Objectives To report the onset of a severe and lethal medium-vessel panarteritis mimicking deep sepsis three weeks after etanercept withdrawal and minocycline therapy in a patient with RA.

Results A 50-year-old man with RA (7/7 1987 ACR criteria) had poorly responded to all DMARDs and low-dose steroids for 12 years. In September 1999, DAS-28 score of 6.8 led to a treatment by etanercept (25 mg twice a week) for 4 months who remained poorly effective too (DAS-28 after 3 months: 6.05). Although no side-effects occurred, the patient wished to stop etanercept. A short cure by minocycline was administered. Ten days after, the patient’s general condition worsened with fever (from 38°C to 39°C), hyperleukocytosis (24.109/ml), confusion, pleural effusion, hypoalbuminemia (17.2 g/l), epigastric pain, and CRP levels up to 295 mg/l. ANA and APL were negative, and ANCA only slightly positive (1/20, without anti-MPO antibodies). RF were 40 times the upper normal value. Extensive search for sepsis was unsuccessful. Highly false positive tests for myoglobin with ELISA kits using rodent antibodies were suggestive of anti-chimeric antibodies. The patient died in March 2000, three months after discontinuance of etanercept. An autopsy confirmed severe plurivisceral vasculitis typical of granulomatous, necrotising panarteritis, with both old and recent lesions involving mainly the medium-vessels in heart, muscles and gall-bladder.

Conclusion Although the relation of the patient’s death to treatments is uncertain, a flare of rheumatoid vasculitis could have been involved. This severe vasculitis occurring soon after etanercept withdrawal and presenting first as unexplained fever mimicked deep sepsis for several weeks and misled the treatment strategy.