When nucleosides interfere with pyrimidine base quantification. Example of dihydropyrimidine dehydrogenase phenotyping

Dihydropyrimidine dehydrogenase (DPD) phenotyping based on plasma uracil (U) and dihydrouracil (UH₂) concentrations is widely used before administration of chemotherapeutic agent 5-fluorouracil. However, analytical and physiological factors may compromise its reliability. We investigated potential analytical biases affecting U and UH₂ quantification using liquid chromatography-mass spectrometry (LC-MS/MS) methods, combined with different sample preparation procedures, including liquid-liquid extraction and protein precipitation with perchloric acid (PCA), acetonitrile (ACN), and trichloroacetic acid. Incomplete chromatographic separation led to UH₂ overestimation due to in-source fragmentation of endogenous compounds, particularly dihydrouridine (DhUrd). This bias was amplified in patients with renal impairment, where DhUrd accumulates. Protein precipitation with PCA resulted in marked increases in both U and UH₂ concentrations compared with ACN or liquid-liquid extraction. Mechanistic investigations demonstrated that PCA induces chemical conversion of DhUrd into UH₂ and partial conversion of uridine (Urd) into U, rather than improving analyte recovery from protein binding as previously described. ACN precipitation was sufficient to release protein-bound U and UH₂, with negligible residuals. Other sources of interference (cytosine, nucleotides, and RNA) contributed minimally. Analytical interferences and sample preparation artifacts significantly impact U and UH₂ quantification and may lead to misinterpretation of DPD phenotyping results. Established UH₂/U ratio decision thresholds may require re-evaluation using harmonized analytical methods accounting for endogenous interferences.