Objective: To investigate if dihydropyrimidine dehydrogenase phenotyping has added value when combined with DPYD genotyping in predicting fluoropyrimidine-related toxicity.
Methods: Retrospective cohort study in which treatment and toxicity data were collected of 228 patients genotyped for four DPYD variants and phenotyped using an ex vivo peripheral blood mononuclear cell assay.
Results: Severe toxicity occurred in 25% of patients with a variant and normal dihydropyrimidine dehydrogenase activity, in 21% of patients without a variant and with decreased dihydropyrimidine dehydrogenase activity, and in 29% of patients without a variant and with normal dihydropyrimidine dehydrogenase activity (controls). The majority of patients with a variant or a decreased dihydropyrimidine dehydrogenase activity received an initial dose reduction (68% and 53% vs 19% in controls) and had a lower mean dose intensity (75% and 81% vs 91% in controls). Fifty percent of patients with a variant and decreased enzyme activity experienced severe toxicity, despite the lowest initial dose and whole treatment dose intensity. They also experienced more grade 4/5 toxicities.
Conclusions: Our results indicate that a combined genotype-phenotype approach could be useful to identify patients at increased risk for fluoropyrimidine-associated toxicity (e.g. patients with a variant and decreased dihydropyrimidine dehydrogenase activity). Because the group sizes are too small to demonstrate statistically significant differences, this warrants further research in a prospective study in a larger cohort.
Keywords: DPD; DPD deficiency; DPYD; decreased DPD activity; fluoropyrimidine toxicity; fluoropyrimidines; pharmacogenetics.