Anticonvulsant hypersensitivity syndrome: a review

Review

10.1517/14740338.4.3.571 © 2005 Ashley Publications Ltd ISSN 1474-0338 571

Ashley Publicationswww.ashley-pub.com

General

Anticonvulsant hypersensitivity syndrome: a reviewNithya J Gogtay†, Sandeep B Bavdekar & Nilima A Kshirsagar†Seth GS Medical College & KEM Hospital, Department of Clinical Pharmacology, Parel, Mumbai 400 012, India

Anticonvulsant hypersensitivity syndrome (AHS), characterised by fever,rash and internal organ involvement, is a rare, but potentially fatal adverseevent that occurs most commonly with first-line aromatic anticonvulsants,but can also occur with non-aromatic anticonvulsants such as lamotrigineand valproic acid. AHS can begin anywhere from 1 to 12 weeks after com-mencement of therapy and has been estimated to occur at a frequency of1/1000 to 1/ 10,000 exposures. Its true incidence, however, remainsunknown due to under-reporting. The disease has protean manifestationsmimicking several other conditions, and the diagnosis is thus difficult. Sev-eral hypotheses have been put forward to explain the pathogenesis ofAHS. These include accumulation of toxic metabolites, graft versus host dis-ease, antibody production and viral infections. The one based on toxicmetabolites has found the greatest acceptance, perhaps due to the factthat it can be proven by an in vitro test; the lymphocyte toxicity assay. Dis-continuation of the offending agent with supportive, symptomatic therapyforms the mainstay of management of AHS. In addition, counselling ofboth the patient and first degree relatives for susceptibility to AHS is animportant aspect of management. In the last decade, several new anticon-vulsants have been introduced for epilepsy. In addition, for resource-poorcountries, inexpensive and effective first-line drugs such as phenytoin andphenobarbitone will continue to remain important treatment options.Thus, the problem of AHS will continue, and attempts should be made tofurther understand the molecular basis of and individual susceptibility toAHS. Adverse event monitoring programs must also actively seek AHSreports to estimate its true incidence.

Keywords: anticonvulsant hypersensitivity syndrome (AHS), aromatic anticonvulsant, lymphocyte toxicity assay (LTA)

Expert Opin. Drug Saf. (2005) 4(3):571-581

1. Introduction

Anticonvulsant hypersensitivity syndrome (AHS), characterised by fever, rash andinternal organ involvement, is a rare, but potentially life-threatening adverse eventthat occurs following exposure to an anticonvulsant, most commonly to aromaticanticonvulsants (phenytoin, phenobarbitone, primidone, carbamazepine) or lamot-rigine. Although AHS was described several decades ago, it remains a challenge toclinicians to diagnose it early and to scientists as its pathophysiology and molecularbasis remain to be completely elucidated. Lack of awareness amongst physicians, theability of the reaction to mimic several disorders, non-availability of a reliable diag-nostic test and an incomplete understanding of its pathogenesis make it prone tounder-reporting.

1. Introduction

2. Anticonvulsant hypersensitivity

syndrome

3. Future perspectives

4. Expert opinion

For reprint orders, please contact:[email protected]

Anticonvulsant hypersensitivity syndrome: a review

572 Expert Opin. Drug Saf. (2005) 4(3)

2. Anticonvulsant hypersensitivity syndrome

2.1 EpidemiologyThe early reports of hypersensitivity to antiepileptic drugsappeared in the 1930s as events secondary to administrationof phenytoin [1]. The systemic symptoms of AHS weredescribed in the 1950s [2]. Although initial reports termedthe reaction as ‘dilantin hypersensitivity syndrome’ and ‘nir-vanol sickness’, the term anticonvulsant hypersensitivity syn-drome was coined when the characteristic constellation ofmanifestations was reported in association with phenobarbi-tone and carbmazepine, as well [3,4]. Over the years, the syn-drome has also been described under various namesincluding ‘dilantin hypersensitivity reaction’, ‘phenytoin syn-drome’, ‘dilantin syndrome’, ‘phenobarbital hypersensitivitysyndrome’, ‘Kawasaki-like syndrome’, ‘drug rash with eosi-nophilia and systemic symptoms’ ‘mononucleosis-like syn-drome’, amongst others. [2]. Other drugs that have beenassociated with similar hypersensitivity syndrome includesulfonamides, dapsone, allopurinol, sorbinil, minocyclineand terbinafine [5-10].

The reaction has been estimated to occur at a frequency of1/1000 to 1/10,000 exposures [2,3]. However, its true incidenceremains unknown due to under-reporting. The study carriedout by Tennis and Stern, provides some idea about the degreeof under-reporting. Using a record linkage study, they esti-mated that the risk of developing AHS within just 60 days ofthe first or second prescription in new users of only phenytoinor carbamazepine was up to 4.5 and 4.1/10,000, respectively[5]. Most of the literature available on AHS is in the form ofcase reports and case series. In these reports, there seems to beno gender variation in the risk of development of AHS.Although the available literature seems to suggest a higher inci-dence of the syndrome amongst the black populations [11], thiscould be related to a higher incidence of epilepsy in the thesegroups rather than a higher propensity to develop the syn-drome [2,11]. Aguir et al. have suggested cranial irradiation as arisk factor for the development of AHS [12], but this hypothesisremains to be substantiated. Children may be at a higher riskof development of AHS as the incidence of seizure disorders ishigher during the first decade of life [1]. AHS can have a fataloutcome and children < 2 years of age, individuals on multipleanticonvulsants, persons with concomitant illnesses and theelderly seem to have a potentially higher mortality rate [13,14].

2.2 PathogenesisThree notions have been put forward to explain the genesis ofAHS. It is believed to be an allergic reaction because it is unre-lated to dosage or serum concentrations [15]. In addition, sub-sequent exposures hasten the appearance of the syndrome, acharacteristic known to be associated with allergic reactions.The allergic reaction is believed to occur through accumula-tion of toxic metabolites. Another hypothesis points towardsthe similarity between the manifestations of AHS and those ofgraft-versus-host disease, whereas yet another hypothesis

delineates a role for viral infections. The one based on thegeneration of toxic metabolites is widely accepted, perhapsbecause it can be confirmed using an in vitro test [16-18].

