If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
11 These authors contributed equally to this work.
Chau Yee Ng
Footnotes
11 These authors contributed equally to this work.
Affiliations
Department of Dermatology, Drug Hypersensitivity Clinical and Research Center, Chang Gung Memorial Hospital, Taipei, Linko and Keelung, TaiwanSchool of Medicine, College of Medicine, Chang Gung University, Taoyuan, Taiwan
11 These authors contributed equally to this work.
Yu-Ting Yeh
Footnotes
11 These authors contributed equally to this work.
Affiliations
Department of Dermatology, Drug Hypersensitivity Clinical and Research Center, Chang Gung Memorial Hospital, Taipei, Linko and Keelung, TaiwanSchool of Medicine, College of Medicine, Chang Gung University, Taoyuan, Taiwan
11 These authors contributed equally to this work.
Chuang-Wei Wang
Footnotes
11 These authors contributed equally to this work.
Affiliations
Department of Dermatology, Drug Hypersensitivity Clinical and Research Center, Chang Gung Memorial Hospital, Taipei, Linko and Keelung, TaiwanChang Gung Immunology Consortium, Chang Gung Memorial Hospital and Chang Gung University, Taoyuan, Taiwan
Department of Dermatology, Drug Hypersensitivity Clinical and Research Center, Chang Gung Memorial Hospital, Taipei, Linko and Keelung, TaiwanSchool of Medicine, College of Medicine, Chang Gung University, Taoyuan, Taiwan
Department of Dermatology, Drug Hypersensitivity Clinical and Research Center, Chang Gung Memorial Hospital, Taipei, Linko and Keelung, TaiwanSchool of Medicine, College of Medicine, Chang Gung University, Taoyuan, Taiwan
Graduate Institute of Clinical Medical Science, Clinical Informatics and Medical Statistics Research Center, Chang Gung University, Taoyuan, TaiwanBiostatistical Center for Clinical Research, Chang Gung Memorial Hospital, Linkou, Taiwan
Graduate Institute of Clinical Medical Science, Clinical Informatics and Medical Statistics Research Center, Chang Gung University, Taoyuan, TaiwanBiostatistical Center for Clinical Research, Chang Gung Memorial Hospital, Linkou, Taiwan
Department of Dermatology, Drug Hypersensitivity Clinical and Research Center, Chang Gung Memorial Hospital, Taipei, Linko and Keelung, TaiwanWhole-Genome Research Core Laboratory of Human Diseases, Chang Gung Memorial Hospital, Keelung, Taiwan
School of Medicine, College of Medicine, Chang Gung University, Taoyuan, TaiwanDivision of Allergy, Immunology and Rheumatology, Department of Medicine, Chang Gung Memorial Hospital, Keelung, Taiwan
12 These authors contributed equally to this work.
Rosaline Chung-Yee Hui
Correspondence
for correspondence regarding clinical analysis and meta-analysis: Rosaline Chung-Yee Hui, MD, PhD, Department of Dermatology, Chang Gung Memorial Hospital, Taipei, 199, Tung-Hwa North Road, Taipei 105, Taiwan.
12 These authors contributed equally to this work.
Affiliations
Department of Dermatology, Drug Hypersensitivity Clinical and Research Center, Chang Gung Memorial Hospital, Taipei, Linko and Keelung, TaiwanSchool of Medicine, College of Medicine, Chang Gung University, Taoyuan, Taiwan
12 These authors contributed equally to this work.
Wen-Hung Chung
Correspondence
Correspondence: Wen-Hung Chung, MD, PhD, Department of Dermatology, Chang Gung Memorial Hospital, Taipei, 199, Tung-Hwa North Road, Taipei 105, Taiwan.
