Genetic Susceptibility to Fatal Rash from Carbamazepine? The Case for HLA Screening

Abstract & Commentary

By Padmaja Kandula, MD, Assistant Professor of Neurology and Neuroscience, Comprehensive Epilepsy Center, Weill Medical College. Dr. Kandula reports no financial relationships relevant to this field of study.

Synopsis: Carbamazepine has been linked with severe forms of hypersensitivity reactions. As such, genome wide approaches to identify patients-at-risk have become increasingly important.

Sources: Chen P, et al. Carbamazepine-induced toxic effects and HLA-B*1502 screening in Taiwan. New Engl J Med 2011;364:1126-1133. McCormack M, et al; HLA-A*3101 and carbamazepine-induced hypersensitivity reactions in Europeans. New Engl J Med 2011;364:1134-1143.

The most severe form of carbamazepine hypersensi-tivity reactions is Stevens-Johnson Syndrome (SJS), characterized by a blistering rash involving the mucous membranes and systemic inflammatory side effects. The most extreme form, toxic epidermal necrolysis (TEN), involves even greater epidermal detachment. Both syndromes carry a high risk of mortality and are the focus of two recent studies from Taiwan and the UK. Previous reports have confirmed the association of SJS and TEN with HLA-B*1502 allele in individuals of Han Chinese descent. In this first paper from Taiwan, the authors prospectively screened patients, by using HLA-B*1502 genotyping, to identify subjects at risk for SJS-TEN before starting clinical treatment with carbamazepine. The second paper from the UK presents data implicating the HLA variant, HLA-A*3101, in the full range of carbamazepine-associated hypersensitivity reactions.

Chen et al screened patients from 23 hospitals in Taiwan and identified 4877 subjects who were carbamazepine candidates (epilepsy, neuralgia, neuropathic pain, tinnitus, bipolar or psychiatric disorders). Patients with a previous history of carbamazepine allergy, bone marrow transplant, and non-Han Chinese descent were excluded. Patients were told to defer taking carbamazepine until results of genetic testing were complete. All individuals had an initial screening visit. Only HLA-B*1502 positive individuals were asked to return for a second office visit to communicate test results and offer alternative treatments. Weekly telephone interviews with all subjects were conducted over the next 2 months. Any clinical suspicion of adverse drug reaction prompted an immediate hospital evaluation by a staff dermatologist. Based on an 8% historical prevalence of the HLA-B*1502 allele in the Taiwanese population, the authors determined that 4419 subjects would render a power of 99% to detect a reduction in the historical incidence of carbamazepine-induced SJS-TEN from 0.25% to 0.03%.

In Taiwan, 4855 patients underwent genotyping; 7.7% (372) were found to be HLA-B*1502 positive and advised not to take carbamazepine and 215 were given alternative treatment (gabapentin, lamotrigine, naproxen, imipramine, or prednisolone). Among all 4855 subjects, 211 (4.3%) developed mild transient rash and pruritis and five were HLA-B*1502 positive. Seven patients had more severe cutaneous symptoms. Three patients had maculopapular eruption, two with hypersensitivity syndrome, and two patients were noted with urticaria. One patient with urticaria was HLA-B*1502 positive and had taken oxcarbazepine before study enrollment. No cases of SJS-TEN were noted in any subjects during the 2-month follow-up.

McCormack et al recruited subjects from either the Liverpool collaborators or the EPIGEN consortium. Subjects were divided into three categories: 1) hypersensitivity syndrome defined as the presence of rash or liver involvement within 3 months of initiation of carbamazepine treatment, accompanied by two of the following: prolonged recovery phase despite drug withdrawal, fever, or involvement of other internal organs (liver, kidney, lung, central nervous system, heart, muscle, thyroid, or lymphoid system); 2) maculopapular exanthema, defined as rash without systemic symptoms; and 3) SJS-TEN defined as skin detachment between 10-30% of body surface area with target skin lesions. Both population and clinical controls (those with epilepsy taking carbamazepine for 3 months or greater without clinical or biochemical hypersensitivity) were used. Single-nucleotide polymorphisms and imputation with high-resolution confirmation sequence-based HLA typing was performed to test for HLA allele association and clinical disease. Follow-up genotyping confirmed the HLA-A*3101 as a risk factor for the following: hypersensitivity syndrome with an odds ratio of 12.41 (27 subjects vs 257 controls without adverse drug reactions), maculopapular exanthema with an odds ratio of 8.33 (106 subjects vs 257 controls), and SJS-TEN with an odds ratio of 25.93 (12 subjects vs 257 controls). Overall, the presence of the HLA-A*3101 allele increased the risk of carbamazepine induced hypersensitivity reactions from 5% to 26%, whereas its absence reduced the risk from 5% to 3.8%.


Both of these papers contribute to the mounting evidence that various HLA alleles predispose select individuals to carbamazepine hypersensitivity reactions. As increased mortality is associated with the most severe of these reactions, SJS-TEN, prescreening individuals may be of increasing clinical importance, particularly in the case of the HLA-B*1502 allele and the associated risk reduction of SJS-TEN in Han Chinese. That said, the greatest criticism of both of these articles is the lack of definitive safety data regarding alternative treatment. It is unknown whether compounds structurally similar to carbamazepine carry similar hypersensitivity risks. In addition, further work will be needed to clarify whether other individuals of European and Asian descent carry the same risk for hypersensitivity reactions.