Electronic Thesis and Dissertation Repository


Doctor of Philosophy


Physiology and Pharmacology


Kim Richard B.


Inhibition of blood coagulation via oral anticoagulant therapy is the mainstay for preventing a cardioembolic stroke in patients with atrial fibrillation (AF). Factor Xa inhibitors (FXaIs), rivaroxaban and apixaban, represent a new class of oral anticoagulants that are now widely prescribed in AF patients as an alternative to traditional warfarin therapy. An important advantage of these drugs is that routine monitoring of anticoagulation response is not necessary. Nevertheless, because of their mechanism of action, FXaI antithrombotic effect can be inferred based on the observed drug plasma concentration, with prolonged periods of elevated FXaI systemic exposure associated with increased risk for major bleeding.

Currently, there is a paucity of data relating to observed interpatient variation in FXaI plasma concentrations in the post-market clinical setting. Given that the patient population that take FXaIs in the post-market setting may vary greatly with their co-morbidities and co-medications from those within clinical trials, we hypothesized the systemic FXaI exposures achieved in these individuals may vary greatly as well, and that this interpatient variation is driven by patient-specific factors (physiology, disease-states, co-medications, genetics).

In the first study, we determined rivaroxaban and apixaban plasma concentrations in a cohort of 243 AF patients during routine clinic visits, and found their measured FXaI concentration tended to be more variable than those observed in clinical trials. Approximately 12% of patients receiving rivaroxaban and 13% of patients receiving apixaban exceeded the predicted maximum FXaI plasma concentration observed in clinical trials.

In the second study, we characterized the observed variation in FXaI concentration among our apixaban-treated AF cohort through regression analysis. Age and renal function accounted for the majority of the explained variation, while female sex and amiodarone use were also significant predictors of apixaban exposure. In addition, we demonstrated, for the first time, that 4β-hydroxycholesterol plasma concentration, as a potential biomarker for apixaban metabolism, was an independent predictor of interpatient variation in apixaban concentration.

Equally important, we noted in our second study, that a large proportion of the interpatient variation remained unaccounted for, and known common genetic variation in metabolizing enzymes and efflux transporters did not significantly predict apixaban concentration within our cohort. Therefore, we applied a custom targeted next-generation exome sequencing approach (PGxSeq) to potentially identify rare or patient-specific single nucleotide variation (SNV) in subjects with unexpectedly high drug concentration that result in aberrant function or expression of apixaban metabolizing enzymes or transporters. Following successful development and validation of PGxSeq, our exploratory analysis of twelve apixaban-treated AF subjects that were sequenced, revealed rare and common SNV(s) within the selected candidate genes.

Taken together, these studies provide insight into the factors that drive variation in FXaI plasma concentration among AF patients in routine care, and serve as a framework for future investigations regarding personalization of FXaI anticoagulant therapy.