A Combination of Ex Vivo and In Vivo Strategies for Evaluating How Much New Oral Anticoagulants Exacerbate Experimental Intracerebral Bleeding

Abstract

Background Intracerebral hemorrhage is the most serious complication of anticoagulant therapy but the effects of different types of oral anticoagulants on the expansion of these hemorrhages are still unclear. Clinical studies have revealed controversial results; more robust and long-term clinical evaluations are necessary to define their outcomes. An alternative is to test the effect of these drugs in experimental models of intracerebral bleeding induced in animals. Aims To test new oral anticoagulants (dabigatran etexilate, rivaroxaban, and apixaban) in an experimental model of intracerebral hemorrhage induced by collagenase injection into the brain striatum of rats. Warfarin was used for comparison. Methods Ex vivo anticoagulant assays and an experimental model of venous thrombosis were employed to determine the doses and periods of time required for the anticoagulants to achieve their maximum effects. Subsequently, volumes of brain hematoma were evaluated after administration of the anticoagulants, using these same parameters. Volumes of brain hematoma were evaluated by magnetic resonance imaging, H&E (hematoxylin and eosin) staining, and Evans blue extravasation. Neuromotor function was assessed by the elevated body swing test. Results and Conclusions The new oral anticoagulants did not increase intracranial bleeding compared with control animals, while warfarin markedly favored expansion of the hematomas, as revealed by magnetic resonance imaging and H&E staining. Dabigatran etexilate caused a modest but statistically significant increase in Evans blue extravasation. We did not observe significant differences in elevated body swing tests among the experimental groups. The new oral anticoagulants may provide a better control over a brain hemorrhage than warfarin.

Keywords: apixaban; dabigatran; intracerebral hemorrhage; rivaroxaban; warfarin.

Fig. 1

Experimental approach used to assess the effect of NOACs and warfarin on the expansion of ICH. The anticoagulants were orally administered to rats (1 or 1.5 mg/kg of warfarin, 18 mg/kg of apixaban or rivaroxaban, or 9 mg/kg of dabigatran etexilate). After the period required for the anticoagulant to achieve its maximum effect (∼1 hour for the NOACs and 24 hours for warfarin), collagenase (0.4 U) was injected into the left brain striatum. After 24 hours, the brains of the animals were analyzed by MRI, histological examinations, and Evans blue extravasation. Functional neurological assessment was performed using the elevated body swing test. MRI, magnetic resonance imaging; NOACs, new oral anticoagulants.

Fig. 2

Profiles of the anticoagulant and antithrombotic activities and bleeding tendency after oral administration of anticoagulants to rats. ( A ) Rat plasmas were collected at different times after oral administration of 9 mg/kg dabigatran etexilate (blue); 18 mg/kg apixaban (green); 18 mg/kg rivaroxaban (black); and 1 mg/kg warfarin (red). Anticoagulant activity based on TT (blue line) or PT (red, black, and green lines) was expressed as T ₁ / T ₀ , which is the ratio of clotting time in the presence or absence of drug ( n  = 5, mean ± SEM, * p  < 0.05 versus control). ( B ) Dose dependence of the antithrombotic effect of warfarin and NOACs on the venous thrombosis model using thromboplastin as thrombogenic stimulus. ( C ) Assessment of bleeding tendency caused by oral anticoagulants based on the amount of blood loss after a tail injury ( n  = 6, mean ± SEM, * p  < 0.05 vs. control). The same doses of the anticoagulants were used in the assays of panels (A) and (C). NOACs, new oral anticoagulants; PT, prothrombin time; SEM, standard error of the mean; TT, thrombin time.

Fig. 3

T1-weighted sequences obtained in a 2-Tesla MRI scanner 24 hours after ICH induced by collagenase injection. Representative images obtained for nontreated animals ( A ), warfarin at the dose of 1.0 ( B ) and 1.5 mg/kg ( C ), 9 mg/kg dabigatran etexilate ( D ), 18 mg/kg rivaroxaban ( E ) or apixaban ( F ) are shown in the panels. Red circles show the contour of the hematoma made using MIPAV software. Panel ( G ) shows the quantification of hematoma volume based on T1-weighted ( n  = 8, mean ± SEM, * p  < 0.05 vs. nonanticoagulated group). ICH, intracranial hemorrhage; MRI, magnetic resonance imaging.

Fig. 4

Images of the histological analysis of the hemorrhagic areas observed on nontreated ( A ) or anticoagulated ( B–F ) animals using H&E staining. Quantitative data are shown in panel ( G ). The anticoagulant doses were 1 mg/kg warfarin (B), 1.5 mg/kg warfarin (C), 9 mg/kg dabigatran etexilate (D), 18 mg/kg rivaroxaban (E), or apixaban (F). (G) Quantification of the hemorrhagic lesion (%) in relation to the whole ipsilateral hemisphere ( n  = 3, mean ± SEM, * p  < 0.05 vs. nonanticoagulated group). SEM, standard error of the mean.

Fig. 5

Evans blue extravasation from the brains of rats treated with saline or oral anticoagulants after 24 hours of collagenase-induced ICH. Evans blue was injected intravascularly and allowed to circulate for 30 minutes. After perfusion, animal was sacrificed and the amount of Evans blue extravasated into the brain was quantified by spectrophotometric analysis. Panel ( A ) shows representative macroscopic images with the left hemisphere uppermost in each pair, while quantitative data are in panel ( B ) ( n  = 8, mean ± SEM, * p  < 0.05 vs. nonanticoagulated group). ICH, intracranial hemorrhage; SEM, standard error of the mean.

Fig. 6

Assessment of motor function by the elevated body swing test. Percentage of right turns in the elevated body swing test was assessed 24 hours after induction of the ICH model ( n  = 8, mean ± SEM, * p  < 0.05 vs. sham group). ICH, intracranial hemorrhage; SEM, standard error of the mean.