Adalimumab-induced autoimmune hepatitis in a patient with Crohn's disease

Biological therapies are increasingly being used for immune-mediated, inflammatory disorders. Minor side effects of treatment with TNFα blockers agents are common, including upper respiratory tract infections, asthenia, fever and headache. More serious side effects are less common. The following case report shows an autoimmune hepatitis (AIH) after beginning of adalimumab, to our knowledge, the first case described in a patient with Crohn's disease.

A 54-year-old male patient was diagnosed with ileocolonic Crohn's disease (Montreal Classification A3, L3, B1), and due to steroid and azathioprine refractory disease, TNFα blockers agent therapy was considered. Cutaneous tuberculin (Mantoux) test was positive, without signs of active tuberculosis infection on chest X-ray, so prophylactic treatment with isoniacide was started. After one month of prophylaxis, infliximab (IFX) was started with standard dose (intravenously 5 mg/kg of weight at weeks 0, 2 and 6). Due to a “serum sickness-like” reaction after the second dose, it was switched to subcutaneous adalimumab (ADA) (160 mg on day 0, 80 mg on day 14 and 40 mg each for 14 days). The presence of antinuclear antibodies was not assessed during infliximab therapy. Liver tests performed before starting azathioprine, isoniacide and infliximab, as well as periodically during such treatments, were normal. After 9 months, isoniacide was stopped. As clinical manifestations persisted, adalimumab dose was escalated up to 40 mg every 7 days, achieving clinical and biochemical (C reactive protein) remission. After six months of adalimumab therapy, blood test showed elevated liver enzymes (ALT 70 UI/mL, normal range up to 35; AST 102 UI/mL, normal range up to 37); results were confirmed two weeks later. Patient denied alcohol intake. Hepatitis C antibodies, Hepatitis B surface antigen and Hepatitis B DNA were negative. Ferritin, transferrin saturation index, ceruloplasmin, serum copper, alfa-1-antitrypsin, HFE mutations were also negative; antinuclear autoantibodies were positive (1/320), with homogeneous pattern; liver ultrasound examination and abdominal CT scan did not show any abnormalities. Percutaneous liver biopsy showed portal areas with inflammatory infiltrate of lymphocytes and plasmatic cells, with no periportal affection. Histiocytes and lymphocytes infiltrate the acini, with cytolysis foci, focal necrosis and mild steatosis. Simplified criteria for the diagnosis of autoimmune hepatitis revealed a score of 6, allowing the diagnosis of “probable autoimmune hepatitis”. Adalimumab was interrupted, and after one month liver tests and autoantibodies became normal. Clinical activity led us to start methotrexate, however ileo-hemicolectomy was necessary three years later because of obstructive symptoms. Azathioprine was not sufficient to prevent anastomotic recurrence so certolizumab was started in standard doses (400 mg subcutaneously at weeks 0, 2 and 4, and 400 mg every 4 weeks there after). No relevant liver test abnormality was observed after 3 years of follow up.

Drug induced autoimmune hepatitis (DIAIH) is only documented in the literature as case reports and case series. Differentiation between AIH, DIAIH, and drug-induced liver injury (DILI) can be extremely difficult; chronologically correlation between liver test and initiation and withdrawal of drugs and liver biopsy may help. As explained in our case, normalization of antinuclear antibodies and transaminase levels after etiological agent removal is highly suggestive of adalimumab induced-AIH, although we cannot confirm the absence of antinuclear antibodies after first biological therapy, infliximab. Moreover, non-immune-mediated liver injury has been described during TNFα blockers therapy.

Autoimmune disorders are documented side effects of TNFα blockers agents, and they are increasing as these agents become more frequently used. Autoimmune hepatitis is one of the most recently described adverse event. Twenty-five cases of TNFα blocker-AIH have been described in the literature, 19 IFX-induced, 4 ADA-induced and 2 etanercept-induced, predominantly in patients with seronegative spondyloarthropathies. Rodrigues et al.¹ reported eight cases of DIAIH among inflammatory bowel disease, rheumatoid arthritis and ankylosing spondylitis patients under antiTNFα treatment; all of them received successful corticosteroid therapy. There are four biopsy-proven ADA-induced AIH cases reported in the literature.^1–4 All of them arise in patients diagnosed with rheumatic disease patients and need for corticosteroids to reach remission. As in the present case, it has been previously described absence of recurrence of antiTNFα induced-AIH after switching of antiTNFα agent,⁵ suggesting that AIH is not a class effect of antiTNFα⋅

In conclusion, DIAIH is a rare adverse event that should be taken into account by all physicians using antiTNFα drugs. High degree of suspicion should promptly drive us to perform a liver biopsy and withdraw the drug. Corticosteroid and other immunosuppressant may be useful, and monitor liver laboratory tests is recommended in case of substitution of the drug by a same-class therapy, although relapse has not yet been described.

Conflict of interest

None.