Page 330 - Personalised medicine of fluoropyrimidines using DPYD pharmacogenetics Carin Lunenburg

Chapter 13
Beyond current DPYD genotyping
It is known that DPYD variants are not the only risk factor for DPD deficiency, and DPD deficiency is not the only risk factor for severe fluoropyrimidine-induced toxicity. Approximately 17% of patients experiencing severe fluoropyrimidine-induced toxicity can be identified as carriers of one of the four currently genotyped DPYD variants. 39-61% of the patients who experienced severe toxicity were identified as DPD deficient patients, thus it was estimated that less than half of the DPD deficient patients could be identified by the four currently genotyped DPYD variants.58 In order to increase the predictability of severe fluoropyrimidine-induced toxicity, we must better predict risk factors for DPD deficiency, and additionally look into factors outside of DPD. Recently, a study was published in which eight years of combining genotyping and phenotyping tests were described.9 This study showed that only 25.3% of the DPD deficient patients was a carrier of one of the four currently genotyped DPYD variants. Patients with a DPD deficiency, but who did not carry the DPYD*2A variant, were genotyped for the entire coding region of DPYD. DPD deficiency could be explained by DPYD variants in 23% of these patients. This results in an expected approximately 42% of DPD deficiency related to DPYD variants. Variants in other regions, which have not been sequenced before, could still contribute to DPD deficiency. Unfortunately, the abovementioned study had no toxicity data of the patients, thus the prediction of DPYD variants for DPD deficiency could be made, but not the prediction for severe toxicity.
It is clear that not only DPYD variants are involved in the onset of severe fluoropyrimidine- induced toxicity. Therefore the DNA of the patients participating in chapter 5 was analysed by genome-wide association study (GWAS), in order to discover novel variants related to the onset of severe fluoropyrimidine-induced toxicity. This study was described in chapter 12. Approximately 700,000 single nucleotide polymorphisms (SNPs) in different genes were genotyped, and imputed to over four million SNPs. While no genome-wide significant SNPs could be identified, six variants were suggestive for the onset of severe toxicity. These variants warrant replication in an independent cohort. After validation, variants can be added to the prospective genotyping panel. In addition to the variants in chapter 12, validation is required for all newly identified variants. For example, some newly identified variants were recently presented in a series of patients who experienced severe toxicity,59 yet it is unclear if these variants could also be identified in patients who did not experience severe toxicity, and thus the clinical value of these variants needs to be determined. As described by Ciccolini et al. in 2010, both genetic and epigenetic factors, such as promotor hyper methylation or variations in transcriptional factor expression, play a role in DPYD dysregulations,60 and should be a focus of future research in DPYD genotyping.
Phenotyping assays
DPD phenotyping could also be used to predict severe fluoropyrimidine-induced toxicity. As described before, the DPD enzyme activity measurement in PBMCs is a well-established method to determine DPD activity.2,8,9 Additionally, DPD phenotyping assays were developed, such as the 2-13C uracil breath test,61-63 the uracil loading dose,54,64 endogenous dihydrouracil/ uracil (DHU/U) ratio and endogenous uracil concentrations.65,66 The status of each DPD
328