Dose-individualisation of fluoropyrimidines based on pre-treatment serum uracil levels: the Alpe2U study

DPYD-guided dosing enhances safety of fluoropyrimidine-based chemotherapy. However, approximately 23 % of patients still experience severe toxicity unexplained by the four commonly tested DPYD-variant alleles. Elevated pre-treatment uracil levels have been proposed as a surrogate marker for reduced DPD activity and an independent predictor of toxicity. This prospective study evaluated whether uracil-guided dose individualisation can reduce severe fluoropyrimidine-induced toxicity in DPYD wild-type patients.

Pre-treatment plasma uracil levels were quantified in patients scheduled to receive fluoropyrimidine-based therapy. DPYD wild-type individuals with uracil concentrations > 16 ng/mL (DPYD_wt/U_high) received a 50 % dose reduction, in accordance with French RNPGx guidelines. The incidence of grade ≥ 3 fluoropyrimidine-related toxicity was compared between dose-reduced DPYD_wt/U_high patients, DPYD_wt patients with uracil ≤ 16 ng/mL (DPYD_wt/U_normal), and a historical cohort of DPYD_wt/U_high patients treated at full dose. Pharmacokinetic data were compared to a second historical cohort.

Among 612 evaluable patients, 22 were DPYD_wt/U_high. The incidence of severe toxicity in the dose-reduced group was significantly lower than in historical full-dose DPYD_wt/U_high patients (20 % vs 43 %, P = 0.03) and comparable during the first 2 treatment cycles to DPYD_wt/U_normal patients (10 % vs 11 %). However, 5-fluorouracil exposure was markedly reduced in nineteen dose-reduced DPYD_wt/U_high patients (177 vs 381 ng*h/mL), while five subsequently treated fully dosed DPYD_wt/U_high patients exhibited comparable exposure to historical wild-type controls (456 vs 381 ng*h/mL). No correlation was found between uracil levels and DPD enzyme activity (R=-0.006, P = 0.98).

Uracil-guided dosing of fluoropyrimidines may reduce toxicity risk but leads to subtherapeutic 5-fluorouracil exposure in DPYD wild-type patients. This indicates that these patients are treated sub-optimally and that uracil is not a reliable predictor of DPD deficiency in DPYD wild-type patients.