12 These authors contributed equally to this work.
Affiliations
Department of Dermatology, Drug Hypersensitivity Clinical and Research Center, Chang Gung Memorial Hospital, Taipei, Linko and Keelung, TaiwanSchool of Medicine, College of Medicine, Chang Gung University, Taoyuan, TaiwanChang Gung Immunology Consortium, Chang Gung Memorial Hospital and Chang Gung University, Taoyuan, TaiwanWhole-Genome Research Core Laboratory of Human Diseases, Chang Gung Memorial Hospital, Keelung, Taiwan
Allopurinol, a common drug for treating hyperuricemia, is associated with cutaneous adverse drug reactions ranging from mild maculopapular exanthema to life-threatening severe cutaneous adverse reactions, including drug reaction with eosinophilia and systemic symptoms, Stevens-Johnson syndrome, and toxic epidermal necrolysis. We have previously reported that HLA-B*58:01 is strongly associated with allopurinol-induced severe cutaneous adverse reactions in Han Chinese, but the associations of the HLA-B*58:01 genotype in an allopurinol-induced hypersensitivity phenotype remain unclear. To investigate the comprehensive associations of HLA-B*58:01, we enrolled 146 patients with allopurinol-induced cutaneous adverse drug reactions (severe cutaneous adverse reactions, n = 106; maculopapular exanthema, n = 40) and 285 allopurinol-tolerant control subjects. Among these allopurinol-induced cutaneous adverse drug reactions, HLA-B*58:01 was strongly associated with severe cutaneous adverse reactions (odds ratio [OR] = 44.0; 95% confidence interval = 21.5–90.3; P = 2.6 × 10-41), and the association was correlated with disease severity (OR = 44.0 for severe cutaneous adverse reactions, OR = 8.5 for maculopapular exanthema). The gene dosage effect of HLA-B*58:01 also influenced the development of allopurinol-induced cutaneous adverse drug reactions (OR = 15.25 for HLA-B*58:01 heterozygotes and OR = 72.45 for homozygotes). Furthermore, coexistence of HLA-B*58:01 and renal impairment increased the risk and predictive accuracy of allopurinol-induced cutaneous adverse drug reactions (heterozygous HLA-B*58:01 and normal renal function: OR = 15.25, specificity = 82%; homozygous HLA-B*58:01 and severe renal impairment: OR = 1269.45, specificity = 100%). This HLA-B*58:01 correlation study suggests that patients with coexisting HLA-B*58:01 and renal impairment (especially estimated glomerular filtration rate < 30ml/minute/1.73 m2) should be cautious and avoid using allopurinol.
EULAR evidence based recommendations for gout. Part II: Management. Report of a task force of the EULAR Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT).
). However, allopurinol administration is associated with cutaneous adverse reactions (cADRs) ranging from mild skin rash, such as maculopapular exanthema (MPE), to potentially life-threatening severe cutaneous adverse drug reactions (SCARs), which include drug reaction with eosinophilia and systemic symptoms (DRESS) (also known as drug-induced hypersensitivity syndrome), Stevens-Johnson syndrome (SJS), and toxic epidermal necrolysis (TEN).
We previously reported a strong association of HLA-B*58:01 with allopurinol-induced SCARs in Han Chinese (
Association of HLA-B*58:01 allele and allopurinol-induced Stevens Johnson syndrome and toxic epidermal necrolysis: a systematic review and meta-analysis.
). Although allopurinol has been widely used for over 30 years, it can induce severe adverse events, particularly in patients with renal insufficiency (
, creatinine clearance-based adjustment for allopurinol has been advocated by many clinicians and was adopted in allopurinol usage guidelines. With the increasing use of pharmacogenomic biomarkers to prevent adverse drug reactions, a comprehensive study of the impact of HLA-B*58:01 and other nongenetic factors is warranted. Although several reports have shown the importance of human leukocyte antigen (HLA) allele (
), the correlation of these risk factors has not been studied. Furthermore, many previous studies were limited by low incidences, difficulties with patient enrollment, and small sample sizes, which complicated the investigation of correlation. Herein, we enrolled a large number of patients with allopurinol-induced cADRs and allopurinol-tolerant control subjects to investigate the associations of HLA-B*58:01, renal function, gene dosage, and drug dosage with the risk of allopurinol-induced cADR development.
Results
Clinical characteristics of study subjects
A total of 146 patients with allopurinol-induced cADRs and 285 allopurinol-tolerant control subjects were recruited. Table 1 shows the clinical characteristics, demographic data, HLA-B*58:01 allele carrier rate, and renal function. Among the 146 patients with allopurinol-induced cADRs, 21% (n = 32) had SJS with less than 10% body surface area (BSA) detachment, 9% (n = 14) had SJS/TEN with 10% or greater BSA detachment, 38% (n = 57) had DRESS, 2% (n = 3) had DRESS with overlapping SJS/TEN, and 27% (n = 40) had MPE. Mean age was higher for patients with cADRs than for control subjects (64 years for cADR subjects and 59 years for control subjects; P < 0.01). The proportion of women was higher among patients with allopurinol-induced cADRs than among control subjects (odds ratio [OR] = 4.71; 95% confidence interval [CI] = [2.36, 9.70]; P < 0.01; Table 2). The mortality rate was 8% (n = 12; Table 1) among patients with allopurinol-induced cADRs, and all deaths occurred among patients with allopurinol-induced SCARs. The mortality rate was highest for subjects with DRESS overlapping SJS/TEN (67%; n = 2), followed by those with SJS/TEN with greater than 10% BSA detachment (43%; n = 6). The renal function was worse in patients with allopurinol-induced cADRs (estimated mean glomerular filtration rate [eGFR] ± standard deviation = 31 ± 26 ml/minute/1.73 m2, range = 3–134 ml/minute/1.73 m2) than in control subjects (mean eGFR ± SD = 63 ml/minute/1.73 m2, range = 3–151 ml/minute/1.73 m2) (P < 0.01).