2.2.1 Accumulation of toxic metabolitesThe first generation anticonvulsants; phenytoin, phenobarbi-tone and carbamazepine, all contain an aromatic benzenering that is metabolised via the cytochrome P450 enzymes(CYP450) to arene oxides. These arene oxides are highly elec-trophilic compounds that can covalently bind to macromole-cules to produce cytotoxicity or form neoantigens that triggeran immunological (hypersensitivity) response. Under normalcircumstances, these unstable metabolites are detoxified bymicrosomal epoxide hydrolase (non-CYP450 enzyme) andget converted to a dihydrodiol. Minor routes of metabolismalso include binding to glutathione and spontaneous rear-rangement to form a phenol, which then reacts non-enzy-matically with glutathione. A deficiency in the activity ofepoxide hydrolase results in accumulation of arene oxides,which is responsible for the tissue damage and clinical mani-festations observed in AHS [19-22]. For a long time, it wasbelieved that quantitative deficiency of epoxide hydrolase wasresponsible for the syndrome. An isolated case report hasraised a possibility of a qualitative defect in the enzyme beingresponsible for AHS [23]. The lymphocyte toxicity assay(LTA) that is used to confirm the diagnosis of AHS is basedon this hypothesis. Some work has been done regardingdelineation of the immune mechanism. Leeder et al. haveidentified antibodies to CYP450 3A1 in patients with AHS.This suggests that arene oxides covalently bind withCYP450, modify it to form a neoantigen and trigger animmune response, which leads to antibody production [24].

The deficient activity of the enzyme may also account forthe teratogenicity of the first-line anticonvulsants. The LTA,(see Section 2.5.1) has confirmed higher rates of cell death inchildren with clinical evidence of the fetal hydantoin syn-drome [25]. It is, therefore, possible for children with evidenceof this syndrome to suffer from hypersensitivity reactions tophenytoin, phenobarbitone and carbamazepine.

In addition, the diversity of clinical manifestations of AHS(see Section 2.3) may relate to the fact that the various organsinvolved contain CYP450 and epoxide hydrolase. Liver hasthe maximum amount of CYP450 activity and hence it is notsurprising to note that liver is the most commonly affectedinternal organ [26]. Despite a lack of natural epoxide hydrolaseactivity, the thyroid and the bone marrow are affected. A pos-sible explanation could be the deficient activity of thyroid per-oxidase and myeloperoxidase, respectively, leading to excessiveaccumulation of arene oxides in these organs [27].

With this theory, occurrence of AHS in association witharomatic anticonvulsants such as phenytoin, phenobarbitoneand carbamazepine can be easily understood. It is also logicalto explain the occurrence of the syndrome in association withprimidone and oxcarbazepine, as primidone is metabolicallyconverted to phenobarbitone in the body and as

Gogtay, Bavdekar & Kshirsagar

Expert Opin. Drug Saf. (2005) 4(3) 573

oxcarbazepine is structurally similar to carbamazepine. Whatthe theory does not explain, is the occurrence of AHS in asso-ciation with lamotrigine [28], which has a non-aromatic struc-ture. Lamotrigine primarily undergoes conjugation, and onlyminor fractions undergo Phase I metabolism. It is not knownif one or more of the Phase I metabolism products is responsi-ble for AHS. However, a recent report has demonstrated thatlamotrigine, when metabolised in rat hepatocyte culture, gaverise to arene oxide metabolites [29]. Although this has not beendemonstrated to occur in humans, it opens an interesting areafor research. A single case report [30] has hinted at the possibil-ity of valproic acid being responsible for AHS. As this drugalso primarily undergoes conjugation with glucuronic acid,the toxic metabolite theory may not be able to explain thisphenomenon.

2.2.2 Viral InfectionsThe contribution of viral infections has been described by sev-eral case reports. Cytomegalovirus (CMV)-specific IgG anti-bodies along with CMV DNA were shown to be significantlyraised in patients with AHS secondary to phenytoin therapy,suggesting that reactivation of CMV may play a role in AHS.Association between AHS and human herpes virus 6 (HHV6)reactivation has also been described. In a study of 10 patientswith AHS, Kano et al. [18] showed a greater than four-foldincrease in HHV6 IgG titers along with decreased total IgGlevels and reduced β-cell counts. Hence, they postulated thatHHV6 reactivation is associated with AHS. Another virusthat has been implicated is HIV. It is a known fact thatpatients with HIV infection have a higher rate of adverse reac-tions to many medications [31]. Beller et al. have reported aprolonged AHS, probably related to lamotrigine, in an HIVinfected female [32].

2.2.3 Graft-versus-host diseaseSimilarities between the manifestations of AHS and ‘graft-ver-sus-host disease (GVHD)’ (initial exposure, an inductionperiod, occurrence after rechallenge, reaction being unrelatedto dosage or serum concentrations) have stimulated scientiststo suggest the idea that AHS is a form of GVHD. It is postu-lated that the anticonvulsants alter the host lymphocytes insuch a manner that they recognise host tissues as foreign.

2.2.4 Antibody productionGiven its highly electrophilic nature, arene oxide can beexpected to form neoantigens by combining with various pro-teins. This neoantigen formation could then trigger an anti-body response. The antibody formation has been blamed forcausing AHS. In fact, patients suffering from AHS have beenshown to possess an antibody that reacted with rat CYP4503A1 isoenzyme. Some investigators have also demonstratedhuman antibodies directed against protein in the endoplasmicreticulum. This target protein has been shown to have ahigher molecular weight than expected from CYP450 alone,raising the possibility of neoantigen formation secondary to

covalent binding of the arene oxide metabolite with theCYP450 enzyme [33,34].

