Abstract Objective. Continued expansion of abdominal aortic aneurysms (AAA) in the absence of an endoleak is thought to result from endotension. It has been observed more frequently with porous polytetrafluoroethylene (ePTFE) stent grafts and may be the result of the transudation of serous blood components and pressure through the stent graft. In this study, a canine model of AAA was used to compare intra-aneurysmal pressure and thrombus formation after exclusion with Dacron and ePTFE stent grafts. Methods. Prosthetic AAA with implanted strain-gauge pressure transducers were treated by stent graft exclusion using FDA-approved devices in 10 mongrel dogs: 5 Dacron (AneuRx, water integral permeability = 211 ± 26 ml/min/cm2) and 5 ePTFE (original Excluder, inter-nodal distance = 25 m). No endoleaks were present. Intra-aneurysmal pressure was measured over 4 weeks after AAA exclusion and indexed to the systemic pressure, represented as a percentage of a simultaneously obtained systemic pressure (value = 1.0). Magnetic resonance imaging (MRI) with densitometry of the intra-aneurysmal thrombus was performed at 1, 2, and 4 weeks after exclusion and expressed as a signal noise ratios (S:N) to control for background signal intensity. Intra-aneurysmal thrombus was characterized histologically, using hematoxylin and eosin, trichrome and Mallory’s phosphotungstic acid hemoxylin (PTAH) stains. Results. In animals excluded with both Dacron and ePTFE stent grafts, the intra-aneurysmal pressure was non-pulsatile and reduced to less than 30% of systemic pressure. Significantly greater pressure transmission was observed after AAA exclusion, using ePTFE compared to Dacron stent grafts (systolic pressure: ePTFE, 0.28 ± 0.12 versus Dacron, 0.11 ± 0.02, P Introductions Endovascular stent grafts are used to treat abdominal aortic aneurysms (AAA) by excluding the aneurysm sac from arterial circulation. Endotension has been proposed to account for the continued expansion of the aneurysm sac after endovascular repair in the absence of perfusion or endoleak.1 This is associated with a nonpulsatile pressure that is approximately one third of systemic pressure. Preliminary clinical data suggest that stent grafts constructed from expanded polytetrafluoroethylene (ePTFE) may result in continued expansion of the AAA despite complete exclusion of the aneurysm from the arterial circulation, as evidenced by the absence of an endoleak.2–4 In addition, aneurysm sac hygromas with the appearance of a gelatinous sac content have been described with ePTFE grafts in several case reports.5–7 This study sought to compare intra-aneurysmal pressure and content after exclusion with Food and Drug Administration (FDA)-approved Dacron and ePTFE stent grafts in a canine model of AAA. Methods Prosthetic AAA with implanted strain-gauge pressure transducers were treated by stent graft exclusion using FDA-approved devices in 10 mongrel dogs: 5 Dacron (AneuRx AAA stent graft system, Medtronic, Santa Rosa, Cal.) and 5 porous ePTFE (original Excluder, WL Gore and Associates, Flagstaff, Ariz.). Each of the dogs weighed 25–35 kilograms. The procedures and handling of the animals were reviewed and approved by the Institutional Animal Care and Use Committee at Weill Medical College of Cornell University. All procedures were in accordance with the “Guide for the Care and Use of Laboratory Animals” by the Institute of Laboratory Animal Resources, Commission on Life Sciences.8 Pressure transducer and prosthetic aneurysm creation. An implantable, solid-state, strain-gauge pressure transducer (Konigsberg Instruments, Pasadena, Cal.) was utilized for daily monitoring of systemic and intra-aneurysmal pressure. The accuracy of the transducer has been confirmed in various media in vitro and in vivo, including liquid, gelatinous and solid environments.9,10 Continuous pressure monitoring and data storage were performed using data recording software from Data Integrated Scientific Systems (DISS, Detroit, Mich.). Prior to implantation and following explantation, calibration of the transducer was performed. Daily systemic and intra-aneurysmal pressures were taken for four weeks after aneurysm creation and exclusion. Balloon dilation was used to create the prosthetic aneurysm from thin walled polytetrafluoroethylene (8 mm diameter, Impra, Bard Peripheral Vascular, Tempe, Ariz.) prior to implantation. The final diameter of the prosthetic aneurysm was 30 mm. The pressure transducer was approximated to the luminal surface of the prosthetic aneurysm and secured in place (Figure 1). Implantation of prosthetic aneurysm and pressure transducers. The animals were fasted overnight, and on the day of surgery were anesthetized (induction, thiopental 8mg/kg; maintenance, isofluorane 2%), intubated and ventilated. The abdominal aorta from the level of the renal arteries to the trifurcation was exposed via a midline abdominal incision. After obtaining proximal and distal control, the aorta was completely transected. The prosthetic aneurysm (diameter = 30 mm) containing an intraluminal strain gauge pressure transducer was sewn to the transected ends of the aorta as an interposition graft (Figure 2). A second strain-gauge pressure transducer was placed in the native aorta proximal to the aneurysm. Systemic anticoagulation with unfractionated sodium heparin (2000 units intravenously) was maintained throughout the period of arterial clamping. Exclusion of the aneurysm. Aneurysm exclusion was performed during the same procedure (Figure 3), via the midline laparotomy incision using one of two different stent grafts (Dacron- five animals, ePTFE- five animals). Exclusion was performed via the midline laparotomy incision because the delivery system (16 Fr) for the stent grafts is larger than the canine femoral artery. The FDA-approved Dacron stent graft used in this study was the AneuRx AAA Stent Graft System (Medtronic, Santa Rosa, Cal.), which is comprised of an exoskeleton of nitinol over a polyethylene terephthalate (Dacron) graft material. The Dacron stent graft porosity is defined by its water integral permeability of 211 ± 26 ml/min/cm2. The FDA-approved ePTFE stent graft utilized in this study was the original Excluder (W. L. Gore & Associates, Flagstaff, Ariz.) made of nitinol-covered ePTFE. Because ePTFE is hydrophobic, the internodal distance (25 um) is used to measure permeability. In five animals, a Dacron stent graft (12 mm x 5.5 cm) was deployed through the distal native aorta without an introducer sheath under fluoroscopic guidance. The remaining five animals had an ePTFE stent graft (12 mm x 7 cm) inserted via a 14 Fr sheath in the distal native aorta, under fluoroscopic guidance. Intraoperative angiography was performed in all cases, after stent graft deployment, to confirm the integrity of the exclusion from antegrade perfusion (Figure 4). The peritoneal cavity was closed, and the cables from the pressure transducers were tunneled laterally through the abdominal wall and tracked subcutaneously to exit the skin between the scapulae. The animals were extubated and allowed to recover overnight. All animals received clopidogrel (75 mg/day, Sanofi-Synthelabo Pharmaceuticals, New York, NY) starting the day of surgery and continuing until euthanasia. Magnetic Resonance Imaging Prior to magnetic resonance imaging (MRI), the animals were premedicated, anesthesized, intubated and mechanically ventilated in a manner similar to preparation for operative procedures. T1 weighted MRI images and gadolinium-enhanced magnetic resonance angiography (MRA) images were obtained using a fast spoiled gradient recalled protocol in a 3.0 Tesla MRI system (GE Medical Systems, Waukesha, Wis.) at 1, 2 and 4 weeks after aneurysm exclusion. Post processing with a GEMS 4.0 Fiesta video imaging software was used to analyze images. A localizer scan was obtained in the coronal, axial and sagittal planes and was used to determine subsequent imaging in the axial plane for T1 weighted images and cine-gated MRA. T1-weighted images were obtained from the renal arteries to the abdominal aortic trifurcation using double inversion recovery fast spin echo sequences. Parameters of T1 weighted images were TR, 350 ms; TE, 14 ms; FOV, 20 cm; and matrix 256 x 256. The T1 weighted images were used to analyze the characteristics of the thrombus, quantified by determining the signal to noise ratio (S:N). The S:N was calculating as the ratio of the signal in the area of the thrombus, standardized with the signal intensity of the air in the same image. A radiologist blinded to the histological results determined the signal intensity of the thrombus in at least three different sections for each time point. Euthanasia Animals were euthanized 30 days after aneurysm exclusion via injection of Sleepaway. Intraoperative angiography confirmed the absence of endoleaks prior to euthanasia. Histology The aneurysm was perfusion fixed in 3% glutaraldehyde buffered to pH 7.4 with sodium cacodylate for histopathologic investigation. The aorta was divided longitudinally to remove the stent graft prior to paraffin embedding. Transverse, longitudinal and oblique sections of the aneurysm content were embedded in paraffin. Intra-aneurysmal thrombus was characterized histologically using hematoxylin and eosin (H&E), and Trichrome and Mallory’s phosphotungstic acid hemoxylin (PTAH) stains. Statistics Intra-aneurysmal pressure measurements were indexed to the systemic pressure, which was obtained simultaneously. The intra-aneurysmal pressure is represented as a percentage of the systemic pressure, with the systemic pressure having a value of 1.0. Results are represented as the mean ± standard deviation. Continuous variables were analyzed using the student t-test, with statistical significance assumed at P £ .05. Results In all animals, complete exclusion from antegrade perfusion was achieved and no endoleaks were present, which was confirmed by MRI (throughout the study) and on angiography at the time of euthanasia (Figure 5). In animals excluded with both Dacron and ePTFE stent grafts, the sac pressure was non-pulsatile and reduced to less than 30% of systemic pressure (Table 1). In aneurysms excluded with ePTFE stent grafts, