Page 43 - Personalised medicine of fluoropyrimidines using DPYD pharmacogenetics Carin Lunenburg

Introduction
The fluoropyrimidine anticancer drug 5-fluorouracil (5-FU) and its oral prodrug capecitabine are frequently used in the treatment of a variety of cancers, including breast, colorectal, head and neck and gastric cancer. The dihydropyrimidine dehydrogenase enzyme (DPD), encoded by the gene DPYD, plays a key role in the metabolism of fluoropyrimidines. Over 80% of the administered dose of 5-FU is metabolized by DPD in the liver into the inactive metabolite 5,6-dihydro-5-fluorouracil, which makes DPD the rate-controlling enzyme for inactivation of 5-FU.1 DPD deficiency occurs in 4─5% of the population and results in decreased inactivation of 5-FU. This can lead to an increase in active metabolites of 5-FU which is associated with an increased risk of severe and even fatal toxicity.2-4 Toxicity could be limited by exposing DPD- deficient patients to a decreased dose of fluoropyrimidines, to keep plasma levels of 5-FU and its metabolites at a therapeutic level for these patients. Over 30 genetic polymorphisms in DPYD have been described among which several lead to reduced function or a nonfunctional DPD enzyme.4-6 Polymorphisms can appear in heterozygous form (one SNP on one allele), homozygous form (two identical SNPs on two alleles) or double heterozygous form (two different SNPs on either one or two alleles, the latter is also called compound heterozygous). Two SNPs on two alleles lead to a larger decrease in DPD enzyme activity, compared with the heterozygous form. An example of a DPYD polymorphism is the splice-site variant DPYD*2A (IVS14+1G>A; c.1905+1G>A; rs3918290), which leads to deletion of exon 14 and hence a nonfunctional DPD enzyme and is the most studied polymorphism in DPYD.
In recent years, genotyping costs have dropped significantly and pre-emptive testing for single or multiple SNPs to guide treatment with fluoropyrimidines has become accessible. Upfront genotype-directed dose adaptation of fluoropyrimidines is feasible and has been shown to increase safety for patients and to be cost-effective for DPYD*2A.7,8 However, only a minority of institutions have implemented screening programs as standard of care.9-11 Some physicians are reluctant to implement upfront genotype-guided dosing due to a lack of results from prospective randomized studies comparing genotype-guided and traditional dosing. The only prospective randomized study was terminated prematurely for ethical reasons as one patient in the control arm died due to 5-FU-related toxicity.12
In addition to DPYD*2A, other SNPs in DPYD have been described to result in decreased DPD enzyme activity, including DPYD*13 (c.1679T>G; I560S; rs55886062), c.2846A>T (D949V; rs67376798) and c.1236G>A (E412E; rs56038477, in haplotype B3).13-15 However, not all of these SNPs result in a similar decrease in DPD enzyme activity as DPYD*2A.3,14,16 As a result of the growing number of alleles and their range of activity, deriving DPD phenotype from genotype is increasingly challenging. In the near future the number of alleles will increase even further, since genetic testing is developing fast and single SNP testing might be replaced by testing SNP panels, whole exome sequencing or even whole genome sequencing. Consequently, there is a need for an individualized recommendation of dose adjustment of fluoropyrimidines, taking into account the specific genetic variants and their resulting reductions in DPD enzyme activity. In this paper we describe a method for translation of DPYD genotype into DPD phenotype making use of the gene activity score. This method accounts for the differences in functionality of the SNPs in DPYD, which results in a more differentiated dose adjustment and thus in optimal safety and effectiveness.
3
DPYD gene activity score
41