HLA-B*5801: utility and cost-effectiveness in the Asia-Pacific Region
Correspondence: Dr Siaw Ing Yeo, Clinical Associate, Division of Rheumatology and Clinical Immunology, University Department of Medicine, Room 807, Administration Block, Queen Mary Hospital, Pokfulam Road, Hong Kong, China.
Gout is a common condition which is mainly treated with the hypo-uricemic agent, allopurinol. Although allopurinol is generally a well-tolerated drug, there is a small risk of developing potentially fatal complications, such as allopurinol hypersensitivity syndrome. Recent advances in pharmacogenomics have made possible the identification of genes which confer susceptibility to specific drugs. A recent multi-national case-control study has reported allopurinol as the most common drug associated with Stevens-Johnson syndrome and toxic epidermal necrolysis. Several studies have established a strong association between the human leukocyte antigen (HLA)-B*5801 gene and development of Stevens-Johnson syndrome and toxic epidermal necrolysis. The allele frequency of HLA-B*5801 is highest in the South East Asian population.Since other hypo-uricemic agents are available, patients may wish to have HLA-B*5801 testing before being started on allopurinol. As the test for HLA-B*5801 is expensive, time-consuming and only available in selected laboratories, there is a need to evaluate the utility and cost-effectiveness of this test in our region.
Gout is a monosodium urate crystal deposition disease with a male preponderance. It is a relatively common condition and its incidence has been increasing, largely due to changes in dietary choices. Zeng et al. reported the prevalence of gout at between 0.15% and 1.98% in China, with the highest prevalence of 11.7% in Taiwanese aborigines. The aims of treatment in gout are reduction and maintenance of serum uric acid levels to below a critical value which allows dissolution of the crystals, and elimination of the uric acid crystals, respectively.
Allopurinol, a xanthine oxidase inhibitor, is the most frequently used drug for the long-term treatment of gout. It is generally well-tolerated, although up to 2% of patients taking allopurinol develop a mild rash, and about 5% discontinue this drug because of another adverse event. However, allopurinol may also cause the rare and potentially fatal, allopurinol hypersensitivity syndrome (AHS), which presents with rash (e.g. Stevens-Johnson syndrome [SJS] or toxic epidermal necrolysis [TEN]), fever, eosinophilia, leukocytosis, hepatitis and renal failure. The mortality rate associated with AHS is as high as 27%.[3, 4] Allopurinol withdrawal and supportive care are the mainstays of treatment. A recent multinational case-control study reported that allopurinol was the most common drug associated with SJS and TEN.
The frequency of AHS has previously been reported to occur at 1:260 (0.4%) in patients treated with allopurinol, and the mortality associated with AHS is said to be much higher than hypersensitivity reactions associated with other drugs. Risk factors for developing AHS include female sex, older age, renal impairment, diuretic use and recent initiation of allopurinol treatment. Criteria for the diagnosis of AHS were suggested by Singer and Wallace and are listed in Table 1.
Table 1. Suggested criteria for the diagnosis of allopurinol hypersensitivity syndrome, according to Singer and Wallace
|A documented intake of allopurinol|
|Lack of exposure to a different drug causing a similar clinical picture|
|Presence of at least two major criteria or one major and one minor criterion|
|Major criteria include:|
|Worsening renal function|
|Acute hepatocellular injury|
|Rash, manifested by toxic epidermal necrolysis, erythema multiforme, diffuse maculopapular rash or exfoliative dermatitis|
|Minor criteria include: fever, leukocytosis and eosinophilia|
The Role of Genomics
Recent advances in genomic research have made possible the identification of genes which confer susceptibility to severe cutaneous adverse drug reactions that are specific to drug, phenotype and ethnicity. Various studies have identified a strong association between human leukocyte antigen (HLA)-B*5801 and the development of SJS/TEN in allopurinol users. The allele frequency of HLA-B*5801 has been reported to be as high as 6–8% among Southeast Asian populations, and <1% among Western European populations, respectively.[7, 8]
In their study, Hung et al. used a case-control association study in the Han Chinese in Taiwan, to identify genetic markers for allopurinol-induced severe cutaneous adverse reactions (allopurinol-SCAR). Allopurinol-SCAR included the drug hypersensitivity syndrome, SJS and TEN. They used two groups of controls, the first being 135 patients who had been on allopurinol for at least 6 months without adverse events, and a second control group of 93 subjects from the general population. They found that all 51 patients (100%) with allopurinol-SCAR carried the HLA-B*5801 gene. The presence of chronic renal insufficiency also increased the risk of developing allopurinol-SCAR (odds ratio 4.7; confidence interval [CI] 2.3–9.3).
Tassaneeyakul and Kaniwa found a strong association between HLA-B*5801 and allopurinol-related SJS and TEN in Thai (100%) and Japanese (4/5) patients, respectively. Kang studied 25 Korean patients with allopurinol-SCAR and found a HLA-B*5801 frequency of 92% in these patients versus 10.5% in controls. Furthermore, Jung discovered that the incidence of allopurinol-SCAR in Korean patients with chronic renal insufficiency was considerably higher if they also carried the HLA-B*5801 gene. In fact, the association between HLA-B*5801 allele and allopurinol-SCAR has been found to be consistent across different populations, both Asian and non-Asian.
