Background Hepatitis B virus (HBV) infection is frequent worldwide. In patients with immune mediated inflammatory diseases (IMID), immunosuppressive therapy may trigger HBV reactivation, leading to significant morbidity and mortality.

Objectives To describe presentation, management and outcome of HBV reactivation occurring in patients treated for IMID, and to evaluate predefined algorithm for its prevention.

Methods French centers have included 35 patients with HBV reactivation diagnosed between January, 2002 and March, 2012. HBV reactivation was defined as an increase >1 log IU/mL of HBV DNA levels or DNA reappearance in negative patients. Hepatitis was defined as an increase >3-fold the baseline value of alanine transaminase (ALT). We further performed an extensive literature review and provided a global analysis of 138 cases of HBV reactivations.

Results Personal cases were treated for rheumatoid arthritis (RA, n=14), connective tissue disease (n=7), vasculitis (n=5), ankylosing spondylitis (n=4) or other diseases (n=5). At baseline, 23 (66%) patients were hepatitis B surface antigen (HBsAg) carriers, 11 had previous history of HBV infection (including 7 with HBs antibodies), and 1 patient had occult HBV infection.

Reactivation occurred after a median time of 39 wk after initiation of corticosteroid (CS) and/or immunosuppressive (IS) therapy. At the time of reactivation, 30 (86%) patients were receiving CS, 11 (31%) methotrexate, 7 (20%) TNF-ablockers, 6 (17%) cyclophosphamide, 4 (11%) rituximab, 4 (11%) azathioprine, and tocilizumab and abatacept in 1 case each (3%). Median HBV DNA and ALT levels were 4.2 log IU/mL and 2-fold the baseline value, respectively, and were correlated (r=+0.49, P=0.004). Patients were clinically asymptomatic in 31 (89%) cases, while hepatitis occurred in 17 (49%), including severe hepatitis (>10-fold the baseline value) in 9 (26%). Management consisted in antiviral therapy in 32 (91%) patients (mainly entecavir and lamivudine), associated with discontinuation or decrease of CS/IS in 16 (46%). Neither fulminant hepatitis was noted, but one patient died of hepatocellular carcinoma 4 months after reactivation.

After global analysis of HBV reactivations, reported patients were clinically asymptomatic in 102 (74%) cases, with severe hepatitis in 46 (33%) and death and/or fulminant hepatitis in 17 (12%). Reactivation kinetics differed according to the treatments used and baseline HBV status, with earlier reactivation occurring under rituximab or cyclophosphamide and in HBsAg+/HBV DNA+ patients (Figure). The use of predefined algorithm could have prevented 108 (78%) reactivations. Two reactivations occurred despite appropriate preemptive antiviral therapy. Finally, according to the algorithm, 28 patients would not have received preemptive therapy, including 2 HBcAb+/HBsAb+ Asian patients with RA receiving methotrexate or adalimumab who died of fulminant hepatitis.

Conclusions This study provides new insights into HBV reactivations in patients with immune mediated inflammatory diseases. Predefined algorithm seems to be effective to reduce the risk of HBV reactivation, but caution is warranted using monitoring of HBV markers.

Disclosure of Interest None Declared