Deschenes et al. raise several important questions relating to the use of hepatitis B surface antigen (HBsAg) positive donor for allogeneic bone marrow transplantation. First, the use of HBsAg positive donors, particularly those with a high pretransplant serum hepatitis B virus (HBV) DNA level, has been associated with the development of post–bone marrow transplantat HBV-related hepatitis.1 To this end, it is highly justifiable to reduce the HBV viral load in the donor in an attempt to reduce the subsequent risk of posttransplant hepatitis. In addition, our center has adopted the policy of treating all bone marrow transplant recipients of HBsAg positive donors with preemptive lamivudine to prevent posttransplant HBV-related liver diseases, irrespective of the recipients' pretransplant HBV serological status. This has restricted us from studying the impact of the HBV DNA level in donor marrow and subsequent posttransplant HBV-related liver diseases.

In our previous study, we did not test donor marrow samples for HBV DNA.2 In our ongoing study, we have tested 10 pairs of serum and mononuclear cells (derived from bone marrow) from the HBsAg positive donors for HBV DNA level, using our in-house quantitative assay based on polymerase chain reaction. Similar to the experience of Deschenes et al., we have found a positive correlation in HBV DNA level between serum and mononuclear cells. However, from the follow-up paired serum-peripheral blood mononuclear cells from these HBsAg positive donors treated with lamivudine, our preliminary results showed that the response to antiviral therapy is different in these 2 compartments, with a less well-defined response in the mononuclear cell compartment. Specifically, in 4 HBsAg positive donors treated with lamivudine for 12 weeks, despite the lowering of serum HBV DNA to more than 3 log, the HBV DNA level in the paired PBMC remain unchanged, suggesting that mononuclear cells from these treated HBsAg positive donor may still be infectious. Our study is investigating whether the HBV DNA level in the peripheral blood mononuclear cells will be lowered with prolonged lamivudine therapy. The discrepancy of the antiviral response observed in the serum and mononuclear cells' HBV DNA level might be related to the replication capacity of the HBV in the liver and mononuclear cells. Indeed, the biological behavior of HBV within the mononuclear cells compartment is still largely unknown. In keeping with this, the presence of covalently closed circular DNA molecules, the key replicative intermediate in the synthesis of the viral pregenome and transcripts during productive HBV infection, has not been convincingly demonstrated in peripheral mononuclear cells.3 Hence, it is not surprising to observe a discrepancy between the reduction of HBV DNA level in the serum, which reflects mainly active and productive HBV replication in the liver, and in the mononuclear cells compartment.

George K.K. Lau* ‡, Hai-Ying Zhang* ‡, John M. Luk‡, Stephen Locarnini?, Raymond Liang†, * Division of Gastroenterology and Hepatology, The University of Hong Kong, Hong Kong SAR, PR China, † Division of Hematology, Department of Medicine, The University of Hong Kong, Hong Kong SAR, PR China, ‡ Center for the Study of Liver Diseases, The University of Hong Kong, Hong Kong SAR, PR China, ? Victorian Infectious Diseases Reference Laboratory, North Melbourne, Victoria, Australia.