Reversal of anticoagulant effects of apixaban with non-specific prohaemostatic agents: an in vitro study

Abstract

Background: As any potent anticoagulant, apixaban, an oral factor Xa inhibitor, exposes to a risk of haemorrhagic complications. Thus in the absence of specific antidote, bleeding management is challenging. Aim: To investigate the efficacy of recombinant activated factor VII (rFVIIa), prothrombin complex concentrate (PCC), and partially activated PCC (aPCC) to compensate, in vitro, the anticoagulant effects of apixaban. Methods: Whole blood (WB) from healthy volunteers was spiked with therapeutic (200 ng.ml-1) or overdose (680 ng.ml-1) amount of apixaban and either rFVIIa (equivalent to 90 and 120 μg.kg-1), PCC (25 and 50 UI.kg-1), or aPCC (80 and 160 UI.kg-1). Reversal was assessed on WB with thromboelastography, and on platelet poor plasma using standard laboratory assays (PT, aPTT) as well as thrombin generation test (TGT). Increase of clot turbidity was also recorded to explore the dynamics of fibrin polymerization. Results: Apixaban concentration dependently lengthened PT, aPTT, and the clotting time (CT) in ROTEM®. It also increased lag time (LT) and decreased endogenous thrombin potential (ETP) as well as peak height in the TGT. Apixaban delayed and slowed down fibrin formation and polymerization with subsequent alteration of the resulting clot. rFVIIa had predominant effects on kinetics parameters. Addition of rFVIIa overcorrected PT, aPTT, and CT at apixaban therapeutic concentration and partly restored them in overdose situation. rFVIIa also decreased LT, and accelerated fibrin polymerization at therapeutic dose. Addition of PCC shortened PT and CT, increased peak height and ETP, but paradoxically lengthened aPTT and little modified the turbidity profile. Addition of aPCC shortened PT, aPTT, and LT in the TGT, whereas it did not affected CT in ROTEM®. aPCC also increased peak high, overcorrected ETP, and induced earlier and faster fibrin polymerization. Conclusions: These non-specific prohaemostatic agents corrected partially or completely several laboratory parameters. In particular aPCC, improving most of them was the most seducing. Further investigation is nevertheless necessary: reversal of aPCC was not confirmed in WB, increasing concentration of the prohaemostatic agents had no additive effect, and putative efficacy was strongly dependent on the observed parameter (rFVIIa was best on CT; PCC on ETP). Overall our observations underline the need for robust endpoints and clinical trials.