2.3 Clinical features and laboratory findingsAHS usually occurs after first exposure to one of the incrimi-nating drugs and commonly begins within 3 weeks afterinitiation of an anticonvulsant, although several literaturereports suggest that the syndrome can begin 1 – 12 weeksafter commencement of therapy [19]. A case-control study byRzany et al. which looked at the risk of developing Ste-vens-Johnson syndrome (SJS) and toxic epidermal necrolysis(TEN) during the first weeks of antiepileptic therapy showedthat the risk was highest in the first eight weeks after initiationof antiepileptic therapy, regardless of the individual drug used[35]. In previously sensitised individuals, the symptoms maybegin within a day of re-exposure. In most cases, the classicaltriad of fever, rash and lymphadenopathy accompanied bysingle- or multi-organ abnormalities are seen [2,3,19,36,37]. Thefirst symptom to occur is usually fever (high and spiking) andis accompanied by malaise and pharyngitis [3]. Subsequently, arash that is variable in severity emerges and can range fromsimple exanthem to SJS and erythema multiforme. Skin rash,which is seen in up to 90% of cases, usually appears withindays to two weeks of the drug exposure. It begins initially as amacular erythema that begins on the face, upper extremitiesand upper trunk and then goes onto involve the lower extrem-ities. Various skin lesions have been known to occur (Table 1).The organs involved include the liver, kidney, bone marrow,lung, heart and the CNS. The liver is the most commonlyaffected internal organ [37]. The protean clinical manifesta-tions of organ derangements are summarised in Table 1. Lym-phadenopathy both as an isolated finding and in conjunctionwith other clinical features is a well-recognised manifestationin AHS associated with phenytoin use and, to a lesser extent,with that associated with carbamazepine [40-43]. Very rarely,lymphoma following antiepileptic drug (AED) therapy hasbeen reported, particularly if the AED is continued in spite ofdevelopment of lymphadenopathy. All the manifestationsenumerated here and in Table 1 may not be present at thesame time. They occur sequentially and only a couple of man-ifestations might be seen at the time of presentation. Hence, ahigh index of suspicion is required to make an early diagnosisand take appropriate steps for management.

Many patients have milder affection and the syndrome canbe diagnosed only on the basis of abnormalities detected onlaboratory tests. A peripheral blood smear shows eosinophilawith lymphocytosis and atypical lymphocytes at times. Aplas-tic anaemia is also a rare finding. The severity of hepatic man-ifestations is dependent upon the alacrity with which theincriminating drug is withdrawn. Severe cases of hepatitis canbe life threatening [1,2,38]. The clinical course is variable. Ini-tially the patients appear toxic. The course is less eventful andinternal organ involvement is milder if the incriminated drugis withdrawn early. In such an event, the clinical manifesta-tions resolve over a few days and rash disappears with mild



desquamation [37]. If the drug withdrawal is delayed due todelay in diagnosis, the manifestations are usually more severethan usual and there is an increase in the risk of death. Mostpatients demonstrate gradual amelioration of manifestationsas soon as the drug is withdrawn. However, some patientsmay show progressive worsening or involvement of new organsystems even after the culprit drug is withdrawn. Somepatients experience a relapse, when corticosteroids are with-drawn [44]. AHS is associated with a mortality rate rangingfrom 5 to 50% and fatal outcomes are most often associatedwith liver dysfunction and TEN [3,37,45]. Although mostpatients would recover over a period of time, some of themcould have long-term manifestations and sequels such asdelayed onset transient hypothyroidism, skin changes andocular complications (Table 2).

2.3.1 Pathological featuresIn AHS, dermatological lesions commonly show perivascularinfiltrates, whereas liver biopsy reveals peri-portal

inflammation. This is associated with varying degrees ofnecrosis, which could be absent in a subset of cases. Renalbiopsies have demonstrated the presence of interstitialnephritis and necrotising vasculitis that is often described asgranulomatous necrotising angiitis [27]. Clinically, theinvolvement of skin is striking, and to the uninitiated, theskin lesions in AHS could mimic those seen with skin lesionsassociated with a variety of conditions such as TEN, SJS andKawasaki disease (KD). Microscopic examination of skinbiopsy specimen offers a reasonable tool for diagnosing theseconditions. TEN is characterised by the presence of a blisterat the epidermo-dermal junction and sparse lympho-histio-cytic infiltrate is present around the dilated blood vessels ofthe superficial vascular plexus. SJS, in contrast, is associatedwith the presence of a dense infiltrate of lymphocytes andhistiocytes around the blood vessels of the superficial dermis,oedema of the papillary dermis and epidermal and subepider-mal vesiculation. The histological examination of skin biopsyin KD reveals sparse perivascular infiltrate of lymphocytesand histiocytes, marked oedema of the papillary dermis withpustular variant showing sterile intra-epidermal spongiformpustules with neutrophilic infiltration [47,48].

2.4 Differential diagnosisGiven its protean manifestations, it is not surprising that sev-eral disorders mimic AHS [2]. Although the clinician couldundertake appropriate studies for confirmation of the diag-nosis, withdrawal of the culprit anticonvulsant drug shouldbe undertaken without delay. Epidemiologically, acute viralinfections such as Epstein-Barr virus, cytomegalovirus, hepa-titis virus, HIV and influenza virus infections and streptococ-cal infections could mimic AHS. These could be detected onthe basis of appropriate viral studies. Collagen vascular dis-eases also have multisystem/multiorgan involvement similarto that found in AHS. Tests for the detection of antinuclear

Table 1. Clinical manifestations of anticonvulsant hypersensitivity syndrome.

Organ/system involved Clinical manifestations

Skin and mucus membranes [1,2,38] Exanthematous eruption, Stevens-Johnson syndrome, toxic epidermal necrolysis, macular erythematous pruritic rash, papules, pustules, periorbital or facial oedema, desquamation of skin

Liver [1,2] Vomiting, tender hepatomegaly, jaundice and clinical evidence of hepatic dysfunction, granulmatous hepatitis or fulminant hepatic necrosis

Kidney [1,2] Haematuria, oliguria, interstitial nephritis, acute renal failure

CNS [2] Aseptic meningitis, status epilepticus following drug withdrawal

Heart [2] Pericarditis, carditis, congestive cardiac failure

Lungs [2] Cough, pneumonitis, adult respiratory distress syndrome, respiratory failure

Musculoskeletal system [2,3] Myalgia, myositis, myopathy, arthralgia, rhabdomyolysis

Other [1,2,37,39] Fever, flu-like symptoms, pharyngitis, exudative tonsillitis, malaise, fatigue, headache, difficulty in swallowing, lymphadenopathy, splenomegaly, decreased appetite, pancreatitis, thyroiditis, delayed onset hypothyroidism, colitis, syndrome of inappropriate secretion of antidiuretic hormone, serositis, uveitis

Table 2. Long-term sequelae.

