Page 328 - Personalised medicine of fluoropyrimidines using DPYD pharmacogenetics Carin Lunenburg

Chapter 13
and 5-fluoro-5,6-dihydrouracil (5-FDHU) plasma levels are used to calculate pharmacokinetic parameters. In a study setting, three patients had marked alterations in pharmacokinetic parameters and possibly severe toxicity was avoided by changing the 5-FU treatment into irinotecan treatment.52 The 5-FU test dose did not result in side effects in any of the patients in this study, which questions the suitability of this test dose, as the metabolizing enzyme DPD has a certain overcapacity. As was stated by van Staveren et al., a test dose of uracil of 500 mg/m2 fully saturates the DPD enzyme.54
Implementation of DPYD genotyping in the Netherlands
Three Dutch hospitals participated in the study of Deenen et al., applying DPYD*2A genotyping in over 2,000 recruited study patients between May 2007 and October 2011. Thereafter, more studies on DPYD variants and their association with the onset of severe fluoropyrimidine-induced toxicity became available. Within this period, some hospitals in the Netherlands implemented routine DPYD genotyping of all patients starting fluoropyrimidines, e.g. the LUMC in April 2013 and the Maastricht University Medical Center in 2013 as well.9,17 In April 2015 we started recruiting patients in our prospective study (chapter 5).15 Seventeen hospitals in the Netherlands participated in this study and implemented or outsourced DPYD genotyping either for study patients only or for all patients starting fluoropyrimidine treatment. In 2016, a survey was published in the Dutch Medical Oncology Journal.55 This survey was sent to oncologists in the Netherlands. Some remarkable results were found. First, 65% of the responders answered that DPD status was determined as standard for all patients starting treatment with fluoropyrimidines. Second, 80% of the oncologists used DPYD genotyping to determine DPD deficiency, compared to 15% of responders who used a DPD phenotyping test. Possibly these results were a little overestimated, as physicians who had experience with requesting these tests were more likely to reply to the survey compared to physicians who did not order DPD deficiency tests. Also, the results of the survey were not adjusted based on the number of respondents per hospital, which could give a misleading image on the status of DPYD genotyping in the Netherlands in 2016. Yet, it is clear that the use of DPYD genotyping in the Netherlands is ahead of the use in many other countries. Some research groups in France, the UK, Italy, Germany and the USA were able to implement DPYD genotyping, whether or not combined with DPD phenotyping, in their hospital or clinical institute and surrounding centres.
Other aspects of implementation
Treatment costs for patients did not increase when applying prospective genotyping of DPYD*2A, or DPYD*2A, DPYD*13, c.2846A>T and c.1236G>A, as was shown by Deenen et al. and in chapter 6 of this thesis.14,16 Expanding the genotyping panel from one variant to four variants did not increase the costs of genotyping much, while more patients at risk could be identified, and thus more (costs of) severe toxicity could be prevented. Currently, most hospitals can offer DPYD genotyping tests for approximately €100. Genotyping assays are becoming less expensive despite the addition of more variants to a genotyping panel, therefore it is expected that DPYD genotyping will probably remain cost-neutral. However, this holds to a current extend. If the panel of predictive variants becomes too large to be
326