Endogenous plasma and salivary uracil to dihydrouracil ratios and DPYD genotyping as predictors of severe fluoropyrimidine toxicity in patients with gastrointestinal malignancies

Abstract

ObjectiveThe aim of this study was to evaluate the use of plasma and saliva uracil (U) to dihydrouracil (UH2) metabolic ratio and DPYD genotyping, as a means to identify patients with dihydropyrimidine dehydrogenase (DPD) deficiency and fluoropyrimidine toxicity. MethodsPaired plasma and saliva samples were obtained from 60 patients with gastrointestinal cancer, before fluoropyrimidine treatment. U and UH2 concentrations were measured by LC–MS/MS. DPYD was genotyped for alleles *7, *2A, *13 and Y186C. Data on toxicity included grade 1 to 4 neutropenia, mucositis, diarrhea, nausea/vomiting and cutaneous rash. Results35% of the patients had severe toxicity. There was no variant allele carrier for DPYD. The [UH2]/[U] metabolic ratios were 0.09–26.73 in plasma and 0.08–24.0 in saliva, with higher correlation with toxicity grade in saliva compared to plasma (rs=−0.515 vs rs=−0.282). Median metabolic ratios were lower in patients with severe toxicity as compared to those with absence of toxicity (0.59 vs 2.83 saliva; 1.62 vs 6.75 plasma, P<0.01). A cut-off of 1.16 for salivary ratio was set (AUC 0.842), with 86% sensitivity and 77% specificity for the identification of patients with severe toxicity. Similarly, a plasma cut-off of 4.0 (AUC 0.746), revealed a 71% sensitivity and 76% specificity. ConclusionsDPYD genotyping for alleles 7, *2A, *13 and Y186C was not helpful in the identification of patients with severe DPD deficiency in this series of patients. The [UH2]/[U] metabolic ratios, however, proved to be a promising functional test to identify the majority of cases of severe DPD activity, with saliva performing better than plasma.