Epilepsy Research (2011) 94, 213—217

journal homepage: www.elsevier.com/locate/epilepsyres

Influence of chemical structure on skin reactions

induced by antiepileptic drugs—–The role of the
aromatic ring
Xiang-qing Wang ∗, Xiao-bing Shi, Ran Au, Fu-shun Chen, Fang Wang,
Sen-yang Lang

Department of Neurology, The Chinese PLA General Hospital, No. 28, Fuxing Road, Beijing 100853, China

Received 18 December 2010; received in revised form 30 January 2011; accepted 6 February 2011
Available online 4 March 2011

KEYWORDS Summary Here we assessed whether the presence of an aromatic ring as a commonality in
Antiepileptic drugs; chemical structures of AEDs can explain skin reaction. We found that 164 cases of skin reac-
Skin reactions; tions associated with the use of AEDs were reported. Aromatic AEDs were suspected in 88.41%
Aromatic ring; (145/164) of patients with skin reactions versus 59.80% (2316/3873) of patients without skin
Gender; reactions. The presence of an aromatic ring in the chemical structure was associated with a
Chinese population significant increased risk of skin reactions (adjusted ROR 3.50; 95% CI 2.29, 5.35). Among the
aromatic AEDs, skin reactions were significantly associated with carbamazepine, lamotrigine,
and oxarbazepine. These results confirm that the presence of an aromatic ring as a common
feature in chemical structures of AEDs partly explains AED-skin reactions. Skin reactions were
reported triple as frequently with aromatic AEDs than with non-aromatic AEDs.
© 2011 Elsevier B.V. All rights reserved.

Introduction la Morena, 2001) and ‘anticonvulsant hypersensitivity syn-

It is well known that the use of antiepileptic drugs
(AEDs) may cause skin reactions in susceptible patients
with a varying clinical presentation. Most skin reaction is
the common and mild maculopapular rashes, which dis-
appear within a few days after discontinuation of the
therapy. The most feared AED-related adverse reactions
are ‘Stevens—Johnson syndrome (SJS)’, ‘toxic epidermal
necrolysis (TEN)’ (Mockenhaupt et al., 2005; Arroyo and de
∗ Corresponding author. Tel.: +86 13661018488;
fax: +86 1066939251.
E-mail address: bjxqwang@yahoo.com.cn (X.-q. Wang).
drome’ (Newell et al., 2009; Cumbo-Nacheli et al., 2008).
The estimated incidence of these severe events ranges from
1 per 1000 to 1 per 10,000 users of AEDs (Mockenhaupt
et al., 2005). Various reports have shown that specific AEDs
such as carbamazepine, phenytoin, phenobarbital and lam-
otrigine were connected with skin reactions (Chadwick,
1999).
The mechanism by which these AEDs induce skin reac-
tions is unknown. One of the main hypotheses is that AEDs
containing an aromatic ring in their chemical structure can
form an arene-oxide intermediate (Knowles et al., 1999)
(Fig. 1). This chemically reactive product may become
immunogenic through interactions with proteins or cellular
macromolecules in accordance with the hapten hypothe-

0920-1211/$ — see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.eplepsyres.2011.02.005

Downloaded for Anonymous User (n/a) at Dr Kariadi General Hospital Medical Center from ClinicalKey.com by Elsevier on May 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
214 X.-q. Wang et al.

with epilepsy taking at least one antiepileptic drugs, seen by at least

Figure 1 Possible metabolic pathway for production of toxic
metabolites of aromatic antiepileptic drugs.
two epileptologists at the Epilepsy Center of the Chinese PLA Gen-
eral Hospital. A cutaneous adverse reaction was defined as any types
of rash (erythematous, maculo-papular, papular, pustular or unspec-
ified), which had no other obvious reason than an antiepileptic drug
effect and resulted in contacting a physician. We recorded the clin-
ical description of rashes and all drugs using in all patients. Date
of occurrence of rash and its clinical description were documented.
We have performed telephone interviews among all patients with
AED-related rashes and some no-rash patients whose documenta-
tion was considered insufficient. All patients had at least one office
visit or a telephone interview after the occurrence of skin reaction.

sis, suggesting that this structural commonality between Exposure definitions

AEDs may be responsible for hypersensitivity reactions.
This hypothesis is based on incidental case reports and
in vitro experiments. Only one observational study sup-
ports this theory which investigates the special type of skin
reactions—hypersensitivity reactions (Handoko et al., 2008).
Skin reactions range with increasing severity from the com-
mon and mild maculopapular rashes, to the hypersensitivity
reactions with fever and internal organ involvement, and to
Stevens—Johnson syndrome and toxic epidermal necrolysis.
So far, no in vivo experimental or observational studies are
present to assess the association between the presence of
an aromatic ring as a structural commonality in AEDs and
skin reactions.
Therefore, we conducted this study to assess whether
an aromatic ring as a commonality in chemical struc-
tures of AEDs is associated with the occurrence of skin
reactions using the epileptic database of our epileptic
center.
AEDs were classified as aromatic anticonvulsants if their chemi-
cal structure contained at least one aromatic ring (carbamazepine,
phenytoin, phenobarbital, oxarbazepine and lamotrigine). All other
AEDs were classified as non-aromatic (Fig. 2).
Data analysis
Skin reactions were analyzed using a case/non-case design. Cases
were defined as those patients with skin reactions induced by AEDs.
Non-cases were patients without skin reactions. The Chi-squared
test, Student’s t-test were used to compare rashes and non-rashes
as appropriate. The strength of the association between the skin
reactions and the aromatic AEDs in comparison with non-aromatic
AEDs was calculated using the skin reaction reporting odds ratio
(ROR) as a measure of disproportionality. The calculation of a ROR is
comparable to the calculation of an odds ratio from a case—control
study. RORs, adjusted for age and sex, were calculated by means
of logistic regression analysis and expressed as point estimates with
corresponding 95% confidence intervals.

Methods
From February 1999 to September 2010, consecutive Chinese
patients with epilepsy were studied retrospectively. We systemati-
cally reviewed the medical records of 4037 consecutive outpatients
Results
4037 Chinese epileptic patients taking at least one of AEDs
were included in this study, of which 2481 (61.46%) were

Table 1 Main patient characteristics and risk of rash for each AEDs drug.

Patient characteristics Patients with skin Patients without skin p-Value Rash frequency
reactions (n = 164) reactions* (n = 3873) (%)

Sex
Female [n (%)] 84 (51.22) 1472 (38.01) 0.0009a 5.40
Mean age (y) ± SD 28.52 ± 14.82 26.46 ± 14.41 0.07b
Drugs on admission
Aromatic AEDs
Carbamazepine 78 1529 4.85
Lamotrigine 28 243 10.33
Oxarbazepine 15 198 7.04
Phenobarbital 9 363 2.42
Phenytoin 15 484 3.01
Non-aromatic AEDs
Valproate 10 1561 0.64
Topiramate 7 660 1.05
Gabapentin 1 51 1.92
Levetiracetam 1 110 0.90
a Pearson 2 test.
b Student’s t-test.
* 1246 patients take more than one antiepileptic drugs on admission.

Downloaded for Anonymous User (n/a) at Dr Kariadi General Hospital Medical Center from ClinicalKey.com by Elsevier on May 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
Influence of chemical structure on skin reactions induced by antiepileptic drugs 215

Figure 2 Aromatic and non-aromatic antiepileptic drugs (AEDs).

