Abstract

Background:

Circulating HBsAg can be derived from both intrahepatic cccDNA and integrated HBV DNA. Functional cure may be dependent on the origins of the HBsAg. We sought to correlate cccDNA and integrated DNA from liver tissues of HBeAg(+) and (-) CHB participants with serum virological parameters and intrahepatic HBcAg and HBsAg immunostaining patterns.

Methods:

Liver tissues from untreated CHB participants of the North American Hepatitis B Research Network (HBRN) were evaluated. For cccDNA analysis, DNA prep was heat denatured and digested by plasmid-safe ATP-dependent DNase (PSAD) to remove rcDNA and integrated DNA prior to qPCR quantification. Mitochondrial DNA COX3 gene qPCR was used for cccDNA normalization. For integrated DNA detection, 100ng of total DNA was subjected to HBV-enriched next generation sequencing (NGS) assay. Enriched DNA library was sequenced by Illumina MiniSeq. The resulting HBV-host junction sequences were identified by ChimericSeq. Integrated DNA quantity was estimated based on percent of HBV-host junction reads per HBV reads and HBV sequence distribution. Serum HBV RNA was measured by Abbott research assay, HBcrAg by Fujirebio Europe, and qHBsAg by Roche Diagnostics’ Elecsys platform.

Results:

The cohort [24 HBeAg(+), 32 HBeAg(-)] was predominantly Asian (75%) with mean age of 44 years. The intrahepatic cccDNA, serum HBV DNA, RNA, HBcrAg and qHBsAg were higher among the HBeAg(+) participants (Table). HBeAg(-) vs. (+) participants had larger proportion of integrated HBV DNA in liver (Figure). There were significant correlations between cccDNA and serum RNA [r=.53, p<.001] and HBcrAg [r=.41, p=.002]. The correlations between serum DNA and RNA were strong among both HBeAg(+) and (-) CHB (Table). Correlations between serum qHBsAg levels with serum DNA or RNA were moderate for HBeAg(+) but very weak for HBeAg(-) participants. For HBeAg(-) participants, integrated DNA levels correlated significantly with qHBsAg (Table). The majority [88%] of HBeAg(+) samples had positive intrahepatic HBcAg stain compared to 13% of HBeAg(-) samples. In contrast, 96% of HBeAg(+) and 88% of HBeAg(-) livers had positive HBsAg stain. Granular cytoplasmic HBsAg staining was noted in 57% and 64% of HBeAg(+) and (-) livers respectively. Membranous HBsAg stain pattern was noted in only 14% of HBeAg(-) compared to 78% of HBeAg(+) samples.

Conclusion:

Serum RNA and HBcrAg reflect the intrahepatic cccDNA concentrations. Participants with HBeAg(-) CHB had high levels of integrated HBV DNA in the liver which correlated significantly to qHBsAg. The HBsAg derived from the integrated DNA likely contributed to the discordant serum DNA and RNA correlations to qHBsAg and the HBsAg immunostaining patterns in HBeAg (-) participants. These findings are important for the design and monitoring of therapies aiming for HBV functional cure.