Residual risk of transfusion-transmitted hepatitis B virus (TT-HBV) infection by NAT-screened blood components: A review of observed versus modeled infectivity from donors with window period and occult HBV infections

Background The residual transfusion-transmitted hepatitis B virus (TT-HBV) risk with different testing strategies depends on the sensitivity of screening assays, the prevalence of hepatitis B surface antigen (HBsAg) compared to HBV-DNA in window period (WP) and occult HBV infections (OBIs), and infectivity of blood in these infection stages. We compared modeled WP and OBI transmission risk in a multiregional individual donation nucleic acid amplification technology (ID-NAT) screening study with observed TT-HBV infection rates in several lookback studies. Study design and methods The WP and OBI risk was estimated from ID-NAT screening data in six geographic regions. The 50% infectious dose (ID50), a key factor in the applied risk models, was assumed to be 100-fold higher in OBI than in WP blood. The relative proportion of WP and OBI TT-risk was estimated for different screening scenarios and expressed as a percentage of the ID-NAT yield rate to allow for comparison with observed TT-rates in lookback studies. Results Despite sevenfold to eightfold higher HBV ID-NAT yield rates in OBI than WP in South-East Asia and Europe, our models predicted that 40 (26-55)% of total residual TT-HBV risk was due to OBI, comparable to 37% observed in a Japanese hemovigilance study. Modeled TT-OBI risk was approximately 10-fold higher than observed rates of 2%-8% in five lookback studies but comparable to one other study (36%). Conclusion Although the observed TT-OBI rate was generally lower than the modeled risk, the relative risk of WP versus OBI transmission was not incompatible with the observational infectivity data. This supports the validity of our assumptions in the infectivity-based models for estimating worst-case residual risk with different testing scenarios.