HLA-B*15:02 Genotyping and SJS Incidence

SUPPLEMENTALDATA

APPENDIX e-1

HLA-B*15:02 Genotyping and SJS Incidence

The presence or absence of HLA-B*15:02 in an individual is determined by sequence based typing of the HLA-B locus, undertaken in accredited pathology laboratories in Hong Kong’s two university hospitals. For the purposes of our study, the test was considered to have 100% sensitivity and 100% specificity for HLA-B*15:02.e1 The incidences of carbamazepine-induced Steven-Johnsons syndrome (CBZ-SJS) and phenytoin-induced Steven-Johnsons syndrome (PHT-SJS) were stratified by HLA-B*15:02 status. Since some Steven-Johnsons syndrome (SJS) or toxic epidermal necrolysis (TEN) patients were selectively tested for the biomarker in the study period, the incidences of CBZ-SJS and PHT-SJS associated with HLA-B*15:02 were derived from: 1) the observed prevalence of the allele in patients who were tested without or before the development of antiepileptic drugs-induced SJS (AED-SJS), 2) the overall incidences of CBZ-SJS and PHT-SJS in the pre-policy period, and 3) the published relative risks of CBZ-SJS and PHT-SJS associated with HLA-B*15:02 in the Hong Kong population. In our previous study, we reported the prevalence of HLA-B*15:02 in the Hong Kong population to be 16.8%, and we observed that 1.18% carbamazepine (CBZ) recipients and 0.26% phenytoin (PHT) recipients in the newly diagnosed and treated epilepsy patients had developed SJS/TEN in the pre-policy period.e2 A study in Hong Kong reported odds ratios (ORs) for CBZ-SJS and PHT-SJS associated with HLA-B*15:02 to be 89.25 [95% confidence interval (CI): 19.25 – 413.83] and 3.50 [95% CI: 1.10 – 11.18], respectively.e3 Since SJS/TEN rarely occurs in non-carriers,e3-e7 these ORs were close approximations of the relative risks. Hence, we estimated 6.69% [95% CI: 5.61% – 6.97%] of the HLA-B*15:02 carriers and 0.07% [95% CI: 0.06% – 0.08%] of the non-carriers would develop SJS/TEN when exposed to CBZ. The incidences for PHT-SJS were 0.65% [95% CI: 0.28% – 1.09%] for HLA-B*15:02 carriers and 0.19% [95% CI: 0.08% – 0.31%] for non-carriers.

Costs

The annual treatment cost of each antiepileptic drug (AED) was calculated based on the tablet costs and the corresponding defined daily dose, which is the assumed average maintenance dose per day.e8 Weighted average annual treatment costs of non-CBZ AEDs were estimated by using prescription weights of each non-CBZ AED among people with newly diagnosed epilepsy. Since the ratio of each non-CBZ AED changed proportionally to the change of CBZ prescription rate, varying CBZ prescription rate affected the weight average annual treatment cost of non-CBZ AEDs. To simulate this effect during sensitivity analyses, a log-linear regression model was established based on the observed annual new AED prescriptions and estimated annual treatment cost of each non-CBZ AED. It was estimated that the weighted average annual treatment cost of non-CBZ AEDs was approximately [R2=0.95] where denoted the prescription rate of CBZ.

SJS/TEN treatment costs were estimated from the average length of stay in the intensive care units, high-dependency units, burn units and general wards. SJS mortality rates were derived directly from observed data. Published costs were US$192 per HLA-B*15:02 test,e1 and US$142 per specialist consultation in the Hospital Authority.e9

All treatment costs were derived from the expenditures reported by HA, if applicable. Since the HA provided subsidized healthcare to Hong Kong residents, the costs were considered to cover all consumer’s costs.e10All costs were converted from Hong Kong to US dollars using the fixed exchange rate of HK$7.8 to US$1.e11

e-REFERENCES

e1.Ng M, Ng H. Re: Genetic testing prior to prescription of carbamazepine [online]. Available at: Accessed November 26, 2014.

e2.Chen Z, Liew D, Kwan P. Effects of a HLA-B*15:02 screening policy on antiepileptic drug use and severe skin reactions. Neurology 2014;83:2077-2084.

e3.Cheung YK, Cheng SH, Chan EJ, Lo SV, Ng MH, Kwan P. HLA-B alleles associated with severe cutaneous reactions to antiepileptic drugs in Han Chinese. Epilepsia 2013;54:1307-1314.

e4.Hung SI, Chung WH, Jee SH, et al. Genetic susceptibility to carbamazepine-induced cutaneous adverse drug reactions. Pharmacogenet Genomics 2006;16:297-306.

e5.Man CB, Kwan P, Baum L, et al. Association between HLA-B*1502 allele and antiepileptic drug-induced cutaneous reactions in Han Chinese. Epilepsia 2007;48:1015-1018.

e6.Chung WH, Hung SI, Hong HS, et al. Medical genetics: a marker for Stevens-Johnson syndrome. Nature 2004;428:486.

e7.Locharernkul C, Shotelersuk V, Hirankarn N. Pharmacogenetic screening of carbamazepine-induced severe cutaneous allergic reactions. J Clin Neurosci 2011;18:1289-1294.

e8.World Health Organization Collaborating Centre for Drug Statistics Methodology. Definition and general considerations [online]. Available at: Accessed November 26, 2014.

e9.Hong Kong Hospital Authority. Hospital Authority Ordinance (Chapter 113): Revisions to List of Charges [online]. Available at: Accessed November 26, 2014.

e10.GovHK. Overview of the Health Care System in Hong Kong [online]. Available at: Accessed March 20, 2015.

e11.Hong Kong Monetary Authority. Linked Exchange Rate System [online]. Available at: Accessed November 26, 2014.

