The influence of age, vitamin K and genetics on anticoagulation outcomes in adults and children

Abstract

The aim of this PhD was to identify the influence of factors that affect anticoagulation response to warfarin and NOACs on clinical response with the aim to further optimise treatment safety and patient management. In a randomised controlled trial, 181 patients with INR >6.0 were randomly allocated to receive orally either a tailored dose based upon index INR and BSA, or a fixed-dose (1 or 2mg) of vitamin K. The tailored dose resulted in a greater proportion of patients returning to within target INR range (68.9% v 52.8%; p=0.026), whilst a smaller proportion of patients remained above target INR range (p<0.001). Studies have confirmed that variant CYP2C9 and VKORC1 allele carriers require more time to achieve stable dosing and have a higher risk of supra-therapeutic INR and serious bleeding risk during initiation. In a retrospective cohort study, neither VKORC1, CYP2C9 genotype, nor their total variant allele count was associated with TTR% during warfarin maintenance or induction phase (p>0.05). A study in children on chronic warfarin therapy showed low levels of plasma vitamin K which are associated with CYP4F2 genotype with the patients with low plasma vitamin K being more likely to carry one or two variant alleles (76.2% vs 45.4%; p=0.011). Chronic low levels of vitamin K, influenced by CYP4F2 genotype, might affect bone development and vascular health in children on warfarin. The link between warfarin treatment and vascular calcification was further explored in the same cohort of children, by assessing plasma uc-MGP concentration, a sensitive biological marker of vascular calcification. No association was found between plasma uc-MGP concentration and either anticoagulation intensity, duration of warfarin therapy, or warfarin daily dose. Patients on warfarin scheduled for surgery have to withdraw treatment for 5 days to avoid peri-operative bleeding. A previously developed algorithm that included genetic, demographic and clinical data was validated using an independent cohort of patients (n=117) who completed a short course of warfarin. There was a strong and highly significant correlation between the observed and predicted INR values by the algorithm (r=0.949, p<0.001). In an ex-vivo study, rivaroxaban in elderly subjects produced a greater prolongation of PT (p<0.05) and modified PT (p<0.001) and a greater suppression in the rate and amount of thrombin generation compared to young subjects, suggesting an age-related increase in sensitivity to the drug