Article Text

THU0443 Anakinra Might Cause Acute Hepatitis in Patients with Systemic Onset Juvenile Idiopathic Arthritis and Adult Onset Still Disease
  1. A. Meyer1,
  2. M. Tebacher1,
  3. F. Beau-Salinas2,
  4. L. Rossi3,
  5. I. Kone-Paut3,
  6. A. Mathian3,
  7. Z. Amoura3,
  8. A. Mekinian3,
  9. O. Fain3,
  10. F. Jazeron2,
  11. S. Rist4,
  12. F. Despert2,
  13. J.-E. Gottenberg1,
  14. J. Sibilia1
  1. 1CHU, Strasbourg
  2. 2CHU, Tours
  3. 3APHP, Paris
  4. 4CH, Orléans, France


Background Systemic onset juvenile idiopathic arthritis (SoJIA) and adult onset still disease (AoSD) are unlicensed indications for anakinra. Although safety has been reported to be good, increase of liver enzymes have been noted in four SoJIA patients1,2,3 and we have recently described acute hepatitis in two patients with AoSD4,5.

Methods The charts of six patients with SoJIA (n=3, ILAR criteria) and AoSD (n=3, Yamaguchi criteria) who developed acute hepatitis during anakinra treatment were retrospectively studied. Strength of association between liver injury and the medication was assessed through the Roussel-Uclaf Causality Assessment Method6.

Results All patients were female taking anakinra for 27 days (range 21-39) at hepatitis diagnosis. Cytolysis was variable (aspartate aminotransferase: 1041UI/L (range 336-2386), alanine aminotransferase (ASAT): 2647 UI/L (range 568-3346)), cholestasis was moderate compared to cytolysis (alkaline phosphatase (ALP): 262UI/L (range 71-376), total bilirubin 94UI/L (range 17-171) Gamma-glutamyl transpeptidase 397UI/L (range 62-538)) with ALAT/ALP ratio >5 in all cases (range 7-13). Factor V amount were systematically >100%.

One patient had vomiting, the others were asymptomatic, and none had clinical evidence of SoJIA or AoSD activity. Median CRP, ferritin and polynuclear count were respectively 0.8mg/dL (range 0.1-7.0), 409µg/L (range77-3108) and 4170/mm3 (range 1500-10650). Non-drug related causes were systematically researched. Acute viral infection were excluded in all patients by serology (VHA, VHB, VHC, (n=6), VHE (n=5), HSV 1&2 (n=6), EBV (n=6), CMV (n=4)) but one patient had detectable EBV genome (2450 copies/ml). None patient tested positive for anti-smooth muscle, anti-mitochondria, anti-LKM1, anti-SLA or anti-liver cytosol antibodies. When tested (n=4), cupremia and ceruloplasmin were normal. Abdominal ultrasound revealed moderate hepatomegaly in one patient and no abnormalities in the others. Four patients underwent liver biopsy that did not show autoimmune hepatitis or hemophagocytosis but nonspecific hepatitis.

Withdrawal of anakinra, made 4 days after the diagnosis of hepatitis (range 1-27), was followed by liver enzymes normalisation in all patients (time to reaction 3.5 days (range 1-56); time to 50% improvement 26 days (range 4-64)). Co-treatments with a compatible latency and that were also withdrawn were omeprazole (n=2), dalteparin (n=1), tetrazepam (n=1). When anakinra was re-administrated (n=2), hepatitis relapsed.

According to the Roussel-Uclaf Causality Assessment Method, anakinra was the definite cause of hepatitis (score>8) in two patients and the probable cause (score 6-8) in the four others.

Conclusions In SoJIA and AoSD patients treated with anakinra, liver enzymes should be monitored as this treatment might cause acute cytolytic hepatitis.


  • Quartier et al. Ann Rheum Dis. 2011; 70:747-54.

  • Nigrovic PA et al. Arthritis Rheum. 2011; 63:545-55.

  • Canna et al. Pediatr Rheumatol Online J. 2009;7:21

  • Meyer et al. Congrès de la Société francaise deRhumatologie. Paris: Elsevier; 2011. A264.

  • Diallo et al. Rev Med Interne. 2012 [Epub].

  • Danan and Benichou. J. Clin. Epidemiol. 1993; 46:1323–1330.

Disclosure of Interest None Declared

Statistics from

Request permissions

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.