Introduction Efficacy and safety of devices for the endovascular recanalization of acute ischemic stroke (AIS) are mostly validated in experimental vascular models and in vivo animal models. Efforts have been made to develop mechanically stable clot models to explore the interactions between the clots and devices; however, little is known about the mechanical similarities between these clot models and the human sources of embolic thrombi. The aim of this study is to explore the structure and mechanical properties of the possible sources of the cerebral emboli extracted from patients and model clots produced in vitro using human, porcine and bovine donors.

Materials and methods Six thrombi were collected during carotid endarterectomies (CEA) and two thrombi were obtained from the internal carotid artery (ICA) of AIS patients. In vitro clots were prepared by simultaneously injecting whole blood/ACD mixture and CaCl₂/thrombin solution into silicone tubing. Three variables of in vitro clotting included species, thrombin concentration and addition of the radio-opaque component, barium sulfate. Dynamic mechanical analyzer was used to acquire the stress–strain behaviors of the clots in the controlled force mode which gave an indication of clot hardness. Clot elasticity was studied in stress–relaxation mode by recording the strain recovery. Scanning electron microscopy (SEM) and Martius Scarlet Blue stained sections were used to investigate the structure and composition of the specimens.

Results Two secant moduli, designated E1 and E2, are calculated from the toe region (initial to 75% strain) and high strain (75–95%) of the stress–strain curve, respectively. The calcified thrombi from CEA (n=2) had lower elongation and higher moduli (E1=737.8±296.9 Pa, E2=12313±165.7 Pa) whereas red cell rich thrombi (n=4 from CEA and n=2 from ICA) showed higher elongation and lower moduli (E1=585.7±98.37 Pa, E2=6651±675 Pa). Interspecies differences in the mechanical properties of clot models were observed. Overall, bovine clots had the highest hardness followed by porcine and human clots. Addition of thrombin significantly increased E1 and E2 of the human and porcine clot models (p<0.05). In the absence of thrombin and barium sulfate, the in vitro bovine clots had strain recovery of 30% whereas human clots were too fragile to measure the strain recovery. Less clot shrinkage and a significant decrease in strain recovery were found in the presence of barium sulfate. The histology and SEM results showed that the clots prepared in vitro were mainly composed of homogeneously dispersed erythrocytes with several interspersed fibrin bands. The human thrombus demonstrated a fibrin matrix with some confined areas rich in erythrocytes. Platelet aggregation, leukocyte deposits and other cell debris were also seen.

Conclusion In this research we move beyond studying composition of thromboemboli by additionally investigating the mechanical properties of these sources of AIS. As newer generations of endovascular recanalization devices become available, it will be important to create a variety of characterized clot models to study the safety and efficacy of the devices preclinically. More human cerebral thrombi will be analyzed to provide a library of mechanical properties to construct a framework for further development of clot models.