The children anticoagulation and pharmacogenetics study (CAPS): Developing a dosing algorithm for acencocoumarol in paediatric patients

Abstract

Background: Dosing of vitamin K antagonists (VKA) in paediatric patients is complex. The large variability in VKA dose requirement asks for elucidating the factors associated with this variability and taking these into account when defining the dose for a patient. For warfarin, paediatric dosing algorithms have been developed, but not for acenocoumarol. Objectives: To develop a dosing algorithm for acenocoumarol in pediatric patients with and without genetic information. Methods: This multicentre retrospective follow-up study was carried out in Dutch anticoagulation clinics and children's hospitals. Patients were selected when they used acenocoumarol for >1 month between January 1995 and December 2014 and were ≤18 years of age. The primary outcome was the mean daily dose during a stable period. A stable period was defined as ≥3 consecutive international normalized ratio measurements within therapeutic range over a period of ≥3 weeks. Clinical information (including height, weight and indication) and saliva samples for genotyping of CYP2C9 (∗2 and ∗3), VKORC1, CYP4F2, CYP2C18 and CYP3A4 (∗1B and ∗22) were collected. Linear regression was used to analyse their association with the log mean stable dose. Results: In total, 175 patients were included of whom 86 patients had a stable period and no missing clinical information (clinical algorithm cohort) and of 80 also genetic information was available (genetic algorithm cohort). The mean age at the stable period was 9 years. The most common indications were Fontan circulation, prosthetic heart valve, deep venous thrombosis and dilated cardiomyopathy. The clinical algorithm, containing body surface area and indication, explained 45.0% of the variability in dose requirement of acenocoumarol. By adding the genotypes of VKORC1, CYP2C18, and CYP2C9∗2/∗3, 61.8% of the variability was explained (genetic algorithm). Conclusions: Clinical factors had the largest impact on the required dose of acenocoumarol in pediatric patients. Including genetic factors in the algorithm, and especially VKORC1, increased this with 16.8%.