Before carbamazepine is prescribed, screening patients for the presence of the HLA-B*1502 and the HLA-A*3101 alleles will identify those most likely to develop potentially lethal hypersensitivity reactions, according to two separate reports in the March 24 issue of the New England Journal of Medicine.
In the first report, researchers identified 8% of 4,855 Taiwanese patients as carriers of the HLA-B*1502 allele, preventing 10 cases of Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) that would have developed if no screening had taken place.
In the second report, Northern European patients who carried the HLA-A*3101 allele were found to have a probability of 26% for developing hypersensitivity reactions to carbamazepine, while those who did not carry the allele had a probability of 3.8%.
Both reports suggest that genetic screening for these alleles before prescribing carbamazepine is now warranted, the investigators said.
Hypersensitivity reactions to carbamazepine include maculopapular exanthema (also known as the DRESS syndrome [drug reaction with eosinophilia and systemic symptoms]), the hypersensitivity syndrome, SJS, and TEN.
In the first study, Pei Chen, Ph.D., of Academia Sinica, Taipei, and associates gave prescriptions for carbamazepine during a 3-year period at 23 hospitals throughout Taiwan. The indications for carbamazepine included epilepsy (14% of subjects), neuralgia (54%), diabetic neuropathy pain (12%), tinnitus (4%), bipolar or other psychiatric disorders (3%), or miscellaneous indications (14%). (Percentages do not equal 100% due to rounding.) Before filling the prescriptions, patients provided blood samples for genotyping and waited 2-3 days for the results.
Of the 4,855 subjects, 372 (8%) were positive for HLA-B*1502 and did not take carbamazepine. Some HLA-B*1502-positive patients were given alternative therapies, while those negative for the allele were started on carbamazepine and closely monitored for 2 months for the development of adverse drug reactions.
No patients developed SJS or TEN, the most serious adverse reactions associated with carbamazepine. SJS is characterized by a rapidly progressive blistering exanthema of macular papules and target-like lesions with mucosal involvement. It has 5% mortality. TEN has a similar presentation but even more extensive detachment of the skin and leads to 25%-35% mortality.
In contrast, at least 10 cases of SJS or TEN would be expected to develop in an unscreened population of this size, the researchers said (N. Engl. J. Med. 2011;364:1126-33).
Adverse cutaneous reactions that did occur were described as "mild, localized, and transient."
A strong association between the HLA-B*1502 allele and carbamazepine hypersensitivity has been noted before in Han Chinese populations, but more recent research has also demonstrated this connection in Malay, Thai, Indian, and Japanese populations, in which the allele is particularly prevalent. "We speculate that in these countries ... HLA-B*1502 screening could provide a benefit," Dr. Chen and colleagues wrote.
In the second report, Mr. Mark McCormack of the Royal College of Surgeons in Ireland, Dublin, and his associates genotyped 22 patients of northern European ancestry who developed carbamazepine-induced hypersensitivity syndrome and compared the results with those from 2,691 healthy controls in a blood-sample database. They found a strong association between the hypersensitivity syndrome and carriage of the HLA-A*3101 allele.
In a separate genomewide association study, the researchers compared samples from 106 patients with carbamazepine-induced maculopapular exanthema and 1,553 control subjects. They found a strong association between carbamazepine-induced maculopapular exanthema and the HLA-A*3101 allele.
The investigators then searched for the HLA-A*3101 allele in samples from 12 patients who had carbamazepine-induced SJS or TEN. Five (42%) of the affected subjects carried the allele, compared with 10 (4%) of 257 control subjects.
On this basis, "we calculated that the presence of HLA-A*3101 had a sensitivity of 26% and a specificity of 96% as a predictor of carbamazepine-associated hypersensitivity" in patients of northern European descent, Mr. McCormack and his colleagues wrote (N. Engl. J. Med. 2011;364:1134-43).
"Although the presence of the HLA-A*3101 allele is neither necessary nor sufficient for the development of hypersensitivity to carbamazepine, it is associated with a significantly increased risk," they wrote.
The prevalence of this allele is estimated to be 2%-5% in northern European populations, 2% in Han Chinese populations, and 9% in Japanese populations, they noted.
Assuming that hypersensitivity to carbamazepine in the general U.K. population is approximately 5%, "the presence of the HLA-A*3101 allele increases the [absolute] risk of hypersensitivity to 26%, whereas its absence reduces the [absolute] risk to 3.8%."
Given these calculations, 83 European patients would need to be screened to prevent 1 case of carbamazepine hypersensitivity, while only 56 Japanese patients would need to be screened to prevent 1 case.
Dr. McCormack and his associates suggested that "consideration be given to adding the association with HLA-A*3101 to the drug label for carbamazepine."