A major limitation of individual studies arises from the low incidence of allopurinol-SCAR, which results in observational studies with small sample sizes and insufficient power. Somkrua and colleagues performed a systematic review and meta-analysis in order to accumulate and quantitatively analyze the genetic association between HLA-B*5801 and allopurinol-induced SJS/TEN as well as to elucidate any between-study heterogeneity. They analyzed four studies which included 55 SJS/TEN cases and 678 matched controls (allopurinol-tolerant control) and five studies with 69 SJS/TEN cases and 3378 population controls (general population). They concluded that allopurinol users with HLA-B*5801 have a 80–97 times increased risk of developing SJS/TEN compared to those who do not have this gene. Furthermore, sensitivity analyses suggested that the summary odds ratios remained significant regardless of populations, thus highlighting the potential of genotyping in different populations.
The pathogenesis of AHS is likely to represent the interplay between different factors, mainly immunological and genetic, and with the drug and accumulation of its metabolite (oxypurinol). It has been postulated that the increased risk of developing SJS/TEN in subjects with HLA-B*5801 may be related to the amplification of cytotoxic T-cells, and the direct involvement of HLA genes, which present an antigen to the T-cell receptor. Thus far, there is no definite proven mechanism, but Sumkrua and Hung have individually made the observation that although HLA-B*5801 plays a central role in allopurinol-related SJS/TEN, it may not be the only factor required for the occurrence of these severe conditions. In addition, most studies have examined the association of allopurinol-related SJS/TEN with HLA-B*5801, so that the interpretation of findings from these studies may only be limited to SJS/TEN cases. However, it should be noted that a number of studies have also reported the potential association of DRESS (drug rash with eosinophilia and systemic symptoms) and HLA-B*5801.[9, 17]
Limitations of HLA-B 5801 Testing
The implication of the strong association of HLA-B*5801 with allopurinol hypersensitivity syndrome is more likely to be significant in populations with a high prevalence of HLA-B*5801. Hence, genotype testing may be of benefit to high-risk groups, for example Asians, before treatment with allopurinol. However, this test is only available in selected laboratories (e.g. laboratories affiliated to transplant centers), is time-consuming with turnaround times around 3–4 weeks, and may be costly. Formal assessment of the cost-effectiveness of routine HLA-B*5801 testing is not available. Somkrua studied the range of genotyping costs that would be cost-effective from the healthcare provider perspective and found that the most influential parameters were the cost of genotyping and SJS/TEN management. Pharmacogenetic screening seemed to be cost-effective if the cost fell in the range of 393–1085 THB (US$13–35).
In their attempt to expedite the application of pharmacogenomic information to proper use of allopurinol in the clinical situation, Maekawa and his group have developed a polymerase chain reaction-restriction fragment length polymorphism (PCR-RLP) assay which is based on their demonstration that several single nucleotide polymorphisms (SNPs) around the HLA region on chromosome 6 were strongly linked with HLA-B*5801. They claim that their use of this surrogate biomarker for carriers of HLA-B*5801 is a robust and inexpensive assay for the screening of subjects prior to starting allopurinol treatment.
On the other hand, Lee et al. have commented that utilization of HLA-B*5801 alone as a population screening test even in a Southeast Asian population is not effective. They argue that although the negative predictive value and sensitivity of HLA-B*5801 in cases of allopurinol-induced SJS/TEN are very high, the positive predictive value (of the screening test) of developing allopurinol hypersensitivity is low because of the very low incidence of allopurinol hypersensitivity itself and the relatively high prevalence of HLA-B*5801 in the Southeast Asian population. They add that separate risk factors, such as other genetic markers, concomitant diuretics, renal impairment and patch-testing, need to be established in order to improve the positive predictive value of screening.
Nevertheless, AHS is a potentially fatal condition which may be preventable. Although the positive predictive value of HLA-B*5801 is low, the test may be useful in patients with Asian ethnic background. Since other hypo-uricemic drugs such as probenecid and febuxostat are available, patients may not wish to take the risk (albeit small) of a serious drug reaction to allopurinol. The option of having this test (on a self-financed basis) should be made available to patients if routine screening has not been or cannot be implemented.
However, it should be stressed that having the HLA-B*5801 test does not result in absolutely no risk of allopurinol-related SJS/TEN. Monitoring for signs and symptoms is still necessary. Other mitigating factors include only prescribing allopurinol for treatment of hyper-uricemia in symptomatic conditions such as gout, urate nephrolithiasis and nephropathy and when cytolytic therapy is considered. Recently, a study by Stamp has shown that the starting dose of allopurinol is an important risk factor for development of AHS. The study suggests a starting dose of 1.5 mg per unit of estimated glomerular filtration rate, with progressive up-titration of the dose to achieve the target serum uric acid level.
Further evaluation of the cost-effectiveness of HLA-B*5801 testing in a population setting should be carried out. This may lead to the development of guidelines which can assist prescribing physicians and ensure that a uniform approach is adopted when the question about genotype testing arises in clinical practice.