Objectives The aim of this prospective study was to examine the safety of anti-tumour necrosis factor (TNF) therapy in patients with rheumatic disease and hepatitis B virus (HBV) infection.
Methods 14 patients with chronic HBV infection, 19 HBV-vaccinated patients and 19 patients with resolved HBV infection were included in the study. All HBV-infected patients received combination therapy with oral antivirals and anti-TNF agents. During treatment the levels of hepatitis B surface antibodies (anti-HBs) in HBV-vaccinated patients and of serum HBV DNA in patients with chronic or resolved HBV infection were monitored.
Results No viral reactivation was observed in patients with resolved HBV infection while anti-HBs titres decreased during anti-TNF treatment in vaccinated patients, similarly to patients treated with methotrexate alone. None of the HBV-infected patients developed liver decompensation or a significant increase in alanine aminotransferase levels. One patient (7%) treated with lamivudine and etanercept showed viral reactivation due to the emergence of a lamivudine-resistant mutant strain.
Conclusions Anti-TNF agents represent a safe option for patients with chronic HBV infection when combined with antiviral therapy, as well as in patients previously exposed to HBV receiving no HBV prophylaxis. Resistant HBV strains may arise in patients with chronic hepatitis B, necessitating the initial use of anti-HBV agents with a low risk of resistance.
Statistics from Altmetric.com
Immunosuppressive agents can induce viral reactivation and serious clinical complications (hepatitis, liver decompensation, death) in patients with chronic hepatitis B virus (HBV) infection not receiving appropriate antiviral prophylaxis and, more rarely, in patients with resolved HBV infection.1 2 Patients with rheumatic diseases treated with high or even low doses of immunosuppression including steroids, disease-modifying antirheumatic drugs3 and anti-tumour necrosis factor (TNF) agents4 can also develop HBV reactivation; appropriate screening for HBV before starting anti-TNF treatment is therefore recommended.3 Data regarding the rate of HBV reactivation in patients treated with anti-TNF agents are limited and are derived mainly from single case reports or retrospective studies.4,–,8
We conducted a prospective study of the liver-related safety of anti-TNF agents when given in combination with antiviral drugs in patients with rheumatic diseases and chronic HBV infection, as well as in HBV-vaccinated patients and in patients with resolved HBV infection.
One hundred and thirty-one consecutive patients with rheumatic diseases starting anti-TNF therapy at an Outpatient Rheumatology Clinic were included in this prospective study (see figure 1 in online supplement and table 1).
Serological screening for HBV infection with hepatitis B surface antigen (HBsAg), antibody against hepatitis B core antigen (anti-HBc) and hepatitis B surface antibodies (anti-HBs) was performed in all patients (AXSYM; Abbott Laboratories, Chicago, Illinois, USA).3 Patients with chronic HBV infection were also tested for hepatitis B virus early antigen (HBeAg), antibody to hepatitis B ‘e’ antigen (anti-HBe) and serum HBV DNA (by PCR assay; Cobas Amplicor HBV Monitor, Roche, Basel, Switzerland; sensitivity 300 copies or 60 IU/ml). Additional diagnostic investigations were individualised in consultation with a hepatologist and included ultrasound of the upper abdomen, α-fetoprotein measurement and liver biopsy (scored for necroinflammation grade 0–18 and fibrosis stage 0–6 according to the Ishak scoring system).9 These patients were further classified into HBsAg inactive carriers and cases with chronic hepatitis B according to standard guidelines (table 2),10 11 and started antiviral therapy with an oral nucleos(t)ide analogue before anti-TNF treatment (table 2). Patients who were already receiving antiviral therapy remained on the same regimen. In general, inactive HBsAg carriers were treated with lamivudine while, in patients with chronic hepatitis B, newer agents with low resistance rates (entecavir, tenofovir) were also administered (see figure 1 in online supplement).
At baseline and at each subsequent visit (every 2–3 months), clinical indices of rheumatic disease activity and the use of antirheumatic drugs were recorded in all patients while alanine aminotransferase (ALT) and HBV DNA levels were measured every 3 and 3–6 months, respectively, in patients with chronic HBV infection. Patients showing virological breakthrough during treatment (detection of serum HBV DNA in patients with previously undetectable HBV DNA or >1 log10 increase compared with treatment nadir)2 underwent sequencing analysis to look for specific drug-induced viral mutations (Trugene HBV Genotyping Kit; Siemens Medical Solutions Diagnostics, Tarrytown, New York, USA). Levels of anti-HBs antibodies were measured in 19 HBV-vaccinated patients (anti-HBs+) treated with anti-TNF agents and 8 patients treated with methotrexate (MTX, control group).
Statistical analysis was performed using the two-sided Fisher exact test for comparison of proportions and the Wilcoxon rank test for comparison of quantitative variables between the different patient subgroups (SPSS Version 16.0, SPSS Inc, Chicago, Illinois, USA). For repeated measurements, the Wilcoxon signed ranks test for paired samples was used. Statistical significance was assessed at the p=0.05 level in a two-tailed test.
Resolved HBV infection
During follow-up, none of the 19 patients with resolved HBV infection (10 anti-HBs/anti-HBc+, 9 anti-HBc+) had raised levels of ALT or HBsAg or HBV DNA positivity (by PCR, median interval between testing 12 months).
Change in anti-HBs titres during antirheumatic treatment
In the 19 HBV-vaccinated patients, anti-HBs levels fell during anti-TNF therapy (median 163 IU/ml before treatment, 105 IU/ml after treatment, p=0.001, see figure 2 in online supplement), but in only one patient (5%) was the titre below 10 IU/ml (cut-off for protective immunity to HBV). A similar decrease in anti-HBs levels was observed in eight vaccinated patients with rheumatic diseases treated with MTX alone for a similar period of time (median 544 IU/ml before treatment, 394 IU/ml after treatment, p=0.02).
