The Lumipulse G HBsAg-Quant assay for screening and quantification of the hepatitis B surface antigen

- The Lumipulse HBsAg assay showed a detection limit of 0.005 IU/mL.
- The assay identified 1% more clinical samples with trace amount of HBsAg.
- The Lumipulse assay showed high correlation (R2 ≥ 0.97, P < 0.01) with the Architect and Elecsys assays in quantifying HBsAg.
- The Lumipulse assay yielded similar patterns of HBsAg fluctuation when used for the treatment monitoring as compared to the Architect and Elecsys assays.
- HBsAg measured by the Lumipulse assay was slightly lower than that by the Architect (0.19 log10 IU/mL) and Elecsys assay (0.07 log10 IU/mL).

Qualitative HBsAg assay is used to screen HBV infection for decades. The utility of quantitative assay is also rejuvenated recently. We aimed to evaluate and compare the performance of a novel ultra-sensitive and quantitative assay, the Lumipulse assay, with the Architect and Elecsys assays.As screening methods, specificity was compared using 2043 consecutive clinical routine samples. As quantitative assays, precision and accuracy were assessed. Sera from 112 treatment-naïve chronic hepatitis B patients, four patients undergoing antiviral therapy and one patient with acute infection were tested to compare the correlations. Samples with concurrent HBsAg/anti-HBs were also quantified.The Lumipulse assay precisely quantified ultra-low level of HBsAg (0.004 IU/mL). It identified additional 0.98% (20/2043) clinical samples with trance amount of HBsAg. Three assays displayed excellent linear correlations irrespective of genotypes and S-gene mutations (R2 > 0.95, P < 0.0001), while minor quantitative biases existed. The Lumipulse assay did not yield higher HBsAg concentrations in samples with concomitant anti-HBs.Compared with other assays, the Lumipulse assay is sensitive and specific for detecting HBsAg. The interpretation of the extremely low-level results, however, is challenging. Quantitative HBsAg results by different assays are highly correlated, but they should be interpreted interchangeably only after conversion to eliminate the biases.