A porcine model of thoracic aortic aneurysms created with a retrievable drug infusion stent graft mirrors human aneurysm pathophysiology.

ABSTRACT

Objective: Aneurysm pathophysiology remains poorly understood, in part from the disparity of murine models with human physiology and the requirement for invasive aortic exposure to apply agents used to create aneurysm models. A retrievable drug infusion stent graft (RDIS) was developed to isolate the aortic wall intraluminally for drug exposure. We hypothesized that an RDIS could deliver aneurysm-promoting enzymes to create a porcine model of thoracic aneurysms without major surgical exposure. Methods: Retrievable nitinol stent graft frames were designed with an isolated drug delivery chamber, covered with polytetrafluoroethylene, and connected to a delivery wire with a drug infusion catheter installed to the outer chamber. Institutional Animal Care and Use Committee-approved Yorkshire pigs (n = 5) underwent percutaneous access of the femoral artery, baseline aortogram and stent placement in the thoracic aorta followed by 30-minute exposure to a cocktail of elastase, collagenase, and trypsin. After aspiration of excess drug, stent retrieval, and femoral artery repair, animals were recovered, with angiograms at 1 and 4 weeks followed by explant. Histological analysis, in situ zymography, and multiplex cytokine assays were performed. Results: The RDIS isolated a segment of anterior aorta angiographically, while the center lumen preserved distal perfusion during drug treatment (baseline femoral mean arterial pressure, 70 ± 14 mm Hg; after RDIS, 75 ± 12; P = .55). Endovascular induction of thoracic aneurysms did not require prior mechanical injury and animals revealed no evidence of toxicity. Within 1 week, significant aneurysmal growth was observed in all five animals (1.4 ± 0.1 cm baseline to 2.9 ± 0.7 cm; P = .002) and only within the treated region of the aorta. Aneurysms persisted out to 4 weeks. Aneurysm histology demonstrated loss of elastin and collagen that was otherwise preserved in untreated aorta. Proinflammatory cytokines and increased matrix metalloproteinase activity were increased significantly within the aneurysm. Conclusions: An RDIS achieves isolated drug delivery while preserving distal perfusion to achieve an endovascular porcine model of thoracic aneurysms without major surgery. This model may have value for surgical training, device testing, and to better understand aneurysm pathogenesis. Most important, although the RDIS was used to simulate aortic pathology, this tool offers intriguing horizons for focused therapeutic drug delivery directly to aneurysms and, more broadly, focused locoregional drug delivery to vessels and vascular beds.