Association of plasma uracil concentration with 5-FU pharmacokinetics, dose-limiting toxicity, and DPYD genotype in patients with gastrointestinal cancer.

Abstract

751 Background: Plasma uracil concentration >16 ng/mL is used in some countries as a screening test for DPD deficiency. We conducted this exploratory study to evaluate the association of uracil concentration with 5-FU exposure, toxicity, and DPYD genotype in patients without known DPD deficiency. The main hypothesis was that lower uracil concentration would be predictive of subtherapeutic 5-FU drug exposure (5-FU AUC < 20 mg*h/L) in patients treated with infusional 5-FU. Methods: The study prospectively enrolled patients with GI cancer who were receiving initial chemotherapy with a regimen containing a 46-hour 5-FU infusion initially dosed at 2400 mg/m2. Study candidates were screened for the DPYD *2A, *13 and c.2846A>T gene variants; variant carriers were excluded. Participants gave pretreatment fasting morning blood samples for determination of plasma uracil concentration and DPYD whole-gene sequencing. Plasma samples were collected at hour 45 of 5-FU infusion in chemotherapy cycles 1 and 2 for determination of 5-FU exposure, measured as the 5-FU AUC. We estimated a linear mixed-effects model to test the association of uracil concentration with 5-FU AUC, adjusting for cycle, sex, creatinine, and 5-FU infusion dose (in mg/m2). Results: There were 29 evaluable participants with a median age of 64 years (range 41-81); nine (31%) were female. The most common cancer sites were colorectal (n = 15) and pancreatic (10). The median (IQR) plasma uracil concentration was 10.4 ng/mL (7.2, 12.6). Data for 5-FU AUC were available for 19 subjects in cycle 1 (median = 22.0 mg*h/L) and 27 subjects in cycle 2 (median = 19.3 mg*h/L). In the mixed-effects model, male sex was significantly associated with 5-FU AUC (beta = -12.0 [95% CI -17.2, -6.7], p <0.001) but plasma uracil concentration was not (beta = 0.35 [-0.16, 0.86], p=0.16). Holding other covariates constant, male sex was associated with a 29% decrease in 5-FU AUC (vs female) and plasma uracil concentration in the lowest quartile was associated with a 12% decrease in 5-FU AUC (vs highest quartile). Three of 29 participants (10%, all male) had uracil concentrations of >16 ng/mL, suggestive of DPD deficiency. The subject with the highest uracil concentration (23.1 ng/mL) had severe toxicity in cycle 1 (neutropenia, mucositis) requiring 5-FU dose reduction; whole gene sequencing identified a rare DPYD missense variant in this subject (c.2185G>A [p.A729T]). The other two subjects with uracil concentration >16 ng/mL did not have culprit DPYD gene variants and tolerated infusional 5-FU at 2400 mg/m2 without early dose-limiting toxicity. Conclusions: Our findings suggest that male sex is more important than low pretreatment plasma uracil concentration as a predictor of risk for subtherapeutic 5-FU drug exposure. This result is consistent with prior evidence that men experience less 5-FU associated toxicity than women.