Stent-grafts are endovascular devices composed of a metal stent backbone to provide radial strength and affix the device to the vascular wall, covered by a synthetic polymer membrane to form a conduit for blood flow [1]. Stent-grafts are used to treat many vascular pathologies including aortic dissection [2] and aneurysms [3], iliac artery atherosclerosis [4], and carotid artery pseudoaneurysms (PSAs) [[5], [6], [7], [8]]. Although stent-grafts are regularly used in the clinic, there are several complications associated with these devices [2]. Their deployment within the artery immediately results in a large surface area of metal and synthetic material in contact with the blood, increasing the risk of device-related thrombosis [9]. Furthermore, this large surface area may delay healing (i.e. establishment of an endothelium) and stimulate neointimal hyperplasia and stenosis [10,11].

The endothelium is a layer of cells that lines the luminal surface of the blood vessels. Endothelial cells act as a non-thrombogenic surface by expressing membrane proteins, such as thrombomodulin, and releasing molecules, such as nitric oxide, that prevent the aggregation of platelets and the formation of thrombus [12]. Endothelial cells also regulate smooth muscle cell proliferation [13] and mediate inflammation by recruiting circulating immune cells [14], both of which play a role in vascular remodeling in response to injury or stent placement. The coverage of blood contacting devices with a functional endothelium, forming a barrier between the device and blood flow, is associated with a decreased risk of device induced thrombosis [15,16] and stenosis [17]. Therefore, promoting the rapid development of an endothelium is a key principle in endovascular device engineering [[18], [19], [20]]. There are many approaches under investigation to promote the rapid endothelialization of endovascular devices. These approaches range from simply minimizing device surface area and streamlining device geometry [21] to selecting materials more suitable for endothelial cell adhesion and proliferation [22]. However, these approaches may have negative implications for the mechanical performance of the device. More complicated approaches include applying bioactive coatings to the device surface with the aim of capturing circulating endothelial progenitor cells and promoting endothelial cell proliferation [23,24]. However, this may result in the non-specific capture of circulating platelets or inflammatory cells, or the excessive proliferation of smooth muscle cells, all of which are unfavorable outcomes [25]. Other groups have investigated pre-seeding the device with cells in vitro, forming a built-in cell layer prior to deploying the device [[26], [27], [28]]. However, this approach has the logistical issue of timing cell seeding with the patient procedure, and the cell layer may be damaged during the device deployment process. In the present study, we investigate a magnetic cell capture technique to promote the rapid endothelialization of stent-grafts. In this approach, cells are delivered specifically to the device surface immediately after it has already been deployed, avoiding drawbacks of the other techniques described above [25,29].

In the magnetic cell capture approach, autologous blood outgrowth endothelial cells (BOECs) are generated from peripheral blood and dosed with superparamagnetic iron oxide nanoparticles (SPIONs). Endovascular devices are then manufactured from magnetizable materials. Once the device is deployed within the artery, the BOECs are injected into its lumen and magnetically adhere to the device surface, facilitating the rapid generation of an endothelium. A schematic of this approach is presented in Fig. 1. Our group has previously established proof of concept for this approach using bare metal stents in porcine coronary arteries [30]. We also tested this approach for stent-grafts composed of a magnetic stent backbone with an electrospun non-magnetic polyurethane covering. We found that the cells only magnetically adhered to areas of the stent-graft immediately adjacent to the stent struts, which was insufficient for improving the healing response [31]. Recently, we developed a technique to incorporate magnetic stainless steel microparticles into an electrospun polyurethane lattice, resulting in a polyurethane-based stent-graft material capable of uniform magnetic capture of BOECs in vitro [32]. In the current work, we construct magnetic stent-grafts out of magnetic stent backbones and the electrospun magnetic polyurethane-based material. The ability of these stent-grafts to occlude PSAs, magnetically capture and retain cells, and the ensuing healing of the device is compared to non-magnetic stent-graft controls in a porcine carotid PSA model.