Organ/system involved

Long-term manifestations/sequel

Skin and mucus membranes [46]

Oral and pharyngeal lesions take up to 10 weeks to healResolution of skin hyperpigmentation may take 12 weeks to 1 yearAlopecia

Eye [46] Symblepharon, corneal ulceration and scarring leading to blindness, SICCA-like syndrome

Lungs Bronchiolitis obliterans in patients with Stevens-Johnson syndrome

Other Delayed onset hypothyroidism



antibodies and antibodies against double-stranded DNAcould help clear the issue. Presence of puffiness of face andanasarca bring up the possibility of serum sickness. However,the propensity of serum sickness to be associated with urti-caria and arthralgia and absence of organ involvement couldhelp clarify the matters. Generalised lymphadenopathy andfever noted in AHS might raise doubts about the possibilityof lymphoma. KD with manifestations such as fever, lym-phadenopathy and skin and mucus membrane lesions isanother disease that closely mimics AHS [49]. Akin to AHS,KD does not have a diagnostic laboratory test and diagnosingit early is of utmost importance if one has to prevent thesequelae in the form of coronary aneurysms. Interestingly,intravenous immunoglobulin (IVIG) therapy is used for theprevention of this complication in KD and is also used inpatients with severe manifestations of AHS.

2.5 Laboratory testsAHS is associated with multi-organ damage. Laboratory testsundertaken to detect the presence of, and evaluate the extentof, organ damage are listed in Table 3. Depending upon thenature of presentation, additional tests may have to be under-taken to confirm or rule out alternative diagnoses. Given thenumber of disorders considered in the differential diagnosis,listing such tests is outside the scope of this review. The reviewwould concentrate on two tests used for diagnosis of AHS,per se.

2.5.1 The lymphocyte toxicity assayThe LTA was developed in the 1980s as an in vitro tool for thediagnosis of AHS [19]. The test is based on the premise thatgeneration of metabolites that are inherently more reactivethan the parent drug are responsible for the development ofdrug hypersensitivity reactions. As stated earlier, aromaticanticonvulsant drugs are metabolised by CYP450 to areneoxides. In the normal course, these are further metabolised byepoxide hydrolase. It has been postulated that individuals

with deficient activity of epoxide hydrolase develop AHS dueto accumulation of arene oxides. These arene oxides are toxicand are responsible for the manifestations of AHS. One of theactions of these reactive metabolites is to induce earlylymphocyte death. LTA makes use of these aspects to diagnoseAHS [19,21,22].

In the assay, lymphocytes from the patient are used as thesubstrate and hepatic microsomes from mice serve as thesource of CYP450. The lymphocytes are exposed to gradedconcentrations of the incriminated anticonvulsant drug in thesystem to simulate the in vivo situation. The proportion oflymphocyte death in vitro is compared with a control for thediagnosis of AHS. The lymphocyte viability is assessed byTrypan blue dye exclusion with live cells not taking up the dyeand the dead cells appearing as dark blue spots under themicroscope (Figure 1). A positive LTA is the one where there isincreased cell death in the subjects with AHS as compared tocontrols (Figure 2). The test has been modified to prove thatthe cell death is related to reduced activity of epoxide hydolaseby adding an inhibitor of epoxide hydrolase to the controlsample. In such a situation, control lymphocytes show celldeath not significantly different than that observed in thepatient, thereby proving the fact that the decreased activity ofthe enzyme was responsible for the increased cell death in thepatient. Yet another way of proving the same is to add pureexogenous epoxide hydrolase to patient lymphocytes. Thisattenuates the observed cell death in the patient sample [19].

Although the LTA does offer a diagnostic tool, it has twomajor limitations. It is difficult to perform the test duringthe acute phase because the lymphocyte yield is suboptimal(personal observations). This is due to the death of lym-phocytes in vivo as seen on a peripheral blood film. In addi-tion, there is an increased chance of eliciting a false-negativeresult during the acute phase [2]. Hence, the test is generallycarried at least two months after the reaction. Secondly, thetest is expensive and requires training, thereby restricting itsuse to research laboratories and limiting its accessibility to

Table 3. Laboratory investigations in anticonvulsant hypersensitivity syndrome.

Organ/system involved Results of laboratory investigations

Liver [2,44] Elevated levels of hepatic transaminases in the serum, hyperbilirubinaemia, coagulopathy

Kidney [2, 44] Hematuria, raised levels of blood urea and serum creatinine

CNS [2] Raised levels of proteins and pleocytosis in the cerebrospinal fluid,

Heart [2] Myocarditis

Lungs [2] Chest radiograph: consolidation, features of adult respiratory distress syndrome

Musculoskeletal system [2] Elevated creatinine kinase levels, rhabdomyolysis

Other [1,2,38] Haematological: Anaemia, leukopenia, pancytopenia, leukocytosis, lymphocytosis, eosinophilia, elevated erythrocyte sedimentation rate, atypical lymophocytosis on peripheral smear, Coomb’s negative haemolytic anaemia, hypoplasia on bone marrow aspiration,Thyroiditis: Hyperthyroid status during the acute phase that evolves into a hypothyroid state within two months of initiation of symptoms, presence of thyroid-stimulating autoantibodies and antimicrosomal antibodies



clinicians. The test can also be used to determine the sus-ceptibility of first-degree relatives (parents and siblings) toAHS [50].

Neuman et al. modified the existing LTA based on Trypanblue exclusion to one based on mitochondrial succinate dehy-drogenase activity, with its activity being measured spectro-photometrically. They used this assay to confirmhypersensitivity syndrome related to sulfonamides and aro-matic anticonvulsants and found it to be objective, faster andreproducible and at least as good as the Trypan blue-basedLTA [51].

2.5.2 Patch testingPatch testing is used as an adjunct for the diagnosis of AHS.Incriminated drugs in the concentration of 1 and 10% inpetroleum jelly are applied on the skin. The test is not with-out its limitations. Its utility is still considered debatable. Itshould not be carried out in the acute phase to avoid

occurrence of false-positive and false-negative results. The testresults should always be interpreted in conjunction withclinical judgment [52,53].

