Van Oijen et al. reported in the Journal [1] an investigation which aimed to ‘assess whether polymorphisms in UGT1A6 and CYP2C9 were related to the prevalence of upper gastrointestinal symptoms in cardiovascular patients using acetylsalicylic acid for the secondary prevention of ischaemic heart disease’. The study was based on the assertion that UGT1A6 and CYP2C9 mediate the glucuronidation and hydroxylation of acetylsalicylic acid (aspirin; ASA), respectively, although no references were provided to support this proposition. There was no significant association between UGT1A6 and CYP2C9 genotypes and the incidence of gastrointestinal adverse effects in 160 patients receiving ASA. Not discounting other important issues associated with the study, particularly unavailability of the ASA dose(s) received by patients, we wish to comment on the theoretical basis for the investigation. The metabolism of ASA is well characterized (reviewed in [2]). ASA itself is not glucuronidated or hydroxylated. Rather, ASA is hydrolyzed to the pharmacologically active salicylic acid (SA) by esterases in the gastrointestinal tract and liver, and only approximately 50–70% of the dose reaches the systemic circulation as unchanged drug following oral administration. ASA escaping first pass extraction is rapidly hydrolyzed to SA with a half-life of about 15 min. Once formed, SA is metabolized to salicyluric acid (SUA; the glycine conjugate), phenolic- (SPG) and acyl- (SAG) glucuronides, and gentisic acid (2,5-dihydroxybenzoic acid; GA). SUA is the major metabolite of SA. In subjects administered 900 mg of ASA, SUA accounted on average for 68% of the dose recovered in urine [3]. SA, SPG, SAG and GA accounted for 12.2%, 10.6%, 8.2% and 1.4% of the recovered dose, respectively. Although formation clearances of SUA and SPG are known to be dose-dependent [4], SA clearance via the SUA pathway is, on average, about three-fold higher than by glucuronidation while hydroxylation to form GA represents a very minor pathway of SA elimination. In contrast to ASA, the elimination half-life of SA (the precursor of GA and the glucuronide conjugates) ranges from a few hours to almost 1 day, depending on the dose. SUA is formed in a two-step process; the first, and activating, step involves formation of a CoA thioester (mediated by a ‘medium chain’ CoA ligase) which is subsequently acylated by glycine N-acyltransferase. The UDP-glucuronosyltransferase(s) (UGT) involved in SA metabolism has not been characterized unambiguously. Ciotti et al.[5] reported that recombinant UGT1A6 glucuronidated SA, but activity was very low (0.05 pmol glucuronide min−1) and relative formation of SPG and SAG was not assessed. More recently, multiple UGTs have been shown to have the capacity to catalyze both the phenolic and acyl glucuronidation of SA [6]; these include UGT 1A1, 1A3, 1A6, 1A7, 1A9, 1A10, 2B4 and 2B7. (It should be noted that UGT 1A7 and 1A10 appear to be expressed only in the gastrointestinal tract, while the other forms are expressed in liver and variably in other tissues). It was not possible from the data presented to ascertain the dominant UGT(s) involved in SPG and SAG formation. To the best of our knowledge, the role of CYP2C9 and other P450s in GA formation has not been assessed in a systematic manner. Thus, the conclusions of van Oijen et al.[1] are hardly surprising. Glucuronidation accounts, on average, for about 20% of SA metabolic clearance and UGT1A6 appears to be just one of several enzymes involved in SAG and SPG formation. GA is a very minor metabolite of SA (approx. 1%) and a role for CYP2C9 as the principal catalyst for the conversion of SA to GA has not apparently been demonstrated. Other studies have also investigated relationships between ASA response and CYP2C9 and UGT1A6 polymorphisms. Bigler et al.[7] reported an inverse relationship between colon adenoma risk and ASA ‘use’ in subjects with a CYP2C9*1/variant UGT1A6 genotype. In contrast, McGreavey et al.[8] found that polymorphisms in CYP2C9 and UGT1A6 (and some other genes) did not influence the protective effect of NSAIDs, either including or excluding ASA use, on the risk of colorectal cancer. Relationships between mutations in genes encoding xenobiotic metabolizing enzyme(s) and the response to compounds eliminated or metabolically activated by a polymorphic enzyme (e.g. CYP2C9, CYP2D6, UGT1A1) have been demonstrated in numerous studies. However, clinically meaningful, generalizable conclusions should not be expected from pharmacogenetic studies based on flawed knowledge of the enzyme(s) responsible for metabolite formation and where the relative contribution of the metabolic pathway to overall elimination (e.g. GA formation) is negligible.
Read full abstract