Aims/methods: Four assays for measuring HBV-DNA quantitatively have been compared with regard to sensitivity, precision and linearity. The methods were 125I-labelled solution hybridisation assay (liquid hybridisation, Abbott), an ELISA-based chemiluminescent RNA-DNA hybrid assay (RNA-DNA, Digene), a chemiluminescent branching oligonucleotide assay (bDNA, Chiron) and a membrane hybridisation assay using slot-blot equipment (slot blot).
Results: The bDNA assay was linear over three orders of magnitude and was the most sensitive assay, being approximately ten times more sensitive than the other assays, so that samples negative on RNA-DNA, liquid hybridisation and slot blot gave quantifiable results on bDNA. Furthermore, intra- and inter-assay variability showed that the bDNA and liquid hybridisation assays had the greatest precision, with coefficients of variation of 6.6% to 11.5% and 2.3% to 10.5%, respectively. However, the nominated amounts of HBV DNA in the standards (from all assays) were not reproducible in the other assays, such that amounts measured with bDNA would give values approximately twice that of RNA-DNA and 60 times that of liquid hybridisation.
Conclusions: The recently developed bDNA assay has advantages compared with the other assays in quantitating samples with low levels of virus present. In addition, since the assays vary considerably by a number of criteria, the method of measurement should always be reported.