, a commercially-available assay could reduce between-assay discrepancies, especially at higher HBeAg-levels.
Copyright © 2013 Elsevier B.V. All rights reserved.