PI-024

A. M. Moyer, J. M. Skierka, L. M. Baudhuin; Mayo Clinic College of Medicine, Rochester, MN

BACKGROUND: UGT1A1 variants are associated with hyperbilirubinemia and irinotecan toxicity. While UGT1A1*28 has been associated with total bilirubin in many studies, evaluation of the *60 and *93 promoter variants has been limited. We evaluated the impact of *60 and *93 on total bilirubin in our clinical cohort.
METHODS: Our retrospective study included 479 hyperbilirubinemic patients with UGT1A1 sequencing data. We analyzed the association of *28/*36/*37, *60, and *93 with total bilirubin (n=479) and the ratio of conjugated to total bilirubin (n=277) using linear regression.
RESULTS: In a univariate analysis, *28 was associated with total bilirubin (p<0.0001) and the ratio of conjugated to total bilirubin (p=0.031). In a multivariate analysis, total bilirubin was associated with *93 (p=0.027), but not *28 (p=0.398) or *60 (p=0.974), and the ratio of conjugated to total bilirubin was associated with *93 (p=0.019), but not *28 (p=0.925) or *60 (p=0.075). Of the 276 patients who were homozygous *28, 275 were also *60 homozygous, while only 211 were *93 homozygous. Of the 101 patients who were WT at *28, 12 were *60 homozygous and only 1 was *93 homozygous. While these SNPs are tightly linked, several rare haplotypes were observed. In the subset of patients with bilirubin ≤3.0, only *28 and *60 were associated with total bilirubin (p=0.003 and 0.026), and the haplotype with all three variants had a higher mean bilirubin than the all WT haplotype (p=0.048). An ethnicity-based variation in bilirubin and the ratio of conjugated to total bilirubin was observed.
CONCLUSION: In a clinical population, we confirmed the associations of *60 and*93 with bilirubin levels as previously reported in smaller cohorts, and found that the association varies by ethnic group, which may reflect varying haplotype structures.