Downloaded for Anonymous User (n/a) at Dr Kariadi General Hospital Medical Center from ClinicalKey.com by Elsevier on May 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
216
Table 2 Use of aromatic and non-aromatic antiepileptic drugs (AEDs) in patients with and without skin reactions.
AED Skin reactions Non-skin reactions*
n = 164 [n (%)] n = 3873 [n (%)]
Non-aromatic AEDs 19 (11.59)
Aromatic AEDs 145 (88.41)
Carbamazepine 78 (47.56)
Lamotrigine 28 (17.07)
Oxarbazepine 15 (9.15)
Phenobarbital 9 (5.49)
Phenytoin 15 (9.15)
a Adjusted for age, and sex.
ROR = reporting odds ratio.
* 1246 patients take more than one antiepileptic drugs on admission.
males and 1556 (38.54%) were females. Overall, 4.06%
(164/4037) of patients experienced a skin reaction to one
of the AEDs marked in China, while 18 patients had his-
tory with skin reactions to other different AEDs. Among the
164 patients experiencing a skin reaction to any antiepilep-
tic drugs, 80 were males and 84 females. The mean age
of patients experiencing skin reactions was 28.52 years,
and of patients without skin reactions was 26.46 years
(Table 1). No significant difference in the ages of the two
groups was detected (p = 0.07). However, females (5.40%,
84/1556) were nearly twice as likely to develop a rash as
were males (3.22%, 80/2481) (adjusted OR = 1.67, CI 1.22-
2.29, p < 0.05). Among the aromatic AEDs, LTG caused the
highest incidence of skin reactions (10.73%, 28/243), fol-
lowed by OXC (7.04%), CBZ (4.85%), PHT (3.01%), and PB
(2.42%). Among the non-aromatic AEDs, GBP had the high-
est rate (1.92%), followed by TPM (1.05%), LEV (0.92%), and
VPA (0.64%) (Table 1).
Aromatic AEDs were suspected in 88.41% (145/164) of
patients with skin reactions versus 59.80% (2316/3873) of
patients without skin reactions. The presence of an aromatic
ring in the chemical structure was associated with a signifi-
cant increased risk of skin reactions (adjusted ROR 3.50; 95%
CI 2.29, 5.35). Among the aromatic AEDs, skin reactions were
significantly associated with carbamazepine (adjusted ROR
1.40; 95% CI 1.02, 1.91), lamotrigine (adjusted ROR 5.12;
95% CI 3.29, 7.98), and oxarbazepine (adjusted ROR 2.91;
95% CI 1.66, 5.10) (Table 2).
Discussion
This study shows that a commonality in chemical structures
of AEDs partly explains skin reactions. We found that skin
reactions were reported triple as frequently with aromatic
AEDs than with non-aromatic AEDs. Among the aromatic
AEDs, skin reactions were significantly associated with car-
bamazepine, lamotrigine, and oxarbazepine.
In general, it is assumed that small molecules with
molecular weights <1 kDa cannot directly induce an immune
response. AEDs, as well as the majority of other drugs, fall
into this category. Therefore, the hapten formation hypoth-
esis proposes that drugs or reactive metabolites of drugs
act as haptens and bind to proteins or other endogenous
macromolecules to induce an immune response (Park et al.,
Downloaded for Anonymous User (n/a) at Dr Kariadi General Hospital Medical Center from ClinicalKey.com by Elsevier on May 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
X.-q. Wang et al.
ROR (95%CI)
Crude Adjusteda
1557 (40.20) 1.00 (reference) 1.00 (reference)
2316 (59.80) 3.63 (2.35, 5.53) 3.50 (2.29, 5.35)
1529 (39.78) 1.39 (1.02, 1.90) 1.40 (1.02, 1.91)
143 (3.69) 5.37 (3.46, 8.34) 5.12 (3.29, 7.98)
133 (3.43) 2.83 (1.62, 4.95) 2.91 (1.66, 5.10)
363 (9.37) 0.56 (0.28, 1.11) 0.55 (0.27, 1.09)
484 (12.50) 0.71 (0.41, 1.21) 0.68 (0.40, 1.17)
1998). In 1974, the metabolic formation of arene-oxides
has been described (Jerina and Daly, 1974). According to
them, arene-oxides were responsible for many toxic and
carcinogenic properties of aromatic hydrocarbons. Another
argument supporting this hapten hypothesis is the cross-
sensitivity that has been reported among patients using
aromatic AEDs (Klassen and Sadler, 2001). Cross sensitivity
