A Method for Creating Artificial Thrombi In Vitro Using a Rotating Mechanical Surface

Effect of Chandler loop shear and tubing size on thrombus architecture

Thrombosis can lead to a wide variety of life-threatening circumstances. As current thrombolytic drug screening models often poorly predict drug profiles, leading to failure of thrombolytic therapy or clinical translation, more representative clot substrates are necessary for drug evaluation. Utilizing a Chandler loop device to form clot analogs at high shear has gained popularity in stroke societies. However, shear-dependent clot microstructure has not been fully addressed and low shear conditions are often overlooked. We herein characterized the impact of wall shear rate (126 to 951 s^-1) on clot properties in the Chandler loop. Different revolutions (20-60) per minute and tubing sizes (3.2 to 7.9 mm) were employed to create different sized clots to mimic various thrombosis applications. Increased shear resulted in decreased RBC counts (76.9 ± 4.3% to 17.6 ± 0.9%) and increased fibrin (10 to 60%) based on clot histology. Increased fibrin sheet morphology and platelet aggregates were observed at higher shear under scanning electron microscope. These results show the significant impact of shear and tubing size on resulting clot properties and demonstrate the capability of forming a variety of reproducible in-vivo-like clot analogs in the Chandler loop device controlling for simple parameters to tune clot characteristics.

Histologic features of thrombosis events with a centrifugal left ventricular assist device

BACKGROUND

Histology of thrombosis events in left ventricular assist devices (LVADs) may point to differences between the etiology of either ingested or de novo thrombus formation within LVADs. Materials ingested by the pump would have features suggestive of lifting and folding, whereas thrombi formed de novo would have uniform, parallel layers. This study tested this hypothesis in a cohort of explanted HeartWare Ventricular assist devices (HVADs) (Medtronic, Miami Lakes, Florida).

METHODS

Histology of thrombi from 59 explanted HVAD pumps were classified as presumed ingested, presumed de novo, or undeterminable on the basis of pre-defined criteria. The apparent size and location of the thrombotic materials were noted.

RESULTS

Histologically, all thrombotic materials were either presumed to be ingested (73%; 95 of 130 total histology cassettes examined) or of undeterminable origin (27%; 35 of 130 histology cassettes). Undetermined origin commonly was due to a lack of sufficient material for analysis. The larger materials (>800 mm³) tended to be in the inflow region. The most common finding was smaller thrombotic materials (<150 mm³) within the pump (64%; 38 of 59 HVADs); when these smaller materials were ingested by the pump, they were most often found within the smaller flow pathways within the pump.

CONCLUSIONS

Our study suggests that the thrombi within HVAD pumps are commonly ingested materials rather than de novo thrombus formation within the pump. Further research to understand the source of this ingested material and the consideration to mitigate this complication should be considered.