Safety, feasibility and cost-effectiveness of genotype-directed individualized dosing of fluoropyrimidines

The fluoropyrimidine anticancer drugs 5-fluorouracil (5-FU) and capecitabine
are standard of care in the treatment of early and advanced breast, colorectal
and gastric cancer. There is ample evidence demonstrating that variation in
activity of the fluoropyrimidine-metabolizing enzyme dihydropyrimidine
dehydrogenase (DPD), encoded by the gene DPYD, causes clinically significant
differences in sensitivity to the toxic effects of 5-FU and capecitabine. DPD
deficiency, occurring in up to 5% of the population, is associated with the
risk of severe, potentially lethal, toxicity. Upfront genotype-directed
dose-adaptation is technically and logistically feasible, but has not been
uniformly implemented as standard of care since improvement of patient safety
and cost-effectiveness of the approach have been studied insufficiently.
Upfront phenotype-directed dosing has not been routinely implemented yet in
clinical practice either, due to insufficient evidence on clinical validity and
utility, bio-analytical issues, logistical and financial issues. In this study
it will be investigated whether genotype-and phenotype-directed dosing of
fluoropyrimidines improves patient safety. It is expected that tailoring
fluoropyrimidine therapy to the patient*s genotype and phenotype improves
safety and reduces the costs of treatment of early and advanced breast,
colorectal and gastric cancer by reducing severe treatment-related toxicity and
should thus change the standard of care.

The primary objective is to determine whether the rate of severe toxicity (CTC
grade 3 to 5) associated with fluoropyrimidine treatment can be significantly
diminished by individualized dosing of fluoropyrimidines based on upfront
genotypic assessment of DPD deficiency.

Secondary objectives are:
• To determine whether individualized dosing based on upfront genotypic
assessment of DPD deficiency is cost-effective.
• To determine the additional value of an array-based screening of additional
decreased activity DPYD-alleles, by determining associations between these
variants and occurrence of severe fluoropyrimidine-induced toxicity.
• To assess the pharmacokinetic profile of capecitabine and 5-FU in DPD
deficient patients (e.g. patients carrying variant DPYD alleles) given reduced
dosages of the respective drugs.
• To determine the clinical sensitivity, specificity, positive predictive value
and negative predictive value of phenotype-based screening using the endogenous
DHU/U ratio.
• To determine the clinical sensitivity, specificity, positive predictive value
and negative predictive value of phenotype-based screening using the uracil
test dose.
• To determine the clinical sensitivity, specificity, positive predictive value
and negative predictive value of phenotype-based screening using the uracil
breath test.

The study will have a duration of about 24 months. A number of 1250 patients
will be included. In this study prospective screening for four single
nucleotide polymorphisms (SNPs) in DPYD (DPYD*2A, c.2846A>T, c.1236G>A/HapB3
and DPYD*13) will be performed using validated real-time polymerase chain
reaction (PCR) assays. Patients with a SNP in DPYD will be treated with a
25-50% reduced starting dose, depending on which SNP is identified. The dose
will be titrated in subsequent cycles, to achieve maximal safe exposure. In
addition to the genotyping, the DPD phenotype of all patients will be
determined by measuring the baseline dihydrouracil/uracil (DHU/U) ratio, in
order to investigate whether phenotype-guided treatment can further improve
patient safety. In a subgroup of patients, other phenotyping methods will be
tested: measuring the plasma levels of uracil after an uracil test dose and a
uracil breath test after a dose of [2-13C] -labeled uracil. To validate these
tests, these phenotyping results will be compared with the results of a DPD
activity assay (which measures DPD enzyme activity in peripheral blood
mononuclear cells), which is considered the gold standard in measuring the DPD
phenotype. The results of the phenotyping tests will be evaluated at the end of
this study and will be the basis for a subsequent study. In this next study the
combined use of upfront genotype- and phenotype based DPD-screening will be
investigated. Patients will be followed for toxicity during the whole treatment
period. The cost-effectiveness of genotype-guided dosing will be calculated.

The intervention will consist of dose adjusting for fluoropyrimidines for
patients heterozygous for a polymorphism in DPD. The degree of dose adjustment
is depending on the SNP that is found. For DPYD*2A and DPYD*13 the dose will be
50% of standard dose and for c.2846A>T and c.1236G>A/HapB3 the dose will be 25%
lowered.

Blood will be drawn from all participating patients for determining the
genotype and the DHU/U ratio, prior to start of the fluoropyrimidine therapy.
Pharmacokinetic analyses will be performed on a subgroup of the participating
patients (all patients carrying a SNP in DPYD), for which hospitalization for
the duration of 9 hours is necessary and blood samples at 9 (for capecitabine
and 5-FU bolus injection) or 11 (for 5-FU continuous infusion) different time
points will be collected. In a subgroup of 15 patients the endogenous DHU/U
ratio at different time points will be determined. Blood samples will be drawn
at 9 time points and hospitalization for the duration of 9 hours is necessary.
The uracil test dose assay and the breath test will be performed on a subgroup
of the participating patients (around 260 patients). For the uracil test dose
assay patients will have to swallow an uracil suspension, hospitalization
during 3 hours is necessary and blood samples at 2 different time points will
be collected. For the uracil breath test patients have to swallow an aqueous
solution of [2-13C]-uracil and exhaled breath samples and blood samples are
collected (hospitalization of around 1 hour is necessary). One extra
bloodsample for these patients undergoing these two phenotyping tests is
necessary to determine the DPD enzyme activity.
Since dosages will be reduced for patients carrying a SNP in DPYD there will be
a small risk of underdosing for these patients. This risk will be very small,
since dose escalation according to tolerance after the first two courses will
be performed.