Page 31 - Personalised medicine of fluoropyrimidines using DPYD pharmacogenetics Carin Lunenburg

Prospective DPYD genotyping: ready for prime time
Group
Falvella et al. 201532
Joerger
et al.
201531
DPYD variant
c.496A>G (rs2297595) c.1896T>C (rs17376848)
c.1896T>C (rs17376848) c.85T>C (rs1801265) c.2846A>T (rs67376798)
Association with 5-FU and/or capecitabine grade ≥3 toxicity
(OR/*RR [95% CI], p-value)
Overall toxicity (5.94 [1.29–27.22], p=0.022) (cap) 2 Overall toxicity (14.53 [1.36–155.20], p=0.027) (cap)
Diarrhoea (p<0.05) (cap) Hand-foot syndrome (p<0.02) (cap)
Brief summary of a few selected studies
with 5-FU and/or capecitabine induced severe toxicity. Included are three meta-analyses and three more recent papers. Results originating with only 5-FU or only capecitabine are explicitly marked. Rosmarin et al. have also tested 5-FU infusion and 5-FU bolus separately. Meulendijks et al. have described RR values, not OR values, as shown by *.
Abbreviations: 5-FU: 5-fluorouracil; in: infusion; bo: bolus; cap: capecitabine; CI: confidence interval; OR: odds ratio; RR: relative risk; AF: allele frequency.
Table 2. Test characteristics of genotyping for DPYD*2A and c.2846A>T
showing the results of DPYD variants and their associations
Test characteristics
Sensitivity
Specificity
Positive predictive value
Negative predictive value
Number needed to screen (i.e. genotype)
Number needed to treat (i.e. apply dose adjustments)
Terrazzino et al.9 Rosmarin et al.8
14.5% 11.8% 97.6% 98.4% 19.8% 23.6% 96.5% 96.4%
210 patients 251 patients 6 patients 5 patients
Clinical utility test characteristics of genotyping for DPYD*2A and c.2846A>T, calculated using “The HuGE Risk translator”33 for Terrazzino et al. and Rosmarin et al.
Clinical implementation of DPD deficiency testing
Advantages and disadvantages of phenotyping and genotyping as possible DPD deficiency screening methods were described previously29 and several institutes53-59 have executed (prospective) screening of DPYD variants or DPD deficiency in a study context. Unfortunately, available literature of clinical implementation remains limited to only a few centres in France, Germany, the Netherlands, Ireland and the United States of America (USA).44,53,60,61 An established and well-recognised DPYD clinical implementation program is that of the ‘Institut de Cancerologie de l’Ouest’ in Angers (France) where screening for DPD deficiency has been a regular procedure for over 10 years. Besides this institute, over 100 centres in France use the ‘Onco Drug Personalized Medicine’ or ODPM ToxTM and 2,000 patients are being screened with this approach every year.62,63 Boisdron-Celle et al. describe a large trial in which 11,104 patients were prospectively screened (combining genotyping and phenotyping) and patients with a DPYD variant or decreased DPD activity received an individual dose adjustment. Genotyping in the trial consisted of 24 mutations in DPYD
29