The phenytoin (hydantoin) hypersensitivity syndrome is rare but potentially fatal. Often, as in this case, it presents with non-specific symptoms and signs, requiring a high degree of clinical suspicion for diagnosis.
Statistics from Altmetric.com
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.
Phenytoin (diphenylhydantoin) is an aromatic ring compound which is metabolised in the liver by hydroxylation of one of its phenyl groups and then excreted in the urine and bile as a glucuronide conjugate. It was first introduced as an anticonvulsant in 1938. Dose-dependent side-effects (neurological impairment, gingival hyperplasia and megaloblastic anaemia) soon became apparent but an idiosyncratic hypersensitivity syndrome to phenytoin (and other hydantoins) was first described in 1959.1
We report a case of unrecognised, fatal, phenytoin hypersensitivity syndrome with characteristic pathological changes in the heart, liver, skin and lymph nodes at autopsy.
The patient was an 85-year-old AfroCaribbean woman who was admitted following collapse at home, with drowsiness, fever, rigors and a pruritic rash. Her medical history included a hysterectomy, ischaemic heart disease, type II diabetes mellitus and paraphrenia. One month prior to admission she was diagnosed as having complex partial seizures and was commenced on phenytoin 300 mg at night. Her other medications included metformin 500 mg bid, tolbutamide 500 mg bid, nifedipine 10 mg bid and risperidone 1 mg od. She was a non-smoker and had stopped taking alcohol since the onset of the fits.
On examination she was unwell and confused with a temperature of 38°C, regular pulse rate of 90 beats/min and a blood pressure of 120/70 mmHg. She had a generalised maculopapular erythaematous rash, marked neck stiffness but no photophobia or focal neurological signs. She had a systolic murmur at the apex. Her chest auscultation was normal. Abdominal examination was unremarkable.
The initial working differential diagnoses were septicaemia, a drug reaction, or possible food allergy (the patient was allergic to salmon and she had consumed some one day prior to admission).
The results of initial investigations are presented in box FB1, the most notable abnormal findings were a raised blood eosinophil count and C-reactive protein; a markedly elevated creatine kinase; and grossly deranged liver function tests. Skin biopsy was not performed.
She was treated empirically with intravenous antibiotics, corticosteroids and antihistamines. All her usual medication apart from the phenytoin was discontinued over a period of 3 days. Phenytoin was completely stopped on day seven and benzodiazepines were used to control any seizures. Unfortunately she developed renal, hepatic and cardiac failures and generalised skin desquamation. She deteriorated rapidly and died 12 days after admission.
At post-mortem examination there was florid desquamation of the entire skin surface. Internal examination revealed heavy lungs (right 520 g, left 420 g) , a large, yellow liver (1640 g), a slightly enlarged spleen (140 g) and widespread soft, fleshy lymphadenopathy (up to 3 cm in diameter). The heart (270 g) appeared normal. The brain (1000 g) was small and firm and showed yellowish discolouration of the cortex suggestive of amyloidosis.
Histological examination of the heart showed an interstitial eosinophil-rich infiltrate with myocytolysis. There was no necrosis, vasculitis or fibrosis (figure). These features are characteristic of hypersensitivity myocarditis.2 The liver displayed a marked periportal acute and chronic inflammatory infiltrate including increased numbers of eosinophils, with lobular hepatocyte necrosis, consistent with a severe drug-induced hepatitis. The lymph nodes sampled showed loss of the normal follicular architecture with expansion of the paracortical areas. Immunoblasts were not increased in number and these features are consistent with phenytoin-induced pseudolymphoma.3 Biopsy of the skin showed a picture of toxic epidermal necrolysis (full thickness epidermal necrosis, separation at the dermo-epidermal junction and a sparse superficial dermal inflammatory infiltrate), also consistent with the phenytoin hypersensitivity syndrome. The sections of lung revealed pulmonary oedema only, with no evidence of infection. The brain displayed classical features of Alzheimer's disease with amyloid angiopathy, extracellular amyloid plaques, neurofibrillary tangles and Hirano bodies.
Phenytoin hypersensitivity syndrome is more common in blacks4 5 and typically has its onset 3 weeks to 3 months after initiation of therapy. Although variable in presentation, its hallmark clinical features are fever, a rash (erythroderma, progressing via a generalised maculopapular to a pustular rash and finally desquamation), lymphadenopathy and hepatosplenomegaly.3-6 Peripheral leucocytosis and eosinophilia are common. Potentially fatal complications associated with the syndrome include hypersensitivity myocarditis,2drug-induced hepatitis with or without hepatocyte necrosis (which carries a mortality of up to 50%),2 7 and compromise of renal and pulmonary functions.4
As with any drug hypersensitivity reaction, prompt cessation of phenytoin is the first essential step in management.8 A benzodiazepine should be substituted for seizure control,9as was done with our patient. However, despite termination of phenytoin, the hypersensitivity reaction is often progressive and it is hence common practise to administer corticosteroids.8 The role of systemic corticosteroids in phenytoin hypersensitivity has not been studied in a randomised, placebo-controlled trial and its use is based only on anecdotal evidence.8 9
The underlying mechanisms of the phenytoin hypersensitivity syndrome are unknown9; immune complex,10 and delayed-type hypersensitivity allergic reactions,7alteration of lymphocyte function and toxic metabolite production4 8 have been postulated. Genetic factors are also thought to contribute, since siblings are said to have a one in four risk of showing a similar reaction.5 8 9Cross-reactivity with other aromatic ring anticonvulsants (eg, phenobarbital and carbamazepine) has been reported in patients with a history of phenytoin hypersensitivity.4 5 8 9
We wish to acknowledge the assistance of Dr S Al-Sarraj, Consultant Neuropathologist, Institute of Psychiatry, London.