29 Aug 2013: Recommendations for HLA-B*1502 genotype testing prior to initiation of carbamazepine in new patients
The Ministry of Health (MOH) has announced that genotyping for the HLA-B*1502 allele prior to the initiation of carbamazepine (CBZ) therapy in new patients of Asian ancestry is now considered the standard of care. These new recommendations by MOH and HSA have been made in consultation with experts in various fields such as neurology, psychiatry and dermatology, following the review of findings from local and international studies.
CBZ has been registered in Singapore since 1988 and is currently available as Tegretol®(Novartis (Singapore) Pte Ltd) and six generic products. It is indicated for the treatment of epilepsy and other conditions such as diabetic neuropathy, trigeminal neuralgia and bipolar disorders. While CBZ is an effective drug and the drug of choice for several conditions, Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN), which are associated with significant mortality and long-term morbidity, have been reported with its use. Between 2003 and 2012, HSA received 131 local serious reports of CBZ-induced SJS/TEN (average of 15 reports per year). Since the beginning of this year, HSA has received five reports of SJS/TEN associated with the use of CBZ. A one-time HLA-B*1502 genotyping test helps distinguish high-risk patients who should avoid CBZ from low-risk patients who are able to continue to use this low-cost yet effective medicine.
SJS and TEN often begin with flu-like symptoms (fever, sore throat and fatigue), followed by development of a red or purplish rash and painful ulcers of mucous membranes.The skin lesions then progress to epidermal necrosis and detachment. SJS is diagnosed when epidermal detachment involves <10% of the body surface area, TEN when >30% of the body surface epidermal detachment is involved and SJS/TEN when 10-30% of the body surface epidermal detachment is involved. These conditions require hospitalisation and can be life-threatening and even fatal.
Summary of evidence to-date
The basis for these new local recommendations came from strong local and international data supporting an association between HLA-B*1502 and CBZ-induced SJS/TEN:
- A HSA-initiated multi-centre study at Changi General Hospital, Singapore General Hospital and National University Hospital found a strong association between the HLA-B*1502 allele and CBZ-induced SJS/TEN (Odds ratio [OR]=181, p<0.0001). The results are consistent with international data that HLA-B*1502 carriers have an elevated risk of developing SJS/TEN when taking CBZ.
- It is cost-effective to genotype all newly diagnosed epilepsy patients in Singapore and prescribe CBZ to those who are tested negative for HLA-B*1502 as compared to prescribing or avoiding usage of CBZ without knowledge of genotype. This was demonstrated in a study by the Duke-NUS Graduate Medical School, in collaboration with HSA.1
- Published studies have documented a strong association between the carriage of HLA-B*1502 allele and risk of CBZ-induced SJS/TEN among Han Chinese in Taiwan and Hong Kong.2,3 Subsequent smaller scale studies have shown that other Asian ethnic groups such as Malays, Indians and Thais have similar associations.4-6
- A large prospective study in Taiwan (n=4,877) found HLA-B*1502 screening prior to the initiation of CBZ therapy to be successful in completely preventing CBZ-induced SJS/TEN.7
HSA, together with MOH, has issued a Dear Healthcare Professional Letter on 30 April 2013 to communicate the new recommendations. Currently, HSA is working with the product licence holders of CBZ products to strengthen the package inserts to reflect the new recommendation that testing for the presence of HLA-B*1502 allele is highly recommended in new patients with Asian ancestry.
In view of the new recommendation for CBZ, healthcare professionals are advised to take note of the following:
- CBZ should not be prescribed prior to the return of HLA-B*1502 test results. This is because of the possibility of development and progression of SJS/TEN in susceptible patients even after prompt discontinuation of the drug.
- The use of CBZ should be avoided and treatment alternatives are recommended in patients who are found to be positive for HLA-B*1502. Preliminary data have shown a suspected association between this allele and phenytoin-induced SJS/TEN, although the effect size is not as large.4,8 As a precaution, patients who are identified to be positive for the HLA-B*1502 allele should also not be prescribed phenytoin.
- Genetic testing should not substitute for appropriate clinical vigilance and patient management. Although rare, patients negative for HLA-B*1502 could still develop SJS/TEN as the role of other factors such as drug dose, concomitant medications and co-morbidities have not been studied.
Where to test
The HLA-B*1502 test is available at the NUH Molecular Diagnosis Centre (MDC). Each test costs $187 (excluding GST) and the estimated turnaround time is two to four working days (see below). Subsidised patients from the MOH-funded restructured hospitals and institutions will qualify for a flat rate subsidy of 75% of the cost of the test. Clinicians may send their patients' samples directly to NUH MDC via the NUH Referral Laboratories (NRL), Tel: 6778 5171. For this test, 3 to 5ml of blood is to be collected in a 5-ml EDTA-anti-coagulated sample tube and kept at 4°C before despatch.
Operating hours of NUH MDC
Monday to Friday: 8.30am – 5.30pm Saturday : 8.30am – 12.30pm Sunday and public holidays: Closed
Tel: 6772 4384; 6772 4175
|Cut-off time for arrival of test sample at NUH (Note: Specimens received after cut-off time will be subjected to the next batch of testing)||Corresponding latest despatch time of results|
|Monday, 3pm; Thursday, 3pm||Tuesday, 6.30pm; Friday, 6.30pm|
- Neurology 2012; 79: 1259-67
- Epilepsia 2007; 48: 1015-8
- Pharmacogenet Genom 2006; 16: 297-306
- Epilepsia 2008; 49: 2087-91
- Indian J Dermatol Venereol Leprol 2009; 75: 579-82
- Asian Pac J Allergy Immunol 2011; 29: 290-3
- N Engl J Med 2011; 364: 1126-33
- Pharmacogenomics 2010; 11: 349-56