Fatal toxic epidermal necrosis: responsibility of diacerein? A controversy.

Toxic epidermal necrosis (TEN) is a life-threatening condition induced by a large variety of drugs, most often anticonvulsants, nonsteroidal anti-inflammatory drugs (NSAIDs) and sulphonamides; reports involving other drugs are available. We report the first case in which diacerein, an anthraquinone derivative recently introduced for rheumatological diseases, has been considered as a possible culprit. A 71-year-old woman was referred for acute, rapidly progressing, widespread mucocutaneous lesions accompanied with major vital signs. She had had joint pains of mechanical origin mainly located in the hips, wrists and shoulders for which she had been continuously treated with diacerein for 3 months before the occurrence of cutaneous lesions with a daily dose of 100 mg. Other recent drug intakes included oral spiramycin for dental abscess stopped 3 weeks before the cutaneous accident; she denied any NSAID intake during the preceding 3 months. Long-duration medications included flunitrazepam and lormetazepam, both of them introduced at least 2 years before and well tolerated. The mucocutaneous lesions developed suddenly with multiple mucous erosions, a generalized, macular, dark red exanthema, fever at 40°C and major general alteration. Nikolsky’s sign quickly appeared whereas diffuse blistering progressed to exfoliation in large sheets of more than 80% of her body surface over 48 h. The histological picture was typical of TEN with necrosis of the whole epidermis, dermo-epidermal cleavage and minor inflammatory infiltrate in the upper dermis. Direct immunofluorescence and a serological test for mycoplasma were negative. She presented two deleterious factors in TEN, lymphopenia and hyperglycaemia, and in spite of supportive care, she rapidly worsened with respiratory distress syndrome and bronchial mucosa exfoliation. She eventually died of lung sepsis 10 days after the beginning of the eruption. We consider that this case of fatal TEN might have been induced by diacerein. The diagnosis of TEN can hardly be questioned since all clinical and histological data were typical of this rare condition. Although flunitrazepam and lormetazepam should also be regarded as culprits, the role of diacerein was considered since it was the only recently introduced drug received by the patient when the TEN occurred; particularly, extensive investigations failed to demonstrate any recent NSAID intake. Spiramycin, a rare but possible TEN-inducing drug, had been stopped 3 weeks earlier, a time lapse theoretically inconsistent with its involvement given its short half-life (6–8 h). By contrast, in our patient, a free interval longer than the usual 2–3 weeks after the beginning of a continuous treatment, although rare, is not an exclusion criterion [1]; this unusual interval can be connected with the unusual delay of efficacy of the drug which is about 45 days [2]. Diacerein is an original molecule supposed to exhibit an anti-IL1 activity, without influencing the cyclo-oxygenase activity. It shows no frank structural homology with NSAIDs. Serious cutaneous sideeffects have not been reported to date. A comprehensive survey of its cutaneous side-effects is advocated.