Table 1Clinical characteristics of 146 patients with allopurinol-induced cADRs and 285 allopurinol-tolerant control subjects
Phenotypes
cADRs
SCARs
SJS (BSA < 10%)
SJS/TEN (BSA ≥ 10%)
DRESS
DRESS Overlap with SJS/TEN
MPE
Control Subjects
Total number
146
106
32
14
57
3
40
285
Age, years
Mean (SD)
65 (16)
64 (16)
63 (18)
60 (17)
64 (15)
77 (8)
69 (14)
59 (14)
Range
18–88
8–98
23–86
32–91
18–98
72–86
27–92
16–87
Sex
Male, n (%)
80 (55)
52 (49)
18 (56)
4 (29)
27 (47)
2 (67)
28 (70)
261 (92)
Female, n (%)
66 (45)
54 (51)
14 (44)
10 (71)
30 (52)
1 (33)
12 (30)
17 (8)
Mortality, n (%)
12 (8)
12 (11)
2 (6)
6 (43)
2 (4)
2 (67)
0 (0)
0 (0)
Baseline eGFR, mL/minute/1.73 m2
Median
31
31
43
33
24
14
31
63
Range
3–134
3–134
3–134
5–114
5–86
7–59
5–85
3–151
HLA-B*58:01 carrier, n (%)
123 (83)
96 (91)
28 (88)
13 (93)
52 (91)
3 (100)
26 (65)
51 (18)
Abbreviations: CADR, cutaneous adverse reactions ranging from mild skin rash; DRESS, drug rash with eosinophilia and systemic symptoms; MPE, maculopapular exanthema; SCAR, severe cutaneous adverse drug reactions; SJS, Stevens-Johnson syndrome; TEN, toxic epidermal necrolysis.
Correlation of HLA-B*58:01 genotype and allopurinol-induced cADR phenotype
The HLA-B*58:01 allele was present in 83% (n = 122) of patients with allopurinol-induced cADRs (OR = 23.32, 95% CI = [13.69, 39.7], P = 2.2 × 10-40) and 91% (n = 96) of patients with allopurinol-induced SCARs (OR = 44, 95% CI = [21.5, 90.3], P = 2.6 × 10-41) but in only 18% (n = 51) of control subjects, which demonstrates positivity for the HLA-B*58:01 allele (Figure 1). Subgroup analysis of the phenotypes of patients with cADRs shows a strong association between HLA-B*58:01 and allopurinol-induced SCARs, irrespective of the cADR phenotype: positivity for HLA-B*58:01 was noted in 91% of patients with SJS with less than 10% BSA detachment (OR = 32.1, 95% CI = [10.8, 95.6], P=3.1 × 10-15), in 93% with SJS/TEN with greater than 10% BSA detachment (OR = 59.6, 95% CI = [7.6, 466.3], P = 8.2 × 10-9), in 91% with DRESS (OR = 47.7, 95% CI = [18.2, 125.4], P = 1.0 × 10-26), and in 100% with DRESS overlapping SJS/TEN (OR = 31.9, 95% CI = [1.62, 626.6], P = 0.01). In contrast, HLA-B*58:01 is present in only 65% of patients with allopurinol-induced MPE (OR = 8.5, 95% CI = [4.2, 17.5], P = 2.3 × 10-9). There was no difference in HLA-B*58:01 positivity between survivors and nonsurvivors (P = 0.4270).
Figure 1The associations of the HLA-B∗58:01 allele with different allopurinol-related cADRs phenotypes in Taiwan population. In the column for odds ratios, values indicate odds ratios and horizontal lines indicate 95% confidence intervals. BSA, body surface area; cADR, cutaneous adverse reaction; CI, confidence interval; DRESS, drug reaction with eosinophilia and systemic symptoms; MPE, maculopapular exanthema; SCAR, severe cutaneous adverse reaction; SJS, Stevens-Johnson syndrome; TEN, toxic epidermal necrosis.
Meta-analysis of HLA-B*58:01 allele in allopurinol-induced cADRs
In addition to our study, 10 studies (Figure 2) investigated the association between HLA-B*58:01 and allopurinol-induced cADRs among populations in Thailand, Korea, Japan, China, Hong Kong, Europe, Portugal, and Australia. These studies of various countries and ethnic groups show a very strong association between HLA-B*58:01 and allopurinol-induced SJS/TEN (pooled OR = 57.33, 95% CI = [35.09, 93.67], Z = 16.61, P < 1.0 × 10-5) compared with the general population. Heterogeneity between studies is low (σ2 = 0.00; I2 = 0%). A meta-analysis of the association between HLA-B*58:01 and allopurinol-induced DRESS yielded a pooled OR of 54.16 (95% CI = [21.44, 138.86], Z = 8.44, P < 1.0 × 10-5) compared with the general population—a strong correlation. Heterogeneity between studies was relatively low (σ2 = 0.58, I2 = 41%). Most studies reported a positive correlation between HLA-B∗58:01 and allopurinol-induced DRESS, except for an Australian study (
) that enrolled five white patients, only one of whom had the HLA-B*58:01 allele. The association of HLA-B*58:01 with allopurinol-induced MPE is much weaker than the association with allopurinol-induced SCARs. The pooled OR compared with the general population was only 5.62 (95% CI = [0.96, 32.74], Z = 1.92, P = 0.05).
