Comparison of the sensitivity of NAT using pooled donor samples for HBV and that of a serologic HBsAg assay


Address reprint requests to: Shinichiro Sato, PhD, Hokkaido Red Cross Blood Center, Yamanote 2-2, Nishi-ku, Sapporo, 063-0002, Japan; e-mail:


BACKGROUND : Studies were conducted using samples from early and late-stage HBV-infected persons to determine the pool size at which PCR had better sensitivity than a sensitive HBsAg chemoluminescence immunoassay (CLIA-HBsAg).

STUDY DESIGN AND METHODS: HBV seroconversion panels were tested for HBsAg by CLIA and for HBV DNA by nested PCR (95% hit rate: 100 copies/mL); PCR was carried out at various dilutions. HBV serologically positive samples that were detected from the simultaneous screening of 540,161 routine whole-blood donations using CLIA-HBsAg and agglutination assays were also characterized for additional markers of HBV infection.

RESULTS: In 9 of 10 HBV seroconversion panels, PCR had better sensitivity than CLIA-HBsAg at dilutions of 1-in-25 or lower. Of 65 CLIA-only confirmed-positive donor samples (agglutination assay-negative), 8 represented early infection, 2 of which were PCR positive at a 1-in-50 dilution but negative at a 1-in-100 dilution. Only 2 of 47 samples from probable late-stage HBV infection that were positive on CLIA only were PCR positive with 0.1-mL sample volume and the S-region primer; the remaining 45 samples required a 1.0-mL sample input and C-region primer for increased PCR positivity. The remaining 10 CLIA-only confirmed-positive donor samples were from HBV vaccine recipients. None of the 12 CLIA- and HBsAg-negative donor samples that were strongly anti-HBc reactive could be detected by PCR at any dilution; all 12 were PCR positive when undiluted, but 4 required a 1.0-mL input volume for PCR positivity.

CONCLUSION: For the detection of samples representing early-stage HBV infection, PCR at dilutions of 1-in-25 or lower (equivalent to a pool of ≤25 members) had greater sensitivity than CLIA-HBsAg. In contrast, samples from late-stage HBV infection were detected by PCR only with undiluted samples (0.1-mL or 1.0-mL input volumes), regardless of CLIA-HBsAg reactivity. Therefore, although NAT using minipools of 25 donations or less may be effective for the detection of early-stage HBV infection, it may not be effective for the detection of persistent HBV infection.