|
|
Wing Tat Poon Department of Clinical Oncology Case History A 52-year-old man presented 19 months prior to this episode with right-sided weakness. He was subsequently diagnosed with inoperable anaplastic astrocytoma of the left basal ganglion. He underwent radical whole brain irradiation and his symptoms were controlled until 3 months ago when he complained of increasing right-sided weakness with ipsilateral pain. Computed tomography of the brain confirmed progressive disease with a new lesion in the left frontal lobe with surrounding oedema. Four weeks prior to the current presentation, he was commenced on dexamethasone 8 mg/day plus physiotherapy. Sodium valproate was given for neuropathic pain. Figure 1. Erythematous rash on trunk. Figure 2. Mucosal involvement with haemorrhagic areas on the oral mucosa. Figure 3. Mucosal involvement with haemorrhagic areas on the conjunctiva. Figure 4. Mucosal involvement with haemorrhagic areas on the genitalia. Two days after the commencement of sodium valproate, the patient developed a widespread erythematous rash. There were multiple target lesions involving the trunk (figure 1) and limbs with blister formation that lacerated easily. There was mucosal involvement with haemorrhagic areas on the oral mucosa (figure 2), conjunctiva (figure 3), and external genitalia (figure 4). Sodium valproate was discontinued. Systemic antibiotic therapy (klacid) was commenced and topical fusidic acid was applied to the open areas. Benzydamine mouthwash and moistening eyedrops were administered. The lesions gradually healed during the following week. Dexamethasone was gradually withdrawn since there was no evidence of neurological improvement. Discussion Stevens-Johnson syndrome is a severe and extensive type of erythema multiforme. It is characterised by a severe necrotising cutaneous reaction and involvement of several mucosal sites (oral, conjunctival, and anogenital). There is a variable degree of constitutional symptoms and visceral involvement. Late complications include pigmentation, scarring, and skin contracture. The mortality rate of Stevens-Johnson syndrome is 5% to 15%. Common aetiological factors include drugs or viral infection. Anti-epileptic drugs are known to be associated with Stevens-Johnson syndrome within the first 8 weeks of treatment, especially phenytoin, phenobarbitone, sodium valproate, and carbamazepine.1-3 Treatment should be individually tailored according to the cause and complications of the syndrome. The most severe cases should be considered for treatment in a burns unit. Systemic corticosteroids should not be used routinely, but may be justified in the early stages of drug induced Stevens-Johnson syndrome. They should be given in relatively high doses (80-100mg prednisolone per day orally) for 2 to 3 days or less if disease is controlled. The dose should be tapered quickly but cautiously since no further benefit can be expected. Treatment should also focus on early detection and prevention of common fatal complications such as infection. Prophylactic antibiotic treatment should be started before signs of infection manifest. Cultures from skin and mucosal erosions, blood, and sputum should be taken regularly and antibiotic therapy adjusted accordingly. Haematocrit, blood gases, fluid, electrolytes and protein levels should be closely monitored. Particular attention must be directed toward pulmonary care and ophthalmologic preventive measures (ocular lubricants, sweeping of conjunctival fornices, and removal of fresh adhesions). Debridement of necrotic skin should not be performed before disease activity ceases. Alternative systemic treatments are still experimental and include haemodialysis, plasmapheresis, and cyclosporin therapy. Topical treatment may include hydrocolloid or gauze dressings. The clinical condition may require from 3 to 6 weeks to resolve, depending on the extent and severity of the lesions. References 1. Berthold R, Osvaldo C. Risk of Stevens Johnson syndrome and toxic epidermal necrolysis during first weeks of antiepileptic therapy: a case control study. Lancet 1999;353:2190-2194.
|
Home | Current Issue | Back Issues | Congress Calendar
Free Subscription | Editorial Board |