Figure 2Meta-analysis of associations of HLA-B∗58:01 allele with allopurinol-induced cADRs in different populations. Meta-analysis of associations of HLA-B∗58:01 with allopurinol-induced cADRs, including SJS/TEN, DRESS, and MPE, in matched- and population-controls in this study and 10 other studies. The associations studies of HLA-B∗58:01 in matched- (allopurinol cADRs) and general population-controls was shown. CI, confidence interval; DRESS, drug reaction with eosinophilia and systemic symptoms; HLA, human leukocyte antigen; M-H, Mantel-Haenszel test; MPE, maculopapular exanthema; SJS, Stevens-Johnson syndrome; TEN, toxic epidermal necrosis.
Gene dosage effect of HLA-B*58:01 in allopurinol-induced cADRs
We further analyzed the impact of the HLA-B*58:01 gene dosage effect on allopurinol-induced cADRs (Table 2). Among the 146 patients with allopurinol-induced cADRs, 70% (n = 102) carried one HLA-B*58:01 allele (heterozygous), 14% (n = 20) carried two HLA-B*58:01 alleles (homozygous), and 16% (n = 24) had no HLA-B*58:01 alleles. Most (82%, n = 234) control subjects did not carry the HLA-B*58:01 allele. The strength of the association of HLA-B*58:01 with allopurinol-induced cADRs correlated with disease severity (OR = 44 for SCARs, OR = 8.5 for MPE; Figure 1). In addition, patients with two HLA-B*58:01 alleles were at greater risk for developing allopurinol-induced cADRs (OR = 81.47, 95% CI = [19.51, 568.95], P <0.001; Table 2) than those who were carriers of a single HLA-B*58:01 allele (OR = 17.42, 95% CI = [9.06, 33.01], P < 0.001). Among the control subjects, only 1% (n = 2) carried homozygous alleles, compared with 17% (n = 49) who carried heterozygous alleles. The distribution of HLA-B*58:01 genotypes according to the subphenotypes of cADRs is shown in Supplementary Figure S1 online.
Effect of renal impairment in allopurinol-induced cADRs
Baseline renal function was significantly worse in patients with allopurinol-induced cADRs than in control subjects (median eGFR = 31 vs. 63 ml/minute/1.73 m2, P < 0.05; Table 1). Patients with moderate or severe renal impairment were at increased risk of developing allopurinol-induced cADRs (P < 0.001). Moreover, severe renal impairment, defined as an eGFR less than 30 ml/minute/1.73 m2, was a significant risk factor for allopurinol-induced cADRs after multivariate adjustment (OR = 4.30, 95% CI = [1.96, 9.62], P < 0.001; Table 2). The poorer renal function of patients with allopurinol-induced cADRs compared with allopurinol-tolerant control subjects was observed in female and male patients (see Supplementary Tables S1 and S2 online). Furthermore, mortality analysis showed that 83% of nonsurvivors had poor baseline renal function (eGFR < 30 ml/minute/1.73 m2) (Supplementary Figure S2 online).
Coexistence of HLA-B*58:01 and renal impairment increased the risk and predictive value of allopurinol-induced cADRs
We analyzed the effect of HLA-B*58:01 and renal function in patients with allopurinol-induced cADRs (Figure 3a). The results were significant compared with age-matched and sex-matched control subjects (i.e., all patients who were not HLA-B*5801 allele carriers and had an eGFR > 30 ml/minute/1.73 m2; see Supplementary Table S3 online). Among patients with severe renal impairment but without the HLA-B*58:01 allele, the OR for allopurinol-induced cADRs was 3.82 (95% CI = [1.8, 8.3], P < 0.001; Figure 3a), compared with an OR of 15.25 (95% CI = [8.4, 27.7], P < 0.001) for patients with a single HLA-B*58:01 allele but no renal impairment. This indicates that the genetic effect has a greater impact than renal impairment on allopurinol-induced cADRs. The risk for allopurinol-induced cADRs was higher for patients with coexisting HLA-B*58:01 and impaired renal function, especially when the eGFR was less than 30 ml/minute/1.73 m2 (OR for heterozygous HLA-B*58:01 carriers = 15.25 for eGFR > 60 ml/minute/1.73 m2 and OR = 264.31 for eGFR < 30 ml/minute/1.73 m2). In fact, the risk of allopurinol-induced cADRs was higher in patients with coexisting heterozygous HLA-B*58:01 and impaired renal function than in those with homozygous HLA-B*58:01 alone (OR for heterozygous HLA-B*58:01 carriers with eGFR < 30 ml/minute/1.73 m2 = 264.31; OR for homozygous HLA-B*58:01 alone = 72.45). The risk of allopurinol-induced cADRs was highest in patients with severe renal impairment and the homozygous HLA-B*58:01 allele (OR = 1,269.45; 95% CI = [192.3, 15,260.1]). Moreover, coexistence of HLA-B*58:01 and renal impairment increased the specificity and positive predictive value for allopurinol-induced cADRs (84% and 5%, respectively, for heterozygous HLA-B*58:01 carriers with normal renal function; 100% and 100%, respectively, for homozygous HLA-B*58:01 carriers with severe renal impairment [eGFR < 30 ml/minute/1.73 m2]).