2.6 ManagementAs with any hypersensitivity reaction, discontinuation of theoffending anticonvulsant and supportive, symptomatic ther-apy forms the mainstay of management. Waiting for confir-mation of diagnosis or for ruling out alternate diagnosisshould be discouraged, as this could have serious undesirableconsequences for the patient in the form of exacerbation ofmanifestations, more severe liver damage and increased risk ofmortality. In view of the possibility of multi-organ involve-ment, laboratory investigations usually include a completeblood count, liver and renal function tests, baseline thyroidfunction tests, urinalysis, chest radiograph if respiratorysymptoms are present) and skin biopsy in case of a blisteringeruption.

An alternative anticonvulsant may be prescribed dependingupon the perceived risk of occurrence of seizures. The varioustherapeutic modalities employed are depicted in Table 4.Careful attention to providing adequate nutrition and main-tenance of fluid and electrolyte balance cannot be overempha-sised. Although, the role of corticosteroids has been debatedupon, most clinicians initiate treatment with corticosteroids ifthe symptoms are severe [2,3,44,46]. Although benefits of otherimmunomodulatory measures (plasmapheresis and intra-venous immunoglobulin) have not been evaluated by largecontrolled trials, it is logical to use these measures in patientswith life-threatening or organ-threatening disease [2]. Forpatients with AHS and status epilepticus, where traditionalaromatic anticonvulsants cannot be given, intravenousdiazepam has been used with success, by virtue of it beingstructurally dissimilar [54]. Adjunctive supportive therapy usu-ally consists of adequate hydration, H2-blockers and topicalcorticosteroids. As it is difficult to differentiate between KDand severe AHS, and as administration of IVIG early in thecourse of both the diseases is vital, Bessmertny and Pham rec-ommend that IVIG be used in severe cases where KD cannotbe ruled out [49]. Once the patients recover from the acute epi-sode, another structurally dissimilar anticonvulsant is usuallystarted. The options available are valproic acid, gabapentinand vigabatrin. Although, lamotrigine is not an aromatic anti-convulsant and is primarily metabolised by conjugation (glu-couronidation), its ability to cause AHS and itscross-sensitivity (in vitro) with other aromatic anticonvulsantshave been demonstrated [57].

2.7 CounsellingPatients should be informed about the nature of the reactionand why it occurs. They should also know that the trait isinherited and, therefore, their first-degree relatives are moreprone to develop this reaction than the general population.

Patients with AHS with a first-line aromatic anticonvulsantshould be advised to avoid all other first-line aromatic

Figure 1. Photomicrograph showing the lymphocytetoxicity assay using Trypan blue dye exclusion, with deadlymphocytes taking up the dye.



anticonvulsants. The patient should carry a card with him/herat all times. The card should indicate that the patient has suf-fered from AHS and should identify the incriminating drug.The card should also indicate the drugs and anticonvulsants(phenobarbitone, primidone, phenytoin, carbamazepine,primidone and oxcarbazepine) that should be avoided. This isnecessary as the awareness about cross-reactivity is not opti-mal and there is a risk that a physician may initiate treatmentwith another cross-reacting drug.

As stated above, the first-degree relatives may have a risk ofdeveloping AHS that approaches 1 in 4 as compared with thereported incidence of 1 in 1000 to 1 in 10,000 exposures inthe general population. Considering the fact that epilepsy itselfruns in families, it is prudent to counsel family members toundergo tests to determine their susceptibility to develop AHS.

3. Future perspectives

AHS is an enigma for scientists and a challenge for clinicians.The syndrome seems to have an immunological basis in sub-jects with deficient functioning of an enzyme system involvedin metabolism of aromatic anticonvulsants. However, severalaspects related to the pathophysiology such as role of bioacti-vation, detoxification, protein binding and the intracellularmechanisms remain to be elucidated completely. It is also notknown if intermediary metabolites are primarily responsiblefor allergic hypersensitivity or serve as cofactors.

A wide range of other issues remains to be elucidated [4].Clinicians face difficulties in the diagnosis of AHS. Its proteanmanifestations ensure that several disorders need to be ruled

out before a diagnosis can be made. Absence of an easilyavailable diagnostic test is an additional roadblock, when earlydiagnosis and prompt withdrawal of the culprit drug is ofgreat importance for improving prognosis and reducing therisk of severe reaction and death. Clinicians would greatlyappreciate the development of an easily available and simplediagnostic test for making a diagnosis. This would be of spe-cial relevance to paediatricians, as fever and rash are the mostfrequent manifestations of AHS. These are also the presentingfeatures of several viral infections that occur with great fre-quency in children. Hence, doctors tend to defer withdrawalof the drug until several other disorders, including viral infec-tions, are ruled out. This could have disastrous consequencesfor the patient. It is not uncommon to find reports, wheredoctors have substituted one aromatic anticonvulsant withanother one after the occurrence of AHS [1]. The healthcareproviders should be aware that considering the high rate ofcross-sensitivity amongst aromatic anticonvulsants (40 –80%) such a substitution would lead to patients continuing tosuffer for a longer duration.

The entire scope of LTA is yet to be defined. Most clini-cians and scientists would agree that the test can be used forretrospective diagnosis of AHS. The issue of using the LTAin every subject to be treated with an aromatic anticonvul-sant for his susceptibility to develop AHS is an issue thatawaits a clear answer. Given the reported low prevalence ofAHS, it does not seem justifiable. However, enhanced physi-cian awareness resulting in more cases of AHS being diag-nosed, increased numbers of laboratories providing facilitiesfor LTA and improved pharmacovigilance increasing the

0

10

20

30

40

50

60

0 8 12 16 20 24 32Concentration of drug (υg)

Control Phenytoin LamotriginePhenobarbitoneCarbamazepine

% C

ell (

lym

ph

ocy

te)

dea

th

Figure 2. Lymphocyte toxicity assay. Representative graph of a patient with AHS showing increased cell death with multipleanticonvulsants in comparison to control.Taken, with permission, from Bavdekar SB et al. Ann. Pharmacother. (2004) 38(10):1648-1650 [57].AHS: Anticonvulsant hypersensitivity syndrome.



number of reported cases could bring about a change in thisequation.

4. Expert opinion

• Since 1993, treatment options for epilepsy have widenedwith a range of newer antiepileptic drugs now becomingavailable. These include felbamate, gabapentin, lamotrig-ine, topiramate, tiagabine, levetiracetam, oxcarbazepineand zonisamide. Over and above this, the older first-lineagents such as phenytoin, phenobarbitone, car-bamazepine, ethosuximide, sodium valproate and primi-done continue to be widely used. Resource-poor countriesaccounting for four-fifths of the 50 million people withepilepsy carry a significant burden. Older anticonvulsants

such as phenytoin and phenobarbitone given once a dayin their lowest effective doses continue to remain first-lineinexpensive options for epilepsy management, both inprimary and tertiary referral centres [58]. In addition,phenytoin, phenobarbitone and carbamazepine are usedin conditions other than epilepsy such as pain syndromes,cardiac arrhythmias and as choleretic agents. HIV-infec-tion is a worldwide epidemic, with > 5 million people indeveloping countries being afflicted [59]. As a large per-centage of them do not receive effective antiretroviraltherapy, seizures related to encephalopathy and opportun-istic infections are a common occurrence. Thus, aromaticanticonvulsants will continue to represent an importanttreatment option, and the threat of AHS will, therefore,continue.

Table 4. Therapeutic interventions used for management of anticonvulsant hypersensitivity syndrome.

Therapeutic modality Justification and remarks

Hospitalisation [46] Possibility of multi-organ involvement, severe clinical course and likelihood of uncontrolled seizures after anticonvulsant is withdrawn

Withdraw the incriminating drug [1-3] Accumulation of drug metabolites that are responsible for the reaction

Use of alternate anticonvulsant [1-3,54,55] To prevent rebound seizures and management of status epilepticusIt might not be possible to use valproate in the acute phase due to the presence of concomitant hepatic dysfunction. Diazepam can be used.Valproate or gabapentin may represent safe alternatives for the long-term management. Cross-sensitivity with lamotrigine and felbamate should be ruled out.

Nutrition and intravenous fluids [1] Enteral nutrition is preferredParenteral nutrition if significant oral intake is not possible for several days in view of mucosal lesions, organ damage and poor general healthIntense fluid management: supplementary intravenous fluids if skin lesions are extensive due to the possibility of fluid loss through the ulcerating skin lesions

Prevention and treatment of infection [46] Admission to isolation room or intensive care unit depending upon the severity and extent of cutaneous sloughingProphylactic antimicrobial agents not requiredPrompt treatment of infections

Skin care Denuded skin should be cleansed carefully as it is a primary portal for infection

Management of ocular disease [46] Daily ophthalmological consultationSeparation of eyelids and mucosal surfaces of the palpebral and bulbar conjunctivae in order to prevent symblepharon

Symptomatic treatment and other general measures [1,46]

H1-antagonists used for the treatment of pruritus: diphenhydramine and/or hydoxyzineH2 antagonists used in patients with mucosal sloughing and suspected SJSSuitable antipyretic and analgesic agentsCareful pulmonary toilet: nebulised medications, incentive spirometry

Corticosteroids [1-3,44,46] Use of prednisolone and methylprednisolone as immunomodulators, as these were used in the treatment of severe cutaneuous adverse reactionsRisks involved: immunosuppression, sepsis, prolonged recovery, relapse after withdrawalOther forms of immunomodulation: plasmaphersis, cyclophosphamide, cyclosporin, IVIG

Intravenous immunoglobulin [44,56] Case reports justifying its use have been published

Patient counselling Provide a list of anticonvulsant drugs that are contraindicated in the patientPatients should wear medication identification braceletScreening of first-degree relatives for the determination of risk for AHS

AHS: Anticonvulsant hypersensitivity syndrome; IVIG: Intravenous immunoglobulin; SJS: Stevens-Johnson syndrome.



• Certain diseases, without a confirmatory test, are diagnosedon the basis of a combination of manifestations. KD is anexample of such a disease. There is a need to evolve consen-sus guidelines for the diagnosis of AHS, which willempower the majority of physicians who do not have accessto tests such as the LTA to diagnose AHS. The consensuscould take into account the clinical and laboratory manifes-tations to make a diagnosis. This would also ensure thatunderdiagnosis is minimised.

• Adverse event monitoring programs should actively seekAHS reports for determination of true prevalence and toidentify newer drugs with potential to cause AHS andcompare how they fare with older AEDs in this regard.

• The matters related to management of the reaction alsoneed to be clarified. Withdrawal of the incriminated drug isthe only therapeutic intervention whose utility has beenproved. Controlled trials should be undertaken to elucidatethe exact role of other therapeutic measures.

• Although there is an assay to diagnose AHS and predict thelikelihood of AHS in siblings, issues related to the cost andavailability of the test and its cost-effectiveness need to beaddressed. The potential of newer anticonvulsant drugs tocause AHS or similar manifestations is not well studied. Itis a matter of speculation whether or not the LTA could beused for these drugs as well.

• New AEDs will continue to be developed and marketed. Itis necessary that their potential to cause AHS be assessedearly in the course of drug development.

• There is a need for basic research to elucidate the underly-ing molecular mechanisms. Once the pathogenesis andmolecular basis of AHS are completely understood, itmight be possible to accurately estimate the risk of devel-opment of AHS in a given patient and perhaps also designdrugs that do not have the potential to cause thissyndrome [4].

Bibliography1. BESSMERTNY O, HATTON RC,

GONZALEZ-PERALTA RP: Antiepileptic hypersensitivity syndrome in children. Ann. Pharmacother. (2001) 35(5):533-538.

2. KNOWLES SR, SHAPIRO LE, SHEAR NH: Anticonvulsant Hypersensitivity Syndrome. Incidence, Prevention and Management. Drug Saf. (1999) 21(6):489-501.

3. VITTORIO CC, MUGLIA JJ: Anticonvulsant hypersensitivity syndrome. Arch. Intern. Med. (1995) 155(21):2285-2290

4. GILMAN JT, DUCHOWNY M: Introduction: Allergic Hypersensitivity to Antiepileptic Drugs: Past, Present, and Future. Epilepsia (1998) 39(Supp 7):S1-S2.

5. TENNIS P, STERN R: Risk of serious cutaneous disorders after initiation of use of phenytoin, carbamazepine, or sodium valproate: a record linkage study. Neurology (1997) 49(2):542-546.

6. ROUJEAU JC, STERN R: Severe adverse cutaneous reactions to drugs. N. Engl. J. Med. (1994) 331(19):1272-1285

7. SPIELBERG S, SHEAR N, HUTSON N et al.: In-vitro assessment of hypersensitivity syndrome associated with sorbinil. Ann. Intern. Med. (1991) 114(9):701-724.

8. KNOWLES S, SHAPIRO L, SHEAR N: Serious adverse reactions induced by minocycline: a report of 13 patients and review of the literature. Arch. Dermatol. (1996) 132(8):934-939.

9. GUPTA AK, KOPSTEIN JB, SHEAR NH: Hypersensitivity reaction to terbinafine. J. Am. Acad. Dermatol. (1997) 36(6 Pt 1):1018-1019

10. KNOWLES SR, GUPTA AK, SHEAR NH, SAUDER D: Azathioprine hypersensitivity-like reactions- a case report and review of the literature. Clin. Exp. Dermatol. (1995) 20(4):353-356.

11. MULLICK F, ISHAK F: Hepatic injury associated with diphenyl hydantoin therapy. Am. J. Clin. Pathol. (1980) 74(4)442-452.

12. AGUIR D, PAZO R, DURAN I et al.: Toxic epidermal necrolysis in patients receiving anticonvulsants and cranial irradiation: a risk to consider. J. Neurooncol. (2004) 66(3):345-350

13. ROUJEAU J-C, CHOSIDOW O, SAIAG P, GUILLAUME J-C: Toxic epidermal necrolysis (Lyell Syndrome). J. Am. Acad. Dermatol. (1990) 23(6 pt 1):1039-1058.

14. DREIFUSS FE, LANGER DH: Hepatic considerations in the use of antiepileptic drugs. Epilepsia (1987) 28(Supp2):S23-S29.

15. SILVERMAN AK, FAIRLEY J, WONG RC: Cutaneous and immunologic reactions to phenytoin. J. Am. Acad. Dermatol. (1988) 18(4 Part 1):721-741.

16. SPIELBERG S: In-vitro analysis of idiosyncratic drug reactions. Clin. Biochem. (1986) 19(2):142-144.

17. GLEICHMANN H: Studies on the mechanism of drug sensitization: T cell

dependent popliteal lymph node reaction to diphynylhydantoin. Clin. Immunol. Immunopathol. (1981) 18(2):203-211.

18. KANO Y, INAOKA M, SHIOHARA T: Association between anticonvulsant hypersensitivity syndrome and human herpesvirus 6 reactivation and hypogammaglobulinemia. Arch. Dermatol. (2004) 140(2):183-188.

19. SHEAR N, SPIELBERG S: Anticonvulsant hypersensitivity syndrome: In vitro assessment of risk. J. Clin. Invest. (1988) 82(6):1826-1832.

20. SPIELBERG S, GORDON G, BLAKE D, MELLITS ED, BROSS DS: Anticonvulsant toxicity in vitro: Possible role of arene oxides. J. Pharmacol. Exp. Ther. (1981) 217(2):386-289.

21. RILEY F, LEEDER S: In vitro analysis of metabolic predisposition to drug hypersensitivity reactions. Clin. Exp. Immunol. (1995) 99(1):1-6.

22. SPIELBERG S, GORDON G, BLAKE D, GOLDSTEIN DA, HERLONG HF: Predisposition to phenytoin hepatotoxicity assessed in vitro. N. Engl. J. Med. (1981) 305(13):722-727.

23. FRIEDMANN P, STRICKLAND I, PIRMOHAMED M, PARK BK: Investigations of mechanisms in toxic epidermal necrolysis by carbamazepine. Arch. Dermatol. (1994) 130(5):596-604.

24. LEEDER J, RILEY R, COOK V, SPIELBERG SP: Human anti-cytorchrome P450 antibodies in aromatic



anticonvulsant-induced hypersensitivity reactions. J. Pharmacol. Exp. Ther. (1992) 263(1):360-367.

25. MARTZ F, FEILINGER C, BLAKE DA: Phenytoin teratogenesis: Correlation between embryopathic effect and covalent binding of putative arene oxide metabolites in gestational tissue. J. Pharmacol. Exp. Ther. (1977) 203(1):231-239.

26. BENET LZ, MITCHELL JR, SHEINER LB: Pharmacokinetics: The dynamics of drug absorption, distribution and elimination. In: Goodman and Gilman’s The pharmacological basis of therapeutics (8th edition). Goodman Gilman A, Rall TW, Nies AS, Taylor P (Eds.) Pergamon Press, New York, (1991) Pp 3-32.

27. MORKUNAS AR, MILLER MB: Anticonvulsant hypersensitivity syndrome. Crit. Care Clin. (1997) 13(4):727-739.

28. BROWN TS, APPEL JE, KASTELER JS, CALLEN JP: Hypersensitivity reaction in a child due to lamotrigine. Pediatr. Dermatol. (1999) 16(1):46-49.

29. MAGGS JL, NAISBITT DJ, TETTEY JN, PIRMOHAMED M, PARK BK: Metabolism of lamotrigine to a reactive arene oxide intermediate. Chem. Res. Toxicol. (2000) 13(11):1075-1081.

30. PLANTIN P, CARTIER H, LE BIHAN G, CLOUARD P, LELLOUCHA F, LEROY JP: Syndrome d’hyeprsensibility medicamenteuse au cours. d’un traitement par acide valproique. Presse Med. (1985) 24(34):1624.

31. NEUMANN MG, MALKIEWICZ IM, PHILLIPS EJ et al.: Monitoring adverse drug reactions to sulfonamide antibiotics in human immunodeficiency virus-infected individuals. Ther. Drug Monit. (2002) 24(6):728-736.

32. BELLER TC, BOYCE JA: Prolonged anticonvulsant hypersensitivity syndrome related to lamotrigine in a patient with human immunodeficiency virus. Allerg. Asthma Proc. (2002) 23(6):415-419.

33. LEEDER JS, RILEY RJ, COOK VA, SPIELBERG SP: Human anti-cytochrome P450 antibodies in aromatic anticonvulsant-induced hypersensitivity reactions. J. Pharmacol. Exp. Ther. (1992) 263(1):360-367.

34. PIRMOHAMED M, KITTERINGHAM NR, BRECKENRIDGE AM, PARK BK: Detection of an autoantibody directed

against human liver microsomal protein in a patient with carbamazepine hypersensitivity. Br. J. Clin. Pharmacol. (1992) 33(2):183-186.

35. RZANY B, CORREIA O, KELLY JP, NALDI L, AUQUIER A, STERN R: Risk of Stevens-Johnson Syndrome and toxic epidermal necrolysis during first weeks of anti-epileptic therapy: a case control study. Study group of the International Case control study on Severe Cutaneous Adverse Reactions. Lancet (1999) 353(9171):2190-2194.

36. HARUDA F: Phenytoin hypersensitivity: 38 cases. Neurology (1979) 29(11):1480-1485.

37. SCHLIENGER RG, SHEAR NH: Antiepileptic drug hypersensitivity syndrome. Epilepsia (1998) 39(Supp 7):S3-S7.

38. BABA M, KARAKAS M, AKSUNGUR VL, HOMAN S, YUCEL A, ACAR MA, MEMISOGLU HR: The anticonvulsant hypersensitivity syndrome. JEADV (2003) 17(4):399-401.

39. CIERNIK IF, THIEL M, WIDMER U: Anterior uveitis and the anticonvulsant hypersensitivity syndrome. Arch. Intern. Med. (1998) 158(2):192.

40. SCHREIBER M, McGREGOR J: Pseudolymphoma syndrome. Arch. Derm. (1968) 97(3):297-300.

41. D’INCAN N, SOUTEYRAND P, BIGNON Y, FONCK Y, ROGER H: Hydantoin-induced cutaneous pseudolymphoma with clinical, pathologic, and immunologic aspects of Sezary syndrome. Arch. Dermatol. (1992) 128(10):1371-1374.

42. YATES P, STOCKDILL G, MCINTYRE M: Hypersensitivity to carbamazepine presenting as pseudolymphoma. J. Clin. Pathol. (1986) 39(11):1224-1228.

43. SHUTTLEWORTH D, GRAHAM-BROWN R, WILLIAMS A, CAMPBELL AC, SEWELL H: Pseudolymphoma associated with carbamazepine. Clin. Exp. Dermatol. (1984) 9(4):421-423.

44. MOSTELLA J, PIERONI R, JOSE R, FINCH CK: Anticonvulsant hypersensitivity syndrome: Treatment with corticosteriods and intravenous immunoglobulin. Southern Med. J. (2004) 97(3):319-321.

45. BROWN CE, SMITH GD, CONIGLIONE T: Anticonvulsant hypersensitivity: an unfortunate case of triple exposure to phenytoin. J. Fam. Pract. (1997) 45(5):434-437.

46. GRIEBEL ML: Acute management of hypersensitivity reactions and seizures. Epilepsia (1998) 39(Supp 7):S17-S21.

47. TOUSSAINT S, KAMINO H. Non-infectious erythematous papular and squamous diseases of the skin. In: Lever’s Histopathology of Skin (8th edition), Elder D (Ed.), Lippincott-Raven, Philadelphia (1997) Pp151-184.

48. ACKERMAN BA. Histologic diagnosis of inflammatory skin diseases, Philadephia Lea and Febiger (1978) Pp581-639.

49. BESSMERTNY O, PHAM T: Anitepileptic hypersensitivity syndrome. Curr. Allergy Asthma Rep. (2002) 2(1):34-49.

50. PIRMOHAMED M, GRAHAM A, ROBERTS P et al.: Carbamazepine hypersensitivity: Assessment of clinical and in vitro chemical cross-reactivity with phenytoin and oxcarbazepine. Br. J. Clin. Pharmacol. (1991) 32(6):741-749.

51. NEUMAN MG, MALKIEWICZ IM, SHEAR NH: A novel LTA to assess drug hypersensitivity syndromes. Clin. Biochem. (2000) 33(7):517-524.

52. JONES M, FERNANDEZ-HERRERA J, DORADO J, SOLS M, RUIZ M, GARCIA-DIEZ A: Epicutaneous test in carbamazepine cutaneous reactions. Dermatology (1994) 188(1):18-20.

53. ALANKO K: Patch testing in cutaneous reactions caused by carbamazepine. Contact Dermatitis (1993) 29(5):254-257.

54. BERTZ RJ, HOWRIE DL: Diazepam by continuous intravenous infusion for status epilepticus in anticonvulsant hypersensitivity syndrome. Ann. Pharmacother. (1993) 27(3):298-301.

55. HAMER HM, MORRIS HH: Successful treatment with gabapentine in the presence of hypersensitivity syndrome to phenytoin and carbamazepine: a report of three cases. Seizure (1999) 8(3):190-192.

56. SCHEUERMAN O, NOFECH-MOSES Y, RACHMEL A, ASHKENAZI S: Successful treatment of antiepileptic drug hypersensitivity syndrome with intravenous immunoglobulin. Pediatrics (2001) 107(1):E14.



57. BAVDEKAR SB, MURANJAN MN, GOGTAY NJ, KANTHARIA V, KSHIRSAGAR NA: Anticonvulsant hypersensitivity syndrome: Lymphocyte toxicity assay for the confirmation of diagnosis and risk assessment. Ann. Pharmacother. (2004) 38(10):1648-1650.

58. MANI KS, RANGAN G, SRINIVAS HV, SRINDHARAN VS, SUBBAKRISHNA DK: Epilepsy control with phenobarbital or phenytoin in rural south India: the Yelandur study. Lancet (2001) 357(9265):1316-1320.

59. DYER O: New 20 year patents threaten to end AIDS drugs for developing countries. Br. Med. J. (2004) 329(7478):1308.

AffiliationNithya J Gogtay MD, DNB1†, Sandeep B Bavdekar MD, DCH2 & Nilima A Kshirsagar MD, PhD1

†Author for correspondence1Seth GS Medical College & KEM Hospital, Department of Clinical Pharmacology, Parel, Mumbai 400 012, INDIATel: +91 22 2417 4420; Fax: +91 22 2414 3435;E-mail: [email protected] GS Medical College & KEM Hospital, Department of Pediatrics, Parel, Mumbai 400 012, INDIATel: +91 22 2413 3767; Fax: +91 22 2414 3435;E-mail: [email protected]

Documents

Anticonvulsant hypersensitivity syndrome: a review