Characterization of an Endovascularly-induced Model of Deep Vein Thrombosis in Domestic Swine

Presenter: Jose Negron-Garcia, Cook Research Incorporated

Background: In vivo models of deep vein thrombosis (DVT) have been described in rodents, non-human primates, and swine. However, these models typically require major surgery or they are not suitable for testing medical devices intended for humans. A large-animal model of DVT produced by minimally invasive methods may result in a more clinically-relevant model.

Purpose: In this series of model development studies, we evaluated the ability to produce both primary and in-stent DVT in domestic swine between 2 and 28 days.

Methods: Vascular access was obtained in the left external jugular and bilateral femoral veins. Following baseline venography, tributary branches to the right external iliac vein were embolized. A balloon designed to deliver a current was advanced to the right iliofemoral vein and activated for 15 minutes. After current activation, the target vein was occluded proximally and distally using balloon catheters and thrombin was injected into the occluded region. The thrombin dwelled in the occluded region for approximately three hours. Following completion venography, the animals were recovered for approximately 2, 7, 14 or 28 days. Animals in the 14- and 28-day groups underwent interim venography and transcutaneous ultrasonic imaging approximately halfway through the follow-up period. At the conclusion of the follow-up period, venography and transcutaneous ultrasound were again performed, in addition to intravascular ultrasound whenever possible. Additional model characterizations conducted with a subset of the animals included the deployment of a venous stent prior to thrombus induction to test the ability to produce an in-stent thrombosis variant as well as evaluation by experienced interventionalists, who used a variety of interventional devices and techniques to attempt to cross and treat the occlusions. At the conclusion of the testing, animals were humanely euthanized. The treated veins and contralateral controls were collected for gross and histological evaluation.

Results: A total of 25 animals successfully underwent interventional procedures to induce thrombus formation. Technical success (≥ 80% lumen occlusion on immediate postoperative venography) was achieved in all animals. At 28 days, lumen obstruction ranged from 65-95% with average obstruction of 75%. Qualitative histopathologic evaluation revealed that in early timepoints thrombus was immature, with minimal to no remodeling and some fibrin deposition. In later timepoints, thrombus maturation was observed with cellular infiltration and inflammation progressing from the wall towards the center of the thrombus and by day 28, thrombus was very mature with abundant fibrovascular tissue and the appearance of multiple vascular channels within.

Conclusions: The porcine model of endovascular thrombus induction is a suitable minimally invasive approach to produce acute and subacute DVT that allows interventions to test the safety and efficacy of devices in development.