Background/Purpose: Allopurinol
is the leading choice of urate-lowering therapy for gout (>95% of treated
cases); however, it is associated with the rare but potentially fatal allopurinol
hypersensitivity syndrome (AHS) (i.e., Steven
Johnson Syndrome [SJS] and Toxic Epidermal Necrolysis [TEN]). Beyond
renal dosing to help reduce the risk of AHS, the recently discovered strong
associations between HLA-B*5801 allele carriage and AHS led to the 2012 ACR recommendation
to screen for this marker among high-risk Asians, but not Whites, given the
higher allele frequency in Asians. This leaves several open questions: 1) are Asians
at a higher risk of developing AHS than Whites, given the allele differences,
and 2) what about other races?  This
obvious information can be immediately useful in clinical care and policy
decisions.  We examined this potential racial
difference in the risk of hospitalized SJS and TEN using a nationwide inpatient
dataset.

Methods: We used the National Inpatient Sample (NIS),
a database representative of hospitalizations in the US from 2009-2011, where severe AHS (caused by
uric acid metabolism drugs) patients were identified by a primary discharge diagnosis
ICD-9-CM code for SJS (695.13) and TEN (695.14-15), followed by a secondary
discharge diagnosis of an adverse effect caused by uric acid metabolism drugs
in therapeutic use (E9447) or gout (274.x). In the context, severe AHS occurs
almost exclusively related to allopurinol. We calculated race-specific
hospitalization rate ratios for AHS using Whites as the reference; in-hospital
mortality and length of stay were also compared. Analyses were performed using NIS
sampling weights to obtain US national estimates.

Results: During
2009-2011 in the US, there were 608 patients admitted with SJS or TEN meeting
our criteria. The mean age was 69 years and 47% were men.  Of the 608, there was a substantial over-representation
of Asians (23%) and Blacks (28%) as compared to Whites (32%) and Hispanics (5%),
given their background population rates from the US census (5%, 13%, 78%, and
16%, respectively).  The hospitalization
rate ratios for AHS among Asians, Blacks, Hispanics, and Whites were 11.4, 5.2,
0.8, and 1.0 (referent). These ratios agreed with their HLA-B*5801 frequencies in the US
population (7.4%, 4%, 1%, and 1%, respectively).  In-hospital
mortality of the 608 was 13%.  The
mortality rate as well as the mean length of stay were higher among Asians than
other races (p=0.001 and <0.001, respectively). 

Conclusion: Our findings based on these nationally
representative inpatient data indicate that race is a strong determinant of severe
AHS risk: Asians (11x) > Blacks (5x) > Hispanics (1x) = Whites (1).  Mortality and length of stay were worse among
Asians.  These data call for extra caution
among Asians (and perhaps Blacks) when considering allopurinol, and support the
ACR recommendation to screen for HLA-B*5801
for high-risk Asians.