among aromatic AEDs occurs in 40—58% of patients in vivo
(Hyson and Sadler, 1997) and has been reported as high as
80% in an in vitro assay (Brown et al., 1997). A rechallenge
with a possible cross-reactive AED resulted in hypersensitiv-
ity reactions in up to 87% of patients (Shear and Spielberg,
1988).
Arif et al. (2007) recently studied predictors of rash asso-
ciated with AEDs. They found higher rash rates in patients
treated with phenytoin, lamotrigine and carbamazepine
(all aromatic AEDs) and lower rates with levetiracetam,
gabapentin and valproate (all non-aromatic AEDs). Handoko
et al. (2008) had found an odds ratio of 2.18 for the associ-
ation between aromatic AEDs versus non-aromatic AEDs and
hypersensitivity reactions induced by AEDs.
Apart from the hapten formation hypothesis, another
immune mechanism might be involved. There is direct,
non-covalent binding of the drug to the T-cell receptor of
specific T-cell clones. Drug-specific T-cells have been iden-
tified for carbamazepine (Pichler, 2003) and lamotrigine
(Naisbitt et al., 2003).
The mechanisms of skin reactions are still not fully under-
stood. Many factors may modify the susceptibility to this
adverse effect and mechanisms may differ. Pharmacogenetic
variations in drug biotransformation may be crucial, but gen-
der, age, the drug titration schedule, the co-medication and
the individual immunological status may also influence the
predisposition to these reactions. We also found higher rates
of skin reactions in females than in males, and a logistic
regression analysis confirmed that female gender is one of
the risk factors for AED-related skin reactions. It has been
reported that gender and age are two factors involved in
drug-induced skin reactions. In most cases of adverse drug
reactions, females are at higher risk than males (Riedl and
Casillas, 2003; Gomes and Demoly, 2005).
Recently, a strong association was found between
Stevens—Johnson syndrome and the HLA-B*1502 gene in
Han Chinese treated with CBZ, indicating that the HLA
genotype may be involved in AED-related reactions (Chung
Influence of chemical structure on skin reactions induced by antiepileptic drugs
et al., 2004). If confirmed in larger studies and other ethnic
cohorts, identification of genetic polymorphisms predispos-
ing AED-induced skin reactions and subsequent avoidance
of aromatic AED treatment could help to prevent life-
threatening events.
Reliability of the study
This study provided important evidence to support the
notion that selecting appropriate AEDs for treatment is very
important to reduce the risk for drug-induced side effects.
This study is based on a retrospective analysis, which is
the major limitation of it, including a remote history of
only vaguely described rashes. Not every skin reaction was
examined by a physician. Recall bias about AED treatment
initiated prior to treatment at our center may have resulted
in underreporting of prior AED-related skin reactions in some
patients, particularly the old medications (phenobarbital)
used decades ago. In our hospital, epilepsy is traditionally
treated and followed up by hospital specialists and detailed
medical records were available for most patients back to our
outpatient department or by telephone interviews.
The approach outlined in this study should be replicated
with other study designs, such as prescription event mon-
itoring, case control surveillance or record linkage by use
of large automated databases. Unraveling the association
between a commonality in chemical structures of AEDs and
other types of skin reactions (such as hypersensitivity reac-
tions, SJS/TEN) with these methods could provide additional
evidence for the association we found.
Conclusion
In conclusion, this study demonstrates that apparently skin
reactions can partly be explained by a commonality in chem-
ical structures of AEDs. A significant association was found
for aromatic versus non-aromatic AEDs and skin reactions.
