Fatal hepatitis B virus reactivation by an escape mutant following rituximab therapy

The chimeric monoclonal anti-CD20 antibody rituximab targets and destroys B lymphocytes and is used increasingly in the treatment of B-cell non-Hodgkin lymphoma.^1,2  In spite of rituximab-induced B-cell depletion, the incidence of infectious complications does not appear to be generally increased, but there have been isolated reports of viral infections^3-5  and hepatitis B virus (HBV) reactivations in hepatitis B surface antigen (HBsAg) carriers.⁶  We present the case of a previously HbsAg-negative patient with high titers of antibodies against HBsAg (anti-HBs) who developed fulminant hepatitis B following the administration of rituximab.

A 73-year-old male had been treated with 5 courses of cyclophosphamide/doxorubicin/vincristine/prednisone (CHOP) chemotherapy from March 2002 to June 2002 and subsequent rituximab from July 2002 to September 2002 for diffuse large-cell lymphoma. The patient had suffered from acute hepatitis B 15 years ago. He had developed complete resolution with high titers of anti-HBs antibodies (> 1000 IU/L) and no serological evidence of HBsAg in the routine test (Table 1). In September 2002, the patient was admitted to the hospital with acute hepatitis. HBsAg and a high load of HBV DNA (8.6 × 10⁸ copies/mL) were detected, whereas anti-HBs antibodies remained positive (868 IU/L). Under antiviral treatment with lamivudine, the viremia decreased to 2.4 × 10⁵ copies/mL within 2 weeks. Nevertheless, hepatic function did not recover, and the patient died 19 days after admission.

Sequence analysis of polymerase chain reaction (PCR) products amplified from the S region revealed 5 remarkable mutations in the major antigenic region (L110R, R122K, Y/F134S, P142L, and D144A), the latter 3 of which have already been noted in HBV escape mutants after vaccination failure.⁷  These mutations are compatible with an escape of an endogenous HBV strain from the patient's own anti-HBs that was nonpathogenic until the administration of rituximab. A highly sensitive quantitative PCR⁸  for the HBV DNA S region revealed a small amount of homogeneous HBV DNA (50 ± 20 copies/mL, Table 1), with an identical HBsAg sequence in a 2-year-old blood sample from the patient that had tested negative for HBsAg.

HBV replication has been shown to persist at very low levels after resolution of acute hepatitis B, regardless of coexisting anti-HBs, and is kept at low levels by antiviral cellular immune responses.⁹  Despite the potential persistence of HBV in hepatocytes and mononuclear cells, symptomatic HBV infection is uncommon in patients with pre-existing anti-HBs antibodies after conventional chemotherapy. However, since the anti-CD20 antibody was introduced, this case is already the second instance of fulminant hepatitis B reactivation in an anti-HBs–positive patient after the combined administration of CHOP and subsequent rituximab.¹⁰  Our findings suggest that HBV escape mutants can be essential for this reactivation. Testing for hepatitis B core and surface antibody is necessary before administration of rituximab, and close monitoring for active HBV infection by quantitative HBV DNA assays needs to be performed during anti–CD20 antibody treatment in positive cases, since HBsAg routine tests may be unreliable in heavily mutated variants. Controlled trials will have to show if prophylactic antiviral therapy is justified for patients with a history of hepatitis B receiving rituximab.