Ongoing viral replication and production of infectious virus in patients with chronic hepatitis B virus suppressed below the limit of quantitation on long-term nucleos(t)ide therapy

Abstract

Nucleos(t)ide analogs are standard-of-care for the treatment of chronic hepatitis B and can effectively reduce hepatitis B virus (HBV) replication but rarely leads to cure. Nucleos(t)ide analogs do not directly eliminate the viral episome, therefore treatment cessation typically leads to rapid viral rebound. While treatment is effective, HBV DNA is still detectable (although not quantifiable) in the periphery of the majority of nucleos(t)ide analog treated HBV patients, even after prolonged treatment. Addressing whether the detectable HBV DNA represents infectious virus is a key unknown and has important implications for the development of a curative treatment for HBV. The minimum HBV genome equivalents required to establish infection in human liver chimeric mice was determined by titration of HBV patient sera and the infectivity in chimeric mice of serum from patients (n = 7) suppressed to the limit of detection on nucleos(t)ide analog therapy was evaluated. A minimum of 5 HBV genome equivalents were required to establish infection in the chimeric mice, confirming this model has sufficient sensitivity to determine whether serum from virally suppressed patients contains infectious virus. Strikingly, serum from 75% (n = 3 out of 4) of nucleos(t)ide-treated HBV patients with DNA that was detectable, but below the lower limit of quantitation, also established infection in the chimeric mice. These results demonstrate that infectious virus is still present in some HBV patients on suppressive nucleos(t)ide therapy. This residual virus may support viral persistence via continuous infection and explain the ongoing risk for HBV-related complications despite long-term suppression on therapy. Thus, additional treatment intensification may facilitate HBV cure.

Fig 1. Viral kinetics of 9 patients on NA treatment and when inoculated in the uPA-SCID mouse model.

a, Viral kinetics of the patients selected for this study following administration of TDF or TDF/FTC. HBV DNA was measured by Roche TaqMan COBAS at the indicated timepoints and reported as IU/mL and converted to copies/mL (5.6 conversion factor). Each line represents an individual patient. The dashed line represents the lower LLOQ of the assay, 162 copies/mL. b, Mice (n = 2 per isolate) were inoculated with 0.1 mL of baseline serum samples from patients chronically infected with HBV genotype B (open symbols) or C (closed symbols). HBV DNA was measured by qRT-PCR and reported as copies/mL. The dashed line represents the LLOQ (4 x 10⁴ copies/mL) and the solid line represents the LLOD (8 x 10² copies/mL). c, HBsAg was measured by CLEIA (LUMIPULSE HBsAg-HQ) and reported as IU/mL. The dashed line represents the LLOQ (0.6 IU/mL). d, HBeAg was measured by CLEIA (LUMIPULSE HBeAg) and reported as IU/mL. The dashed line represents the LLOQ (12 IU/mL).

Fig 2. Establishment of the lower LLOD in human liver chimeric mice.

Mice were infected with 0.1 mL of sera from patients chronically infected with HBV diluted to contain a total of 100 copies (n = 4), 10 copies (n = 7), and 5 copies (n = 3). a, Summary of PCR results over time. Blank squares indicate HBV DNA values were LLOD but LLOQ. b-d Mean infectivity of mice (n = 2) infected with of 100, 10, and 5 copies of HBV DNA in the uPA-SCID model. HBV DNA was measured by qRT-PCR and reported as copies/mL. Mice infected with genotype B and C isolates are blue and purple, respectively. The dashed line represents the LLOQ (4 x 10⁴ copies/mL) and the solid line represents the LLOD (8 x 10² copies/mL).

Fig 3. Patient serum near the LLOQ are infectious.

Human liver chimeric mice were inoculated with 0.1 mL of sera from CHB patients from a timepoint where the HBV DNA levels were at or below the LLOQ but still TD. a, Summary of PCR results over time. Blank squares indicate HBV DNA values were LLOD but LLOQ. b-d, Infectivity of a single inoculation of patient serum into mice (n = 2). HBV DNA was measured by qRT-PCR and reported as copies/mL. The dashed line represents the LLOQ (4 x 10⁴ copies/mL) and the solid line represents the LLOD (8 x 10² copies/mL).

Fig 4. Patient serum samples below the LLOQ contain infectious virus.

Human liver chimeric mice (n = 31) were inoculated with sera from CHB patients with HBV DNA levels LLOD but LLOQ. b-d, Infectivity of sera from NA-treated patients suppressed to below the LLOQ. HBV DNA was measured by qRT-PCR and reported as copies/mL. The dashed line represents the LLOQ (4 x 10⁴ copies/mL) and the solid line represents the LLOD (8 x 10² copies/mL). Shaded gray area represents the two-week period over which mice were administered a total of 5 consecutive serum inoculations.