Complete DPYD genotyping combined with dihydropyrimidine dehydrogenase phenotyping to prevent fluoropyrimidine toxicity: A retrospective study

Abstract

AbstractIntroductionIn April 2019, French authorities mandated dihydropyrimidine dehydrogenase (DPD) screening, specifically testing uracilemia, to mitigate the risk of toxicity associated with fluoropyrimidine‐based chemotherapy. However, this subject is still of debate as there is no consensus on a standardized DPD deficiency screening test. We conducted a real‐life retrospective study with the aim of assessing the impact of DPD screening on the occurrence of severe toxicity and exploring the potential benefits of complete genotyping using next‐generation sequencing.MethodsAll adult patients consecutively treated with 5‐fluorouracil (5‐FU) or its oral prodrug at six cancer centers between March 2018 and February 2019 were considered for inclusion. Dihydropyrimidine dehydrogenase deficiency screening included gene encoding DPD (DPYD) genotyping using complete genome sequencing and DPD phenotyping (uracilemia or dihydrouracilemia/uracilemia ratio) or both tests. Associations between each DPD screening method and (i) severe (grade ≥3) early toxicity and (ii) fluoropyrimidine dose reduction in the second chemotherapy cycle were evaluated using multivariable logistic regression analysis. Furthermore, we assessed the concordance between DPD genotype and phenotype using Cohen's kappa.ResultsA total of 551 patients were included. Most patients were tested for DPD deficiency (86%) including DPYD genotyping only (6%), DPD phenotyping only (8%), or both (72%). Complete DPD deficiency was not detected in the study population. Severe early toxicity events were observed in 73 patients (13%), with two patients (0.30%) presenting grade 5 toxicity. Despite the numerically higher toxicity rate in untested patients, the occurrence of severe toxicity was not significantly associated with the DPD screening method (p = 0.69). Concordance between the DPD genotype and phenotype was weak (Cohen's kappa of 0.14).ConclusionDue to insufficient numbers, our study was not able to demonstrate any added value of DPYD genotyping using complete genome sequencing to prevent 5‐FU toxicity. The optimal strategy for DPD screening before fluoropyrimidine‐based chemotherapy requires further clinical evaluation.