UGT1A1 polymorphism predicts irinotecan toxicity: Evolving proof

Abstract

The antineoplastic agent, irinotecan (CPT-11) is metabolized by enzymes known to exhibit polymorphic activity. Its active metabolite SN38 is glucuronidated to an inactive product by UDP-glucuronosyltransferase, UGT1A1, the isoform catalyzing bilirubin glucuronidation. Thus, glucuronidation may be an important determinant of net SN-38 concentration in bile (termed SN-38 biliary index). Additional factors that determine SN-38 concentrations relative to its glucuronidated product include the activity of gut beta-glucuronidase, which affects recirculation of SN-38 and direct gut exposure to SN-38. Recent results suggest that inter-patient variability in SN38/SN-38 glucuronide kinetics and possibly irinotecan toxicity results from genetic variations in UGT1A1 expression. For example, genetic defects in UGT1A1 determine Crigler-Najjar and Gilbert's syndromes characterized by unconjugated hyperbilirubinemia. Gilbert's syndrome often remains undiagnosed and occurs in up to 19% of individuals homozygous for the UGT1A1 (TA) allele (TA insertion in the TATAA promoter). Furthermore, since irinotecan toxicity is inversely related to SN-38 glucuronidation rate, individuals with low UGT1A1 expression may experience severe toxicity. In recent studies, decreased SN-38 glucuronidating activity has been observed in livers obtained from individuals carrying the (TA) allele. Ando et al attempted to determine whether UGT1A1 genotype is predictive of irinotecan toxicity, in a retrospective and case-controlled study (note: there was a 3.5:1 control to case ratio). Because of small data sets analyzed and failure to control for variations in treatment patterns and other determinants of toxicity unrelated to UGT1A1, their conclusions are somewhat limited. Despite these limitations, it is clear that certain promoter polymorphisms were associated with severe toxicity. In their analysis of Japanese patients, multivariate analysis suggested that genotypes either heterozygous or homozygous for UGT1A1*28 would be a significant risk factor for severe irinotecan toxicity (P < 0.001; odds ratio, 7.23; 95% confidence interval, 2.52-22.3). Individuals heterozygous for UGT1A1*27 also encountered severe toxicity. One must caution however that the same genotype in another racial group may be less predictive of toxicity as other variant alleles may be more frequently expressed. Nevertheless, variable promoter TA repeats have been demonstrated to alter promoter function and transcriptional activity; this could therefore replace direct phenotyping (glucuronidation activity). However, a detailed human genotype-phenotype analysis with respect to UGT1A1 expression and function is still needed. These studies could lead to strategies for optimizing therapy with antineoplastic agents that inherently have a low therapeutic index. In the future, UGT1A1 genotyping may serve to spare patients from excessive toxicity resulting from therapy with irinotecan.