Comparison of two immunoassays for determining hepatitis B virus serum markers

Clin Chem Lab Med. 2011 Sep 28;50(1):153-7. doi: 10.1515/CCLM.2011.721.

Abstract

Background: To evaluate and compare the detection efficacy of serum hepatitis B virus (HBV) markers by two immunoassays: electrochemiluminescence immunoassay (ECLIA) and enzyme-linked immunosorbent assay (ELISA).

Methods: ECLIA and ELISA were used to analyze 359 serum samples, including 64 HBsAg/anti-HBs coexistence serological pattern samples (samples positive for both HBsAg and anti-HBs), 24 HBeAg/anti-HBe coexistence serological pattern samples (samples positive for both HBeAg and anti-HBe), and 271 normal serological pattern samples (negative for either of the combinations).

Results: In the normal serological pattern samples, the concordance rates of the two methods in detecting serum HBV markers were as follows: 97.05% for HBsAg, 92.62% for anti-HBs, 100% for HBeAg, 76.75% for anti-HBe, and 58.67% for anti-HBc. The differences in the qualitative criteria for anti-HBc and anti-HBe were primarily responsible for the discrepancy between the two methods (κ-values of 0.657 and 0.253, respectively). Most weak positive results, determined by ECLIA, were negative determined by ELISA, whereas the results of HBsAg, anti-HBs, and HBeAg detection were generally consistent. In the HBsAg/anti-HBs coexistence serological pattern samples, the concordance rates of HBsAg and anti-HBs detection were 98.44% and 34.38%, respectively. The positive rate of ELISA does not vary as the COI (cut-off index) varies which was determined by ECLIA; in the HBeAg/anti-HBe coexistence serological pattern samples, the concordance rates of HBeAg and anti-HBe detection were 45.83% and 79.17%, respectively. Most weak positive results, determined by ECLIA, were negative when determined by ELISA.

Conclusion: The discrepancies between the two assays in normal serological patterns samples and HBeAg/anti-HBe coexistence serological pattern samples were mostly due to the different sensitivity of the two assays, but for the HBsAg/anti-HBs coexistence serological pattern samples, the discrepancy was not caused by the different sensitivity. It is the difference in determining anti-HBs which led to the discrepancy.

Publication types

  • Comparative Study
  • Evaluation Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Biomarkers / blood*
  • Hepatitis B / blood*
  • Hepatitis B virus*
  • Humans
  • Immunoassay / methods*
  • Immunoassay / standards
  • Reproducibility of Results
  • Sensitivity and Specificity

Substances

  • Biomarkers