Diagnostic algorithm for HBV safe transfusion.

Abstract

Nucleic acid testing (NAT) has been introduced in blood donation screening in the late 1990s and was progressively extended with the availability of commercial assays initially for HCV, to HIV-1 and more recently to HBV1, 2. The main objective of HCV and HIV-1 RNA screening was the detection of window period infections (WP) for viruses that replicate very quickly with a viral load doubling time of 14 and 20 hours, respectively3. As a result, substantial increase in concentration of viral genomes is achieved in a short period of time ( 10E4IU/mL. Under such circumstances, the dilution factor introduced by pooling 24–96 plasma samples from individual donations did not substantially affect the identification of HCV and HIV-1 WP infections. However, plasma pooling being introduced for the economic purpose of decreasing the cost of NAT, several cases of HIV-1 transmission by lack of assay sensitivity related to sample pooling were reported4–6. Despite these few failures, the impact of NAT for HCV and HIV-1 reduced the residual risk of transfusion-transmission of these viruses below 1 per million donations in most developed countries where it was introduced2, 7. The issue of decreased sensitivity due to sample pooling became considerably more critical when HBV DNA screening was added to HCV and HIV-1 in triplex assays able to simultaneously detect all three viral genomes. Contrary to the relatively simple situation of HCV and HIV-1 where nearly 100% of samples containing viral RNA after the WP were seropositive for virus specific antibodies, discrepancies between serological HBV screening with HBsAg and DNA were found with considerable frequency8. Depending on the studies and the HBV genotype, 2–15% of HBV infected samples contained HBsAg without detectable DNA9 or DNA without detectable HBsAg10. As a result, 65–95% of potentially infectious HBV-containing donations were reactive to both types of assays. In addition, during the window period, the viral replication doubling time is considerably longer than for HCV and HIV-1 (estimated between 2 to 3 days) and it takes several weeks before viral load reaches concentrations ranging between 10E3 to 10E5 IU/mL and HBsAg becomes detectable3, 11. In this situation, the effect of pool size becomes more critical for WP identification. In addition, preliminary data collected in the past 10 years as well as more recent data obtained in areas where HBV DNA was screened in blood donor populations revealed substantial numbers of what is now known as 'occult' HBV infection (OBI) or carriage. This condition is defined as viral DNA without detectable HBsAg observed after the initial period of primary infection and most of the time accompanied by the presence of anti-HBc12. OBIs identified in apparently healthy blood donors have been further studied and characterised serologically, molecularly and to some extent clinically8, 13–15. OBIs are found mostly in older men, the viral load is generally below 1,000 IU/mL, in most cases <100 IU/mL, accompanied by anti-HBc and normal ALT level. Approximately half of these cases also carry anti-HBs indicating viral persistence in otherwise resolved infections. In very few cases, OBI may be negative for both anti-HBc and anti-HBs corresponding to 'primary' occult HBV infection16, which is defined as acute HBsAg negative infection. The very low viral load observed in most OBIs has two important consequences: 1. sample pooling considerably reduces the yield of cases and affects blood safety improvement; 2. It complicates the screening algorithm as well as the confirmation process. The present review will focus on the issue of confirmation of blood donation samples positive for HBV DNA but negative for HBsAg.