Role of hepatitis B antibody in predicting reactivation of resolved hepatitis B virus infection in leukemia patients.

Background & Aims: Quantification of anti-HBs and anti-HBc predicts the risk of HBV reactivation (HBVr) in lymphoma patients receiving rituximab treatment. However, it remains unclear whether the quantification is predictive of HBVr in leukemia patients undergoing immunosuppression.

Methods: and patients: Clinical and laboratory data of the leukemia patients with resolved HBV infection diagnosed between January 2013 and March 2018 were retrospectively collected. Data series of HBV seromarkers and HBV DNA levels before the patients receiving chemotherapy and/or hematopoietic stem cell transplantation (HSCT) and during follow-up duration were analyzed.

Results: In total, 533 leukemia patients with resolved HBV infection were included. The incidences of HBVr were 5.7% (25/441) and 2.2% (2/92) in patients receiving HSCT and chemotherapy, respectively. In patients receiving HSCT, acute lymphoid leukemia had a significantly higher incidence of HBVr than acute myeloid leukemia (8.9% vs 3.9%, P < 0.05). The incidence varied almost zero to 40% due to the differences in the profiles of HBV antibodies. High anti-HBs (cut-off of 79.2 IU/L) or low anti-HBc levels (cut-off of 4.475, S/CO) at baseline were associated with a low risk of HBVr. Anti-HBe status did not affect the incidence of HBVr. However, the cut-offs were only predictive of HBVr in the patients who had negative anti-HBe.

Conclusion: The baseline profiles of HBV antibodies are predictive of the risk of HBVr in leukemia patients undergoing immunosuppression. However, seronegative anti-HBe is a prerequisite for using baseline anti-HBs and anti-HBc quantification to predict HBVr risk.