Females seem to be at higher risk than males.
Acknowledgement
No sources of funding were used in the preparation of this
study.
References
Arif, H., Buchsbaum, R., Weintraub, D., Koyfman, S., Salas-Humara,
C., Bazil, C.W., Resor, SR., Hirsch, L.J., 2007. Comparison and
predictors of rash associated with 15 antiepileptic drugs. Neu-
rology 68, 1701—1709 (PubMed: 17502552).
Arroyo, S., de la Morena, A., 2001. Life-threatening adverse events
of antiepileptic drugs. Epilepsy Res. 47, 155—174 (PubMed:
11673030).
Downloaded for Anonymous User (n/a) at Dr Kariadi General Hospital Medical Center from ClinicalKey.com by Elsevier on May 21, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
217
Brown, C.E., Smith, G.D., Coniglione, T., 1997. Anticonvulsant
hypersensitivity: an unfortunate case of triple exposure to
phenytoin. J. Fam. Pract. 45, 434—437 (PubMed: 9374970).
Chadwick, D., 1999. Safety and efficacy of vigabatrin and car-
bamazepine in newly diagnosed epilepsy: a multi centre
randomized double blind study. Vigabatrin European Monother-
apy Study Group. Lancet 354, 13—19 (PubMed: 10406359).
Chung, W.H., Hung, S.I., Hong, H.S., Hsih, M.S., Yang, L.C., Ho,
H.C., Wu, J.Y., Chen, Y.T., 2004. Medical genetics: a marker
for Stevens—Johnson syndrome. Nature 428, 486 (PubMed:
15057820).
Cumbo-Nacheli, G., Weinberger, J., Alkhalil, M., Thati, N., Baptist,
A.P., 2008. Anticonvulsant hypersensitivity syndrome: is there a
role for immunomodulation? Epilepsia 49, 2108—2112 (PubMed:
18637830).
Gomes, E.R., Demoly, P., 2005. Epidemiology of hypersensitivity
drug reactions. Curr. Opin. Allergy Clin. Immunol. 5, 309—316
(PubMed: 15985812).
Handoko, K.B., van Puijenbroek, E.P., Bijl, A.H., Hermens, W.A.,
Zwart-van Rijkom, J.E., Hekster, Y.A., Egberts, T.C., 2008. Influ-
ence of chemical structure on hypersensitivity reactions induced
by antiepileptic drugs: the role of the aromatic ring. Drug Saf.
31, 695—702 (PubMed: 18636788).
Hyson, C., Sadler, M., 1997. Cross sensitivity of skin rashes with
antiepileptic drugs. Can. J. Neurol. Sci. 24, 245—249 (PubMed:
9276112).
Jerina, D.M., Daly, J.W., 1974. Arene oxide: a new aspect of drug
metabolism. Science 185, 573—582 (PubMed: 4841570).
Klassen, B.D., Sadler, R.M., 2001. Induction of hypersensitivity to a
previously tolerated antiepileptic drug by a second antiepileptic
drug. Epilepsia 42, 433—435 (PubMed: 11442165).
Knowles, S.R., Shapiro, L.E., Shear, N.H., 1999. Anticonvulsant
hypersensitivity syndrome: incidence, prevention and manage-
ment. Drug Saf. 21, 489—501 (PubMed: 10612272).
Mockenhaupt, M., Messenheimer, J., Tennis, P., Schlingmann, J.,
2005. Risk of Stevens—Johnson syndrome and toxic epider-
mal necrolysis in new users of antiepileptics. Neurology 64,
1134—1138 (PubMed: 15824335).
Naisbitt, D.J., Farrell, J., Wong, G., Depta, J.P., Dodd, C.C.,
Hopkins, J.E., Gibney, C.A., Chadwick, D.W., Pichler, W.J.,
Pirmohamed, M., Park, B.K., 2003. Characterization of drug-
specific T cells in lamotrigine hypersensitivity. J. Allergy Clin.
Immunol. 111, 1393—1403 (PubMed: 12789244).
Newell, B.D., Moinfar, M., Mancini, A.J., Nopper, A.J., 2009.
Retrospective analysis of 32 pediatric patients with anticonvul-
sant hypersensitivity syndrome (ACHSS). Pediatr. Dermatol. 26,
536—546 (PubMed: 19840307).
Park, B.K., Pirmohamed, M., Kitteringham, N.R., 1998. Role of drug
disposition in drug hypersensitivity: a chemical, molecular, and
clinical perspective. Chem. Res. Toxicol. 11, 969—988 (PubMed:
9760271).
Pichler, W.J., 2003. Delayed drug hypersensitivity reactions. Ann.
Intern. Med. 139, 683—693 (PubMed: 14568857).
Riedl, M.A., Casillas, A.M., 2003. Adverse drug reactions: types and
treatment options. Am. Fam. Physician 68, 1781—1790 (PubMed:
14620598).
Shear, N.H., Spielberg, S.P., 1988. Anticonvulsant hypersensitiv-
ity syndrome: in vitro assessment of risk. J. Clin. Invest. 82,
1826—1832 (PubMed: 3198757).