Background
Functional hepatitis B virus (HBV) cure is defined by clearance of HBV surface antigen (HBsAg) from blood and occurs more often in people with HIV and HBV (PWHHB) taking tenofovir-based antiretroviral therapy (ART) than people with HBV mono-infection on tenofovir. This suggests that HBsAg “clearance” is not binary but occurs across a continuum of immunological control that is reduced by HIV and restored by ART.
Methods
We tested stored baseline plasma specimens from a cohort of 2314 people who inject drugs (PWID) and men who have sex with men (MSM) living with HIV enrolled as part of a cluster-randomized trial. Specimens were tested for HBV biomarkers using the ARCHITECT platform including core antibody (anti-HBc) and surface antibody (anti-HBs) and HBV DNA (Abbott RealTime HBV viral load). HBsAg was measured using the ARCHITECT HBsAg Next Qualitative assay, which at a signal-to-cutoff (S/Co) of 1 has a sensitivity of 0.005 IU/ml (compared to ~0.05 IU/ml in most commercially available assays). We examined the association of HBsAg above (S/Co >10) and below (S/Co 1-10) the threshold with detection of anti-HBs and HBV DNA. Multivariable logistic regression identified correlates of detectable HBV DNA.
Results
Overall, 257 participants of 2198 tested had a positive anti-HBc and an HBsAg Next S/Co ≥1, of whom median age was 31 years; 28% were initiated on ART (for a median of 137 days), and 16% had undetectable HIV RNA (<150 c/mL). Median CD4+ T cell count was 379 cells/µl. 34% (n=87) had an HBsAg Next S/Co 1-10 and may have been classified as negative by most commercially-available HBsAg assays. 73/87 (84%) had HBV DNA results available; 37% (n=27) had detectable HBV DNA (Figure). Prevalence of anti-HBs in those with S/Co 1-10 was 23% vs 5% in those with S/Co of ≥10 (p<0.001). In the group with S/Co 1-10, compared to undetectable HBV DNA, those with detectable DNA had higher S/Co values for anti-HBc (aOR 1.57, 95% CI 1.26-1.95) and were less likely to be on ART at baseline (aOR 0.10, 95% CI 0.02-0.64). There was no association with anti-HBs or CD4+ T cell count.
Conclusions
These data support the paradigm that functional cure of HBV is not binary but represents a continuum of the quantity of HBsAg in plasma. This may partially explain why functional cure is gained (and lost) more often in PWHHB and that the phenomenon of occult HBV (in which HBV DNA is detected in plasma without HBsAg) may actually represent low level expression of HBsAg.
Table or Figure