SUPPLEMENTAL TABLE

Table e-1. Variables input, sensitivity range and distribution for cost-effectiveness analysis of current HLA-B*15:02 screening policy
Variable / Input Value* / Range for Sensitivity Analysis / Distribution
HLA-B*15:02 prevalence / 0.168 / 0.084 - 0.336 / Beta
SJS/TEN Incidence
CBZ-SJS in HLA-B*15:02 positive patients / 0.0669 / 0.0561 - 0.0697 / Beta
CBZ-SJS in HLA-B*15:02 negative patients / 0.0007 / 0.0006 – 0.0008 / Beta
Non-CBZ AED-SJS without screening policy / 0.0026 / 0.002 - 0.0032 / Beta
Non-CBZ AED-SJS with screening policy / 0.0025 / 0.0019 - 0.0031 / Beta
Mortality in SJS/TEN / 0.075 / 0.0375 - 0.15 / Beta
Costs (US$)
Average annual CBZ treatment cost (in both periods) / 278 / 139 - 556 / Gamma
Weighted average annual other AED treatment cost
Without the screening policy / 105 / 50 - 210 / Gamma
With the screening policy / 161 / 80 - 320 / Gamma
HLA-B*15:02 genotyping cost / 192 / 0 - 384 / Gamma
Specialist consultation cost / 142 / 71 - 284 / Gamma
Weighted average SJS/TEN treatment cost / 10,110 / 5,055 – 20,220 / Gamma
Utility/QoL
Epilepsy (60% seizure control) / 0.829 / 0.746 - 1 / Beta
SJS/TEN / 0.373 / 0.25 - 0.5 / Beta
AED prescription behaviour
CBZ recipients among newly treated patients without screening policy / 0.1221 / 0.0611 - 0.2442 / Beta
With screening policy
CBZ recipients among HLA-B*15:02 negative patients / 0.6095 / 0.3047 - 1 / Beta
CBZ recipients among HLA-B*15:02 positive patients / 0 / 0 - 0.05 / Beta
CBZ recipients among patients do not wait for test result / 0.0186 / 0 - 0.0372 / Beta
CBZ percipients among patients do not perform the test / 0.0177 / 0 - 0.035 / Beta
Test-ordering behaviour
HLA-B*15:02 test rate / 0.0625 / 0 – 0.125 / Beta
Proportion of physician waited for test result before prescribing an AED / 0.4905 / 0.2452 - 1 / Beta
Pre-emptive testing rate / 0.0343 / 0 - 0.0686 / Beta
Repeat HLA-B*15:02 test rate / 0.0183 / 0 - 0.0366 / Beta
Non-same day test result rate / 0.9863 / 0 - 1 / Beta
* Utilities were estimated from the literature. Other input variables were derived from our previous observations.e2 See text for details.
AED, antiepileptic drug; CBZ, carbamazepine; SJS, Stevens-Johnson syndrome; TEN, toxic epidermal necrolysis

SUPPLEMENTAL FIGURES

Figure e-1. Cost-effectiveness acceptability curves of no screening and HLA-B*15:02 screening policy strategies for patient with newly diagnosed epilepsy in Hong Kong.

Figure Legend: Red line with triangles and blue line with squares represent the percentage of iterations would fall below certain incremental cost-effectiveness ratios for no HLA-B*15:02 screening policy and the HLA-B*15:02 screening policy in current situation, respectively.

Figure e-2. Cost-effectiveness acceptability curves of no screening and HLA-B*15:02 screening policy in ideal situation strategies for patient with newly diagnosed epilepsy in Hong Kong.

Figure e-3. Cost-effectiveness acceptability curves of no screening and extended HLA-B*15:02 screening policy in ideal situation strategies for patient with newly diagnosed epilepsy in Hong Kong.

Figure Legend: Red line with triangles and blue line with squares represent the percentage of iterations would fall below certain incremental cost-effectiveness ratios for no HLA-B*15:02 screening policy and the extended HLA-B*15:02 screening policy in ideal situation, respectively.

Figure e-4. Two-way sensitivity analysis of HLA-B*15:02 genotyping cost and incidence of phenytoin induced Stevens-Johnson syndrome in HLA-B*15:02 carriers of the cost-effectiveness analysis of the extended HLA-B*15:02 screening policy in the ideal situation.

Figure Legend: Area in red represents the incremental cost-effectiveness ratio of the extended screening policy is greater than US$50,000/QALY for the corresponding combinations of the genotyping cost and the incidence rate. Area in blue represents the extended screening policy is cost-effective when the corresponding combinations of the two variables are achieved.