Chronic HBV infection
Fourteen patients with HBeAg-negative chronic HBV infection (HBeAg−/anti-HBe+)12 fulfilled criteria for anti-TNF treatment13: six received etanercept, four adalimumab and four infliximab (see table 2 and figure 1 in online supplement). Three non-responding patients were switched to another anti-TNF agent (one from adalimumab to infliximab and two from etanercept to adalimumab). Eight patients were classified as inactive HBsAg carriers and six as suffering from chronic hepatitis B (table 2).2 10 11 In nine patients (64%) the HBV infection was diagnosed simultaneously or after the rheumatic disease (median interval 4.9 years), while in the remaining five patients the chronic HBV infection was known before the diagnosis of the rheumatic disease (median interval 2 years). In three patients with rheumatoid arthritis referred to our centre, an acute exacerbation of chronic hepatitis B (ALT >10 times the upper limit of normal values) was observed during treatment either with low-dose steroids (5 mg/day) and/or methotrexate (10 mg/week).
None of these patients had clinical, imaging or histological evidence of cirrhosis, while all had normal or near normal liver enzymes (ALT, 24±10 IU/l) and undetectable HBV DNA (<60 IU/ml or 300 copies/ml) at initiation of anti-TNF therapy. All received antiviral therapy with an oral nucleos(t)ide analogue (11 lamivudine, 2 entecavir, 1 telbivudine) starting either before (n=9, median interval 6.8 months) or together with (n=5) anti-TNF therapy (see figure 1 in online supplement). Among the three patients with chronic hepatitis B treated initially with lamivudine, two were subsequently switched to another agent. One was switched to entecavir when a liver biopsy showed chronic hepatitis (grading score = 4, fibrosis score = 3), and the second developed viral reactivation due to the emergence of a lamivudine-resistant viral strain while on etanercept therapy and was switched successfully to tenofovir (figure 1). ALT levels did not change significantly during treatment (see figure 3 in online supplement), although mild (1–2× ULN) transient increases in ALT levels were noted in four patients (29%) without an associated increase in HBV DNA levels.
A number of studies have clearly shown that immunosuppressive agents given without appropriate antiviral prophylaxis can induce viral reactivation and liver decompensation in patients with haematological or neoplastic diseases and chronic HBV infection.2 There are limited data regarding the rate of HBV reactivation in patients with rheumatic diseases treated with anti-TNF agents.3 In a recent literature review, Zingarelli et al4 identified 27 HBV-infected patients treated with anti-TNF agents; HBV reactivation was documented in 14% of patients treated with lamivudine compared with 73% of patients without HBV prophylaxis.
Our data show that the pre-emptive use of an oral anti-HBV agent can prevent HBV reactivation in most patients (>90%) with chronic HBV infection treated with anti-TNF agent(s). Only one patient (7%) developed viral reactivation due to a lamivudine-resistant strain during anti-TNF treatment. This rate of viral resistance among patients with chronic hepatitis B treated with lamivudine/anti-TNF agents (33%) is similar to published rates of resistance after 2–3 years of lamivudine monotherapy (38–49%).10 14 Our patient is the second published case of viral reactivation due to a documented lamivudine-resistant viral strain emerging during anti-TNF therapy.7 The significance of these findings are twofold: (1) they emphasise the need for close monitoring with serum HBV DNA during anti-TNF therapy, as it allows for early diagnosis and prompt treatment of viral reactivation; and (2) they support the current view of using anti-HBV agents with low risk of resistance (such as entecavir or tenofovir) as initial therapy for patients with chronic hepatitis B given long-term immunosuppression.11 15
Patients with resolved or occult HBV infection can develop HBV reactivation after chemotherapy, transplantation2 or even anti-TNF therapy.16 17 However, in our prospective study, none of the 19 patients with resolved HBV infection experienced viral reactivation during anti-TNF treatment. Similar findings were reported recently by Charpin et al.18 Nevertheless, continuous vigilance with frequent measurements of HBsAg and/or HBV DNA (every 6–12 months) is needed in order to detect early viral reactivation.
In agreement with a recent report,18 we observed a slight decrease in anti-HBs levels during anti-TNF treatment in HBV-vaccinated patients. A comparable decrease was observed in a control group of patients treated with MTX alone, indicating no specific effect of anti-TNF on HBV protective immunity.
The limitations of our study include its relatively small size and short follow-up (approximately 2 years). The high proportion of patients with chronic HBV among our population treated with anti-TNF agents (11%) may be explained by the fact that the study was performed at a referral centre for liver diseases.
In conclusion, our study suggests that anti-TNF agents can be safely used without increasing HBV replication in patients with chronic HBV infection receiving oral anti-HBV therapy as well as in patients exposed to HBV receiving no HBV prophylaxis. For patients with chronic hepatitis B, agents with a low risk of resistance should be used (entecavir or tenofovir) as initial therapy in order to avoid the emergence of resistant strains.
Detail has been removed from these case descriptions to ensure anonymity. The editors and reviewers have seen the detailed information available and are satisfied that the information backs up the case the authors are making.
Funding This work was supported in part by a research grant from the Hellenic Society for Rheumatology.
Competing interests None.
Ethics approval This study was conducted with the approval of the institutional review board of the Hippokration General Hospital, Athens, Greece and all patients signed an informed consent prior to their enrolment in the study.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.