Page 45 - Personalised medicine of fluoropyrimidines using DPYD pharmacogenetics Carin Lunenburg

in a meta-analysis by Terrazzino et al. a strong correlation between the DPYD*2A allele and overall grade >3 toxicity was found (odds ratio [OR] 5.42, p<0.001).33 Deenen et al. described
a mean capecitabine dose reduction of 50%, guided by toxicity, in patients carrying DPYD*2A, compared with a mean dose reduction of 10% in wild-type patients.42 Also, an initial dose reduction of capecitabine or 5-FU of 50% of standard dose has proven to decrease the risk 3 of severe toxicity in DPYD*2A carriers.7,8 The above mentioned in vitro, ex vivo and in vivo studies provide solid evidence for the nonfunctionality of DPYD*2A and a 50% reduced function in patients heterozygous for DPYD*2A.
c.2846A>T (rs67376798)
The c.2846A>T variant allele was first described by van Kuilenburg et al. in 2000.28 The c.2846A>T polymorphism leads to a structural change in the DPD enzyme that interferes with cofactor binding or electron transport.16 Reported allele frequencies of c.2846A>T vary from 0.1 to 1.1% in African-Americans and Caucasians, respectively.13,19,24,46 In vitro data show that homozygous expression of the c.2846A>T variant results in an activity of 59% compared with wild-type (p=0.0031).13 Although the enzyme activity of c.2846A>T is significantly impaired, it is not comparable to the extent observed for DPYD*2A, where homozygous expression resulted in a completely nonfunctional enzyme.27 This finding that homozygous expression of c.2846A>T results in ∼50% reduction, suggests that a heterozygous carrier would have around 25% reduction in DPD activity. Furthermore, also in clinical practice a difference between the effect of the DPYD*2A variant and the c.2846A>T variant has been observed. Deenen et al. described an average 25% dose reduction for c.2846A>T heterozygous patients in response to fluoropyrimidine-related toxicity, compared with 50% for DPYD*2A heterozygous patients.42 Although there are less publications for c.2846A>T than for DPYD*2A, several studies and two meta-analyses found an association between the c.2846A>T variant and increased risk of severe fluoropyrimidine-associated toxicity, which indicates that a dose reduction is warranted.4,24,33,36,41,42,44,45,47 In the study by Rosmarin et al. an OR of 9.35 (p=0.0043) was found between c.2846A>T and capecitabine-related severe (grade >3) toxicity.47 The evidence described above shows that c.2846A>T has rest-activity left, but that a dose reduction would still be required to prevent toxicities that would occur using a full dose of fluoropyrimidines. Therefore, based upon the available evidence we can assume that a dose reduction of 25% is most rational.
DPYD*13 (rs55886062)
DPYD*13 was first described by Collie-Duguid et al. as “T1679G”.48 The allele frequency was found to vary from 0.07 to 0.1% in Caucasians.19,24 The precise functional consequences of the DPYD*13 variant have not yet been unraveled, but are thought to be related to destabilization of a sensitive region of the protein.16 DPYD*13 has been found in patients with decreased enzyme activity, not in patients showing normal DPD enzyme activity.29 Homozygous expression of this variant resulted in a 75% reduction of DPD enzyme activity compared with wild-type, as reported in an in vitro study by Offer et al.27 This suggests that this variant almost completely inactivates the protein. Decreased DPD enzyme activity in patients with the DPYD*13 variant was determined only in a limited number of ex vivo
DPYD gene activity score
43