Figure 3Analysis of the impact of the HLA-B*58:01 allele and renal impairment on allopurinol-induced cADRs.(a) ORs for allopurinol-induced cADRs in relation to renal function (eGFR >60, 30–60, and <30 mL/minute/1.73 m2) and presence of the HLA-B*58:01 risk allele (heterozygous and homozygous). Compared with allopurinol-tolerant subjects, the ORs were 3.82, 15.25, 72.45, 264.31, and 1,269.45 for the risk factors eGFR < 30 mL/minute/1.73 m2 alone, heterozygous HLA-B*58:01 with normal renal function, homozygous HLA-B*58:01 with normal renal function, heterozygous HLA-B*58:01 with an eGFR < 30 mL/minute/1.73 m2, and homozygous HLA-B*58:01 with an eGFR < 30 mL/minute/1.73 m2, respectively (P < 0.001 for all). (b) ROC curve for poor renal function and HLA-B*58:01 risk allele. The AUC was highest for concomitant presence of the homozygous HLA-B*58:01 allele and an eGFR < 30 mL/minute/1.73 m2. AUC, area under the curve; cADR, cutaneous adverse reaction; CI, confidence interval; eGFR, estimated glomerular filtration rate; OR, odds ratio; PPV, positive predictive value; ROC, receiver operating characteristic.
Receiver operating characteristic (ROC) analysis (Figure 3b) showed that the area under the curve (AUC) for patients with concomitant severe renal impairment and the HLA-B*58:01 allele was 0.89, which was significantly higher than for patients with one of these risk factors (AUC = 0.84 for the HLA-B*58:01 allele, AUC= 0.74 for severe renal impairment).
Discussion
Presentation of allopurinol-induced cADRs ranges from mild skin rash (MPE) to severe, potentially life-threatening hypersensitivity reactions (SJS, TEN, and DRESS). Although HLA-B*58:01 is strongly associated with severe allopurinol hypersensitivity reactions, its role and strength in variable clinical manifestations is yet to be determined. Our results suggest that HLA-B*58:01 is strongly associated with allopurinol-induced SCAR and more weakly associated with allopurinol-induced MPE. Although HLA-B*58:01 genetic testing has become increasingly useful in clinical practice, clinicians should be cautious in interpreting genetic test results. To our knowledge, there were no previous reports regarding the gene dosage effect of HLA-B*58:01 and its increased predictive accuracy with coexisting renal impairment in patients with allopurinol-induced cADRs.
The OR for allopurinol-induced SCARs associated with HLA-B*58:01 was lower in the present study than in our previous study (
)—44.0 (95% CI, [21.5, 90.3]) compared with 580.3 (95% CI, [34.4, 9780.9])—because of an increase in sample size and the inclusion of extended phenotypes of allopurinol-induced cADRs. Although HLA-B*58:01 is strongly associated with allopurinol-induced cADRs, the HLA-B*58:01 allele alone is not the only risk factor for predicting development of allopurinol-related cADRs. Other factors, such as age, sex, and renal insufficiency could also be involved in allopurinol hypersensitivity (
) The percentages of women and older people (age > 65 years old) are higher among patients with allopurinol-induced cADRs than among control subjects, which is consistent with previous findings (
Association between adverse reactions to allopurinol and exposures to high maintenance doses: implications for management of patients using allopurinol.
). A previous study has shown that the risk of chronic kidney disease (CKD) was significantly higher in female patients with hyperuricemia than in male patients (
). Hence, there may be more female patients treated with allopurinol for asymptomatic hyperuricemia with CKD in clinical practice. Because CKD is a significant risk factor for allopurinol-induced cADRs, we believe that this is one reason why women had a higher risk of allopurinol-induced cADRs than men.
In this study, we performed a meta-analysis of the associations of the HLA-B*58:01 allele with the various phenotypes for allopurinol-induced cADRs in different ethnic groups. The HLA-B*58:01 allele and allopurinol-induced SCARs were strongly associated regardless of ethnicity. A subgroup analysis of Asian and non-Asian populations (
Association of HLA-B*58:01 allele and allopurinol-induced Stevens Johnson syndrome and toxic epidermal necrolysis: a systematic review and meta-analysis.
) yielded similar findings. Furthermore, the risk for allopurinol-induced cADRs was greater for subjects with the homozygous HLA-B*58:01 allele than for the heterozygous allele. This gene dosage effect of HLA-B*58:01 on the risk of allopurinol-induced cADRs may be explained by the findings of our recent study on the functional role of HLA-B*58:01 in restricting activation of allopurinol-specific T cells (
Immunologic basis for allopurinol-induced severe cutaneous adverse reactions: HLA-B*58:01-restricted activation of drug-specific T cells and molecular interaction.
). Another important finding of the present study was that the OR, positive predictive value, and AUC in ROC analysis were higher in patients with the concomitant risk factors of HLA-B*58:01 and poor renal function. The risk for this subgroup was higher than the risk for patients with a single risk factor, which increased predictive accuracy, although it is uncertain if coexistence of the HLA-B*58:01 allele and poor renal function (especially an eGFR < 30 ml/minute/1.73 m2) would be rare in allopurinol-tolerant control subjects if the sample size were increased. Impaired renal function was reported to increase the risk of allopurinol-induced cADRs (
). Furthermore, patients with coexisting HLA-B*58:01 and impaired renal function had a higher risk for allopurinol-induced cADRs, especially those with severe renal impairment (eGFR < 30ml/minute/1.73 m2). The possibility that renal impairment plays a role in the pathogenesis of allopurinol-induced cADRs is similar to our recent findings that, in addition to predisposition because of susceptible HLA alleles, a defect in drug metabolism, or impaired drug clearance, is an important factor in drug-induced SCARs (
The mortality rate was higher among patients with allopurinol-induced cADRs with poor baseline renal function, perhaps because of renal dysfunction resulting in reduced elimination of the drug or its metabolites. Accumulation of oxypurinol, a primary metabolite of allopurinol that is primarily eliminated by the kidneys, may be important in triggering the hypersensitivity reactions in cADRs (
Insights into the poor prognosis of allopurinol-induced severe cutaneous adverse reactions: the impact of renal insufficiency, high plasma levels of oxypurinol and granulysin.
). Allopurinol is commonly used in clinical practice for treating hyperuricemia in patients with chronic kidney disease, which may explain the high prevalence and mortality of allopurinol-induced SCARs in Taiwan (data not shown). A previous study found that allopurinol may slow down the progression of renal disease in patients with CKD (
). This study has lead to the increased use of allopurinol in asymptomatic hyperuricemic in patients with CKD. However, renal insufficiency has also been found as a risk factor for allopurinol-induced cADRs in this study; hence, physicians should evaluate the risks and benefits before administration of allopurinol in patients with CKD.
Allopurinol has been widely used for over 30 years, and the standard dosing guidelines have been established (
Using allopurinol above the dose based on creatinine clearance is effective and safe in patients with chronic gout, including those with renal impairment.
). Recently, our and other studies have suggested that adverse drug reactions are weakly associated with initial allopurinol dosage in a Taiwanese population and can occur even at low doses (
). In this study, the initial dosage for allopurinol was not obviously higher for patients with allopurinol-induced cADRs than for allopurinol-tolerant control subjects (data not shown). However, our recent study showed that the plasma concentration of oxypurinol was significantly higher in patients with an eGFR less than 30 ml/minute/1.73 m2 than in patients with an eGFR of 30 ml/minute/1.73 m2 or greater (
Insights into the poor prognosis of allopurinol-induced severe cutaneous adverse reactions: the impact of renal insufficiency, high plasma levels of oxypurinol and granulysin.
). This suggests that poor renal function contributes to pathogenesis of allopurinol-related cADRs by increasing the plasma concentration of allopurinol or its metabolite. According to these results, the genetic dosage effect of HLA-B*58:01 and poor baseline renal function, especially an eGFR of less than 30 ml/minute/1.73 m2, appear to be more important factors than initial drug dosage in the development of allopurinol-induced cADRs.
In conclusion, this HLA-B*58:01 genotype and allopurinol-induced cADR phenotype association study provides comprehensive information on the clinical profile and impact of HLA-B*58:01 and the increased predictive importance of renal function in the development of allopurinol-related cADRs. The results show that not only the gene dosage effect of HLA-B*58:01 but also coexistence of the HLA-B*58:01 allele and severe renal impairment increase the risk of allopurinol-induced cADRs. In addition to the gene effect of HLA-B*58:01, assessment of renal impairment (eGFR < 30ml/minute/1.73 m2) is also necessary to prevent allopurinol-induced cADRs.
Materials and Methods
Patient recruitment and disease assessment
We recruited 146 patients with allopurinol-induced cADRs during the period of 1998–2014, as recorded in the Chang Gung Memorial Hospital health system and Taiwan Severe Cutaneous Adverse Reaction Consortium (T-SCARS) consortium, which includes Chung Shan Hospital, National Taiwan University Hospital, Veterans General Hospital, National Cheng Kung University Hospital, and Kaoshiung Medical University Hospital, Taiwan. Among these 146 patients, 41 patients with SCARs (23 patients with SJS/TEN and 18 patients with DRESS) were previously reported (
). In addition, we included 105 new patients with allopurinol-induced cADRs: 23 patients with SJS/TEN, 39 patients with DRESS, 3 patients with overlapping SJS/TEN and DRESS, and 40 patients with MPE. Patients with SJS/TEN were subdivided into two groups according to the extent of BSA detachment: less than 10% of BSA (SJS) and 10% or greater of BSA (includes overlapping SJS/TEN and TEN). For control subjects, we enrolled from the same hospital system 285 allopurinol-tolerant individuals who had received allopurinol for at least 6 months without evidence of adverse reactions. Of the 285 control subjects, 135 were previously reported (
). All patients were Taiwanese of Han Chinese ethnicity. This study was approved by the institutional review board of the ethical standards committee of each study site/center before study initiation and was conducted in accordance with the Declaration of Helsinki and all relevant local laws, regulations, and guidelines for the use of human subjects. Written informed consent was obtained from all patients and control subjects.
All patients were assessed by at least two dermatologists. The phenotypes were classified using the criteria of the RegiSCARS study group (
Correlations between clinical patterns and causes of erythema multiforme majus, Stevens-Johnson syndrome, and toxic epidermal necrolysis: results of an international prospective study.
). SJS and TEN are characterized by rapidly developing blistering exanthema with purpuric macules and target-like lesions accompanied by mucosal involvement and skin detachment. SJS is defined as less than 10% BSA skin detachment, TEN is defined as more than 30% BSA skin detachment, and SJS-TEN overlap is defined as 10–29% BSA skin detachment (
). For SJS/TEN the culprit drug was determined by the Algorithm of drug causality for epidermal necrolysis (ALDEN) algorithm of drug causality assessment, and for the other allopurinol-induced cADRs, the Naranjo algorithm was used (
ALDEN, an algorithm for assessment of drug causality in Stevens-Johnson Syndrome and toxic epidermal necrolysis: comparison with case-control analysis.
). Only patients with the criteria as probable to definite cases of SJS/TEN caused by allopurinol (ALDEN score ≥4 and/or Naranjo algorithm ≥5) were enrolled. Briefly, this assessment included prior drug reaction history; clinical manifestations of typical drug reactions; chronology or temporal relationship between drug use and onset of reaction; rechallenge, dechallenge, and improvement after discontinuation of suspected drugs; and notoriety of suspected drugs. The DRESS criteria and scoring system included cutaneous involvement with typical skin rash (e.g., exfoliative dermatitis, diffuse maculopapular exanthema), fever, eosinophilia, lymph node enlargement, atypical lymphocytes, internal organ involvement (liver, kidney, central nervous system, lung, heart, muscle), time of resolution, and evaluation of other potential causes (
). The DRESS with overlapping SJS/TEN phenotype matches both SJS/TEN and DRESS criteria as according to the RegiSCAR cases validation of probable or definite cases. The photographs of one case for “DRESS with overlapping SJS/TEN” is shown in Supplementary Figure S3 online. Furthermore, the statistical analysis of Table 2 excluding the 3 patients with DRESS overlapping SJS/TEN is shown in Supplementary Table S4 online, and the result is very similar to the original analysis. The MPE phenotype is characterized by generalized cutaneous erythematous macules and papules and is self-limited, without systemic involvement (
HLA-B genotypes were determined by sequence-specific oligonucleotide reverse line blots (DYNAL Biotech Ltd, Bromborough, UK) or SeCore1 HLA sequence–based typing (Invitrogen, Life Technologies, Carlsbad, CA). Ambiguous results were resolved by sequence-based typing.
Evaluation of renal function
To evaluate renal function, creatinine clearance was determined, and eGFR was calculated using the Modification of Diet in Renal Disease Study equation (
A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group.
). Baseline eGFR was obtained before onset of cADR. Patients with missing data for baseline eGFR were excluded. We classified renal function into three categories: normal (eGFR > 60 mL/minute/1.73 m2), moderate renal impairment (eGFR = 30–60 mL/minute/1.73 m2), and severe renal impairment (eGFR < 30 mL/minute/1.73 m2).
Meta-analysis
A comprehensive search was performed with databases including MEDLINE, Pre-MEDLINE, Cochrane Library, EMBASE, International Pharmaceutical Abstracts, CINAHL, PsychInfo, the WHO International database, Clinical Trial Registry, and ClinicalTrial.gov from their inceptions to June 2015. Only studies investigating the association between HLA-B*58:01 with allopurinol-induced SJS/TEN, MPE, and DRESS were included. All studies were evaluated and selected by two independent authors. The carrier allele frequency of each respective population was extracted from the Allele Frequency Net Database (http://www.allelefrequencies.net/). We compared HLA-B*58:01 in each allopurinol-related cADR phenotype. The pooled ORs were calculated using a random effect model.
Statistical analysis
Comparisons of allele frequencies between groups were performed using Fisher’s exact test. All P-values were two tailed. Since the P-values of less than 0.05 were considered significant, alpha was firstly set at 0.05, and the Holm-Bonferroni method was used to control the type I error in multiple comparisons with an alpha of 0.00625 (0.05/8). Hardy-Weinberg equilibrium was used for evaluation of genotype distribution among control subjects, and the results showed a χ2 of 0.96, with less than 3.84 indicating that the measured genotype frequencies were not significantly different from expectations. ORs were calculated using Haldane’s modification, which added 0.5 to all cells to account for possible zero counts. Measures of the validity of HLA-B*58:01 screening, that is, sensitivity, specificity, and positive predictive value, were also assessed and corrected with the disease prevalence rate (prevalence rate of allopurinol-induced cADRs = 1.056%; n = 83 out of 5,513) at Chang Gung Memorial Hospital, Taiwan. Univariate and multivariate logistic regression models were used to estimate ORs for HLA-B*58:01 and eGFR. ROC curve analysis was applied. AUC results are considered excellent for AUC values of 0.9–1.0, good for values of 0.8–0.9, fair for values of 0.7–0.8, poor for values of 0.6–0.7, and failed for values of 0.5–0.6.
P-values, ORs, and 95% CIs were calculated by logistic regression analysis and adjusted for sex and age. In real-life test, we compared the OR and performance characteristics of all subjects other than those with positive HLA-B*58:01 allele and eGFR less than 30ml/minute/1.73 m2. All analyses were conducted using R statistical software (GNU, The University of Auckland, Auckland, New Zealand) and SPSS Statistics (IBM, Chicago, IL).
Conflict of Interest
The authors state no conflict of interest.
Acknowledgments
This work was supported by grants from the National Science Council, Taiwan (MOST101-2320-B-010-072-MY3, MOST101-2321-B-010-027, MOST101-2628-B-182- 001-MY3, MOST101-2321-B-182-008, MOST102-2314-B-010-014-MY3, MOST102- 2321-B-182-006, MOST103-2321-B-182-001, MOST103-2321-B-182A-006, MOST101-2628-B-182A-001-MY3) and Chang Gung Memorial Hospital (BMRPG290011, OMRPG2C0011, OMRPG2C0021, OMRPG3E0041, CMRPG290051-3, CMRPG3D0351-2, CMRPG3D0361-2, CLRPG2E0051, CORPG3F0041). We thank analytic and technical support from Whole-Genome Research Core Laboratory of Human Diseases, Chang Gung Memorial Hospital, Keelung, Taiwan, for this study.
Correlations between clinical patterns and causes of erythema multiforme majus, Stevens-Johnson syndrome, and toxic epidermal necrolysis: results of an international prospective study.
Insights into the poor prognosis of allopurinol-induced severe cutaneous adverse reactions: the impact of renal insufficiency, high plasma levels of oxypurinol and granulysin.
A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group.
Immunologic basis for allopurinol-induced severe cutaneous adverse reactions: HLA-B*58:01-restricted activation of drug-specific T cells and molecular interaction.
Association between adverse reactions to allopurinol and exposures to high maintenance doses: implications for management of patients using allopurinol.
ALDEN, an algorithm for assessment of drug causality in Stevens-Johnson Syndrome and toxic epidermal necrolysis: comparison with case-control analysis.
Association of HLA-B*58:01 allele and allopurinol-induced Stevens Johnson syndrome and toxic epidermal necrolysis: a systematic review and meta-analysis.
Using allopurinol above the dose based on creatinine clearance is effective and safe in patients with chronic gout, including those with renal impairment.
EULAR evidence based recommendations for gout. Part II: Management. Report of a task force of the EULAR Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT).
30871-5&id=gr1.jpg){kind=link}
30871-5&id=gr2.jpg){kind=link}
30871-5&id=gr3.jpg){kind=link}