Gut: first published as 10.1136/gut.27.11.1390 on 1 November 1986. Downloaded from Gut, 1986, 27, 1390-1393 Case reports Fatal hepatitis associated with diclofenac E G BREEN, J McNICHOLL, E COSGROVE, J MCCABE, AND F M STEVENS From the Department of Medicine, Regional Hospital, Galway, Eire SUMMARY Non-steroidal anti-inflammatory agents (NSAIDS) are a well recognised cause of hepatotoxicity. Diclofenac, a relatively new NSAID, was first introduced into the UK in 1979. Five cases of hepatitis have recently been reported, principally in the French literature. -5 We report the first fulminant case of hepatitis in the English literature in a patient taking diclofenac and indomethacin. Diclofenac is a member of the arylalkanoic group of 100 mg per day for five weeks. Ferrous sulphate one NSAIDS (Fig. 1). Three other agents in this group tablet daily was added on 16 May. The patient was have been shown to be significantly hepatotoxic. admitted to hospital on 26 June. A week before this Ibufenac was withdrawn from circulation because of he had felt unwell with anorexia, nausea, abdominal the frequent rise in transaminases,6 7 the use of discomfort, and dark urine. On admission he was benoxaprofen was stopped in Britain after 10 icteric, the liver edge was palpable 4 cm below the patients died with hepatitis8 9 and more recently a costal margin and there were no signs of chronic fatal case of hepatitis due to pirprofen has been liver disease. Ultrasound showed early ascites with reported."' Early reports about diclofenac showed no obstruction of the biliary tract. A week after that it caused a rise in liver function tests.'' Since admission he developed a maculopapular rash 1983 five reports of hepatitis caused by diclofenac maximal on the trunk, diffuse lymphadenopathy, http://gut.bmj.com/ have been published, including that of a 55 year old and a pyrexia of 39°C. The eosinophil count was French woman, who died with fulminant hepatitis 399x 1)9/l. Laboratory investigations were consist- after a three week course of diclofenac.'-5 ent with fulminant hepatitis and serological markers Indomethacin has been in clinical use for over 20 for viral infection were absent (Table). The patient years and two patients have died from fulminant had no history of foreign travel, blood transfusion, hepatitis caused by indomethacin. 12 '" Other reports alcohol abuse, exposure to toxic chemicals or of hepatotoxicity exist but are infrequent.'4 ' We describe a patient who developed fulminant hepati- on September 28, 2021 by guest. Protected copyright. tis after one month's treatment with indomethacin followed by two months' treatment with diclofenac. CH2COOH Case report A 56 year old farmer started treatment for osteo- arthritis on 9 April, 1985. Before this he had no medical illness and was not taking any medication. He received an intramuscular injection of diclofenac %NH (50 mg) and started indomethacin 75 mg per day for 10 days, and 50 mg per day for the following two weeks. On 2 May he received a further intramus- cular injection of diclofenac (50 mg) and oral diclofenac was substituted for indomethacin. The dose was 75 mg per day for two weeks followed by Address for corrcsponidcncc: I)r E (G 13rccn. Depirtmient of Mclicinie. Rcgional Flospital. (Cilway. Eire. Rcccivcd for publiciation 1f) Mairch 1980. Fig. C(hemicalstrff(trcre of diclofienac. 1390 Gut: first published as 10.1136/gut.27.11.1390 on 1 November 1986. Downloaded from Fatal hepatitis associated with diclofenac 1391 homosexuality. His condition gradually deterio- and hydrocortisone were started. On 13 July aster- rated. The liver reduced in size, the prothrombin ixis and foetor hepaticus were present and the time rose to 37 sec, the serum albumin fell to 23 g/l patient became comatose. On 17 July urine output and gross ascites developed. Lactulose neomvcin fell to 300 ml per day, and the platelet count was 30 000, with no laboratory evidence of disseminated intravascular coagulation. The patient died on 20 Table Laboratory data oni admissioni July, 24 days after admission. Necropsy showed a small liver which weighed 950 g. The histopathology Liver futnc tion telsts Normtial ritoige was of massive hepatic cell necrosis with extensive Albumin 36 g/l (34-48 g/l) centrilobular liver cell dropout (Fig. 2). The portal Prothrombin time 22 secs (14-16 sec) triads showed intense inflammatory infiltrate and Alkaline phosphatasc 321 U/I (30-115 U/1) bridging necrosis was present in the relatively better Bilirubin 115 LImol/l (1-17 LtlloUI) preserved aireas. There was no fibrosis. Bile plug- Alanine aminotransferasc 2110O U/I ((0-4 U/I) Viral screcn ging was present in the renal tubules and the other Hepatitis A- IgM Negativc organs were normal. Hepatitis B - sAg Negative Hepatitis B - IgM - cAntibody Negatis c Discussion Cytomegalovirus Negativc Ebstein-Barr virus Negative Herpes Simplex virus Negatitsc This patient died from fulminant hepatitis, as shown Autoantihodies by the absence of clinical, or histological evidence of Nuclear Negatis c previous liver disease, the occurrence of hepatic Mitochondrial Negativ c Smooth muscle Negatis e encephalopathy three weeks after the onset of White ccll count 5 7xl())/l jaundice, the abnormal prothrombin time and the Ncutrophils 6h5/., massive hepatic cell necrosis. There are three Lymphocytes 2700% possible causes: non-A non-B hepatitis, indometha- Eosinophils 7°/o cin, and diclofenac. The diagnosis of non-A non-B hepatitis is one of exclusion. In this patient all http://gut.bmj.com/ on September 28, 2021 by guest. Protected copyright. Fig. 2 Microscopy of liver showing centriloblular necrosis. Gut: first published as 10.1136/gut.27.11.1390 on 1 November 1986. Downloaded from 1392 Breen, McNicholl, Cosgrove, McCabe, and Stevens screening for viral infection was negative. Two quently accompanies hepatitis caused by benoxa- injections with sterile disposable needles and syr- profen and fenclofenac is now known to cause inges were administered, but transmission of non-A interstitial nephritis. 9 23 This patient, however, non-B hepatitis by this route seems most unlikely. developed oliguria as a terminal event, and necropsy The manifestations of hypersensitivity favour a drug failed to show any specific findings suggestive of a aetiology rather than non-A non-B hepatitis.'" The drug induced nephropathy. predominance of centrilobular necrosis also favours Review of the five published reports of diclofenac a drug aetiology.'7 hepatitis, shows that four patients were women, and Reports of hepatotoxicity with indomethacin are the mean age was 56*8 years (range 45-72). The rare. Two deaths due to fulminant hepatitis have mean duration of treatment was 14 2 weeks (range been published.'2 13 On both occasions treatment 3-24). Two patients had eosinophilia and one was with 100 mg per day for two and a half and patient had a skin rash. On withdrawing the drug, seven months respectively, and treatment was in three patients showed rapid correction of liver progress when the hepatitis occurred. This patient function tests, and the other two patients died from had a smaller dose of indomethacin for a much hepatic failure, one five weeks and the other five shorter period, and treatment had stopped two months, after starting treatment. Three authors months before the onset of hepatitis. There were no suggest toxic metabolic mediated liver damage, and adverse effects or signs of a hypersensitivity reaction one author suggests an immunoallergic mechanism. during treatment with indomethacin. Those that did We consider the latter to be the most likely occur were unlikely to have been caused by a drug mechanism in this patient, in view of the presence of administered two months previously. 17 signs of hypersensitivity, and the exceptional nature Diclofenac is the most likely cause of hepatitis in of the liver damage. The eosinophil count was at the this patient. Hepatitis including fulminant hepatitis upper limit of the normal range, and thus neither due to diclofenac is well established.'-- It usually supports nor excludes this mechanism. occurs after an average of three months' treatment The recent increase in reports of hepatoxicity and may be associated with signs of a hypersensitiv- induced by NSAIDS warrants attention. This is the ity reaction. This patient had taken two months' first report in the English literature of fulminant treatment with diclofenac and hepatitis occurred hepatitis induced by diclofenac. Fortnightly assess- while taking the drug. The signs of a hypersensitivity ment of liver function for the first six months' reaction and the temporal relationship between the treatment with these new agents is advisable. http://gut.bmj.com/ onset of hepatitis and treatment with diclofenac, in Immediately stopping treatment on finding raised the absence of convincing evidence of another liver function tests or at the onset of symptoms such cause, supports diclofenac as being the offending as nausea, anorexia, or abdominal discomfort, is agent. obligatory. Corticosteroid treatment was not suc- The mechanism of arylalkanoic derivative in- cessful in this patient and its role in diclofenac duced hepatitis is not clear. Dunk and Danan favour induced hepatitis has yet to be evaluated. a metabolic rather than a hypersensitivity on September 28, 2021 by guest. Protected copyright. mechanism. I 1) They base this on the absence of hypersensitivity signs in some patients, the initial delay in the onset of hepatitis from the start of References treatment, and the late response to readministration of the drug.' 5 18 Alclofenac, another arylalkanoic I Dunk AA, Walt RP, Jenkins WJ, Sherlock SS. derivative, has been shown to form unstable epox- Diclofenac hepatitis. Br Med J 1982; 284: 1605-6. ides 2 Babany G, Pessayre D, Benhamou JP. Hepatitis which destroy cytochrome P-450.'9 This caused by diclofenac. Gastroenterol C/in Biol 1983; 7: mechanism may be at the basis of other arylalkanoic 316.