Anticonvulsant hypersensitivity syndrome with fatal outcome

ARTICLE

Anticonvulsant hypersensitivity syndrome is a disease characterized by multisystemic involvement, fever, lymphadenopathy, mucocutaneous rash, hypertransaminasemia and peripheral eosinophilia. This rare syndrome seems to be related to arene oxide metabolites of aromatic anticonvulsants (phenytoin, phenobarbital and carbamazepine). Anticonvulsant hypersensitivity seems to be much more aggressive in patients undergoing concomitant radiotherapy. We report a case of anticonvulsant hypersensitivity syndrome developing toxic epidermal necrolysis with fatal outcome in a patient receiving cranial irradiation and aromatic anticonvulsants for seizure prophylaxis. This report attempts to emphasize the importance of an early diagnosis of this syndrome, the knowledge of the common cross-reactivity among the major anticonvulsants and the need for an appropriate measurement of the true benefits of seizure prophylaxis in patients with brain tumors.

(Key words: anticonvulsant hypersensitivity syndrome, cranial irradiation, toxic epidermal necrolysis, carbamazepine, phenytoin, phenobarbital.)

Anticonvulsant hypersensitivity syndrome (AHS) is a disease characterized by multisystemic involvement, fever, lymphadenopathy, mucocutaneous rash, hypertransaminasemia and peripheral eosinophilia. Nephritis, hematologic abnormalities and lung involvement can also be present [1-4].

This rare syndrome has been related to arene oxide metabolites of aromatic anticonvulsants (phenytoin [5], phenobarbital [3, 6] and carbamazepine [7, 8]) and with a new anticonvulsant drug, lamotrigine [9].

It seems to be an idiosyncratic reaction due to an abnormal detoxification of arene oxide metabolites by cytochrome P-450 [10, 11].

Cross-reactivity among the major anticonvulsants is common and should be considered before shifting to an alternative drug [2].

We report a case of anticonvulsant hypersensitivity syndrome with fatal outcome in a patient with a nasopharyngeal carcinoma undergoing cranial irradiation. Phenytoin plus phenobarbital substituted the initial treatment with carbamazepine for seizure prophylaxis after an episode of toxicoderma. Forty-five days later the patient developed a new generalized rash, which evolved in three days to a toxic epidermal necrolysis and death.

The benefits of anticonvulsant prophylaxis in patients with intracranial tumors are commented.

Case report

A 72-year-old white woman was diagnosed as having a nasopharyngeal carcinoma with intracraneal spreading. The disease was treated with palliative radiotherapy, deflazacort (60 mg three times daily) with progressive tapering and carbamazepine (200 mg three times daily) for seizure prophylaxis. Forty days later she was admitted in our Emergency Department presenting a pruritic centrifugal rash, which rapidly became widespread. She had completed radiotherapy and the steroid course five days before. Physical examination revealed a generalized maculopapular erythematous rash with occasional urticarial lesions. The rash involved the trunk and extremities with a negative Nikolsky sign. Mucous membranes were intact. Her vital signs were stable and she was afebrile.

Laboratory tests demonstrated a haemoglobin level of 11 g/dl and a leucocyte count of 16,640/mm³ with eosinophils at 5,030/mm³. Liver function tests were moderately elevated with an AST of 116 U/l (normal 15-37 U/l), and ALT 78 U/l (normal 9-42 U/L). The patient refused a skin biopsy.

On the basis of a clinical diagnosis of toxicoderma, carbamazepine therapy was stopped, and prednisone 60 mg daily, antihistamines and topical steroids were started. Within six days there was a notable improvement of the rash. A diagnostic challenge with oral carbamazepine was rejected due to her basal condition. Carbamazepine patch tests were negative.

Four months later, the patient presented again with a five-day history of skin rash and fever. Forty-five days before the onset of the eruption, the patient had begun treatment with phenytoin, phenobarbital and dexamethasone to prevent seizures from intracraneal involvement.

On physical examination, the patient was febrile at 38º C with conjunctival injection. The individual skin lesions were poorly defined macules with darker centers, coalescing into larger plaques that involved the trunk, extremities, palms and soles. Phenytoin and phenobarbital were discontinued. The lesions progressed despite treatment with 6-methilprednisolone (60 mg IV twice daily), and three days afterwards the patient developed an extensive denudation of the trunk that rapidly extended to the entire body (Figs. 1 and 2) with severe mucosal involvement. Nikolsky sign was positive throughout the body.

Laboratory tests demonstrated white blood count with leucocytosis (17,530/mm³), eosinophilia (2,350/mm³) and hypertransaminasemia (AST 78, ALT 92). A biopsy specimen showed full-thickness epidermal necrosis with subepidermal bulla formation and a sparse dermal mononuclear infiltrate. A diagnosis of toxic epidermal necrolysis was made (Fig. 3). A cyclophosphamide course was administered after a lack of response to intravenous steroids. An episode of sepsis developed with rapid deterioration of the patient's general condition, which eventually led to death.

Discussion

Several drugs may be involved in the pathogenesis of hypersensitivity reactions. Anticonvulsant hypersensitivity syndrome (AHS) lacks a typical reaction pattern; it has a low reaction rate (1: 10,000 exposures), a variable latency and is not dose dependent. Current pathogenetic theories suspect a genetic defect involving drug metabolism [2, 4, 12].

Aromatic anticonvulsants (phenytoin, phenobarbital and carbamazepine) are metabolized by cytochrome P-450 to hydroxilated aromatic compounds. In this process, reactive toxic aromatic metabolites (arene oxides) are developed. These metabolites are detoxificated by cell enzymes called epoxide hydrolases. In patients with AHS, it has been postulated that the existence of a defect in these enzymes might lead to the accumulation of toxic metabolites. Arene oxides working as haptens can be linked to cell macromolecules and initiate an immunologic cell mediated response [1, 2, 10, 13, 14]. Because of the possible cutaneous metabolization of these metabolites, it has been suggested that the cellular mediated skin tests (patch tests) could be useful in the diagnosis of this syndrome although results vary [1]. Patch tests were negative in our patient.

Patients treated with more than one anticonvulsant represent an important diagnostic issue for several reasons. First, the latency of the symptoms and the non-specific clinical features hinder the identification of the responsible drug. Moreover, challenge tests can trigger severe reactions in the most severely affected patients.

AHS seems to be the clinical expression of a variety of enzymatic defects, which would concern the metabolic pathways of the major anticonvulsants. Depending on the metabolic level disrupted, these patients may show cross reactivity to all aromatic anticonvulsants or to only one or two [1, 2]. For this reason, patients presenting with AHS should be treated with non-aromatic anticonvulsants (valproic acid, felbamate or lamotrigine) [1, 2, 12].

The latencies reported for the onset of AHS range from 2 weeks to 3 months, and in cases of readministration, the interval can be a matter of hours [1, 12].

The clinical features include fever (94%), malaise, skin rash (87%) and generalized lymphadenopathy (75%). Skin rashes range from mild morbiliform eruptions to erythema multiforme or toxic epidermal necrolysis [1-3]. Signs of internal organ involvement, when present, are usually late and include jaundice, hepatitis, nephritis, pneumonitis or hematologic alterations. They are all manifestations of an organ-specific granulomatous vasculitis [12].

Laboratory studies find leucocytosis with eosinophilia (30%), hepatitis (51%), nephritis (11%) or humoral immunity alterations [2]. The most common skin manifestation is a morbiliform exanthema that resolves spontaneously. Occasionally, it can become an exfoliative dermatitis, Stevens-Johnson syndrome, or toxic epidermal necrolysis.

We believe that in our patient, the preexisting sensitization to carbamazepine was a determining factor in the severity of the second skin rash (epidermal necrolysis). This same reason could explain the presence of peripherical eosinophilia, otherwise rare in patients affected by Lyell syndrome [15]. So we believe that this unusual overlap (TEN with eosinophilia) appeared because Lyell syndrome developed in the context of a pre-existing hypersensitivity syndrome.

Evolution cannot be anticipated because despite withdrawal of the anticonvulsant at an early stage, clinical features can worsen and some previously uninvolved organs can be affected [2, 4].

Severe toxicodermas have more often been reported in immunosuppressed patients. When patients are receiving cranial irradiation and anticonvulsant drugs for seizure prophylaxis, the immunosupression [12] seems to be mainly due to three reasons:

- A decrease in the CD8⁺ lymphocyte population induced by radiotherapy. This allows an easier development of hypersensitivity reactions to drugs.

- The high doses of corticosteroids administered for cerebral edema prophylaxis [16-18].

- The immunosuppressor tumoral effect [19].

To the best of our knowledge, there are 22 cases reported of erythema multiforme, Stevens-Johnson syndrome or toxic epidermal necrolysis arising in patients receiving radiation therapy, plus aromatic anticonvulsants [16, 20, 24]. As in our experience, the combination of radiotherapy, anticonvulsants, and sometimes concomitant tapering of steroids, triggered a cutaneous eruption [23]. The most frequently reported association is whole brain irradiation and phenytoin [16-18, 20], although carbamazepine [23] and phenobarbital [22] plus radiotherapy have also been reported.

Redondo [12] postulated that steroids and radiotherapy might lead to a reduction in the glutation levels. Glutation is the main endogenous antioxidant and its decrease could be responsible for an increased accumulation of the arene oxide metabolites.

It should always be kept in mind that only 20% of the patients suffering from intracraneal metastases show seizure disorders [24]. For this reason, the potential benefits obtained with anticonvulsant prophylaxis in patients with intracraneal tumors should always be carefully considered, particularly in those who have been treated with whole brain irradiation [18, 22]. Several authors [12] advocate anticonvulsant prophylaxis only in patients with metastatic melanoma due to the higher risk of seizures. In these cases, valproic acid should be considered the main option [12, 22].

In conclusion, this report attempts to emphasize the importance of an early diagnosis of this syndrome, the knowledge of the common cross-reactivity among the major anticonvulsants and the need for an appropriate measurement of the true benefits of seizure prophylaxis in patients with brain tumors.

References

1. Morkunas AR, Miller MB. Anticonvulsant hypersensitivity syndrome. Critical Care Clinics 1997; 13: 727-99.

2. Shear NH, Spielberg SP. Anticonvulsant hypersensitivity syndrome. In vitro assessment of risk. J Clin Invest 1988; 82: 1826-32.

3. Braitberg G, Miller MB, Curry SC. Anticonvulsant hypersensitivity syndrome. Emerg Med (Australia) 1995; 7: 170.

4. Vittorio CC, Muglio JJ. Anticonvulsant hypersensitivity syndrome. Arch Int Med 1995; 155: 2285-90.

5. Chaiken BH, Goldberg BI, Segal JP. Dilantin hypersensitivity: report of a case of hepatitis with jaundice, pyrexia, and exfoliative dermatitis. N England J Med 1950; 242: 897-8.

6. McGeachy TE, Bloomer WE. The phenobarbital hypersensitivity syndrome. Am J Med 1953; 14: 600-4.

7. Crill WE. Carbamazepine. Ann Intern Med 1973; 79: 844-7.

8. Breathnach SM, McGibbon DH, Ive FA, Black MM. Carbamazepine (Tegretol) and toxic epidermal necrolysis: report of three cases with histopathologic observations. Clin Exp Dermatol 1982; 7: 585-91.

9. Jones D, Chiap V, Resor S, Appel G, Grossman ME. Phenytoin-like hypersensitivity associated with lamotrigine. J Am Dermatol 1997; 36: 1016-8.

10. Enrique E, Simó M, García-Ruiz G, Caig P, San Miguel MM, García-Ortega P. Síndrome de hipersensibilidad a anticonvulsivantes. Rev Clin Esp 1999; 199: 257-8 (letter).

11. Lertratanangkoon K, Horning MG. Metabolism of carbamazepine. Drug Metab Dispos 1982; 10: 1-10.

12. Redondo P. Reacciones cutáneas adversas a drogas en Neurología. Rev Neurol 1997; 25: 309-19.

13. Strickler SM, Dansky LV, Miller MA, Seni MH, Andermann E, Spielberg SP. Genetic predisposition to phenytoin-induced birth defects. Lancet 1985; ii: 746-9.

14. Mauri-Hellweg D, Bettens F, Mauri D, Brander C, Hunziker Tpichler WJ. Activation of drug-specific CD4⁺ and CD8⁺ T cells in individuals allergic to sulfonamides, phenytoin and carbamazepine. J Immunol 1995; 155: 462-72.

15. Roujeau JC, Chusidou O, Saiag P, Guillaume JC. Toxic epidermal necrolysis (Lyell syndrome). J Am Acad Dermatol 1990; 23: 1039-58.

16. Delattre JY, Safai b, Posner JB. Erythema multiforme and Stevens-Johnson syndrome in patients receiving cranial irradiation and phenytoin. Neurology 1988; 38: 194-8.

17. Janinis J, Panagos G, Panousaki A. Stevens-Johnson syndrome and epidermal necrolysis after administration of sodium phenytoin with cranial irradiation. Eur J Cancer 1993; 29A: 478-9.

18. Cockey GH, Amann ST, Reents SB, Lynch JW. Stevens-Johnson syndrome resulting for whole-brain radiation and phenytoin. Am J Clin Oncol 1996; 19: 32-4.

19. Dröge W, Eck HP, Betzler M. Plasma glutamate concentration and lymphocyte activity. J Cancer Res Clin Oncol 1988; 114: 124-8.

20. Borg M, Probert J, Zwi L. Is phenytoin contraindicated in patients receiving cranial irradiation? Australas Radiol 1995; 39: 42-6.

21. Hoang-Xuan K, Delattre J, Poisson M. Stevens-Johnson syndrome in a patient receiving cranial irradiation and carbamazepine. Neurology 1990. 40: 1144-5.

22. Duncan KO, Tigelaar RE, Bolognia JL. Stevens-Johnson syndrome limited to multiple sites of radiation therapy in a patient receiving phenobarbital. J Am Acad Dermatol 1999; 40: 493-6.

23. Chevenet C, Ferrier M, Reverte M, Tao Y, Perea R, Roger H, et al. Érythème polymorphe dû à l´association carbamazépine et irradiation cérébrale. Ann Dermatol Venereol 1992; 119: 929-31.

24. Posner JB. Brain metastases: a clinician´s view. In: Gilbert HA, Posner JB, Weiss L, eds. Brain metastasis. Boston: GK Hall; 1980: 2-29.

Article accepted on 16/7/02