Development, validation and application of a liquid chromatography tandem mass spectrometry assay for uracil and 5,6-dihydrouracil along with 5-fluorouracil and its metabolites, in human plasma

5-Fluorouracil (5-FU), a pyrimidine analogue, is one of the commonly prescribed chemotherapeutic drugs for solid tumours, principally of the digestive tract, breast plus head and neck. Conventional 5-FU dose normalisation relies on body surface area adjustment. The body surface area approach bases the dose per mg/m2 and relies on maximum tolerated doses established in early clinical trials. However, prior work has established an optimal exposure window, minimising toxicities balanced against adequate responses. Up to one-hundred fold variability in pharmacokinetic parameters between individuals has been observed. More significantly, greater than 50% of patients dosed by body surface area have plasma 5-FU concentrations outside the optimal range. The plasma 5,6-dihydrouracil/uracil (UH2/U) ratio is a possible phenotypic marker of dihydropyrimidine dehydrogenase (DPD) metabolic activity, hence may serve as an index of 5-FU response and toxicity. The UH2/U ratio can potentially be useful to predict the initial dose, aiming to diminish early toxicity events. Studies have re-affirmed the significance of 5-FU and 5,6-dihydro-5-fluorouracil (FUH2) plasma concentrations for therapeutic drug monitoring (TDM). However, FUH2 has limited stability in plasma which hinders compliance with routine sample handling. Additional 5-FU metabolites, α-fluoro-β-ureidopropionic acid (FUPA) and α-fluoro-β-alanine (FβAL) are stable in plasma and thus, are candidates to facilitate practical 5-FU TDM. Aims: To develop, validate and provide proof-of-concept application of a liquid chromatography tandem mass spectrometry (LC-MS/MS) assay quantifying endogenous U and UH2, as well as 5-FU, FUH2, FUPA and FβAL following 5-FU continuous infusion, in human plasma. Methods: Systematic optimisation experiments for mass spectrometry parameters, liquid chromatography and sample preparation were conducted to maximise analyte responses. For the final method, plasma samples were consecutively precipitated by ZnSO4 and acetonitrile/methanol, and then extracted by liquid-liquid extraction (LLE) using ethyl acetate/ isopropanol with pH lowering, in two-stages. Supernatants were mixed, dried and reconstituted. Analytes were resolved on the Luna pentafluorophenyl (PFP) column and analysed by MS/MS via electrospray ionisation in positive polarity (ESI+). Stable isotope-labeled internal standards (IS) were used for each analyte. The method was validated according to the FDA (Guidance for Industry: Bioanalytical method validation) and applied to samples from cancer patients (n=10) on 5-FU. Results: The analytical response was linear (r2 ≥ 0.99) in the concentration (ng/mL) ranges: 50-10, 000 for FβAL and FUH2, 50-5,000 for FUPA, 50-100,000 for 5-FU, 5-200 for U and 10-400 for UH2. Within- and between-run accuracy and precision were ≤ 10.2 and ≤ 9.8% respectively across the range of analytes. The IS normalised matrix effects were within 93-112% with coefficient of variation (CV) ≤ 9.7% and normalised recoveries within 91-107% with CV ≤ 9.8%. Deviations were ≤ 2.0% from the mean values when incurred samples were reanalysed. A selective, sensitive, accurate and robust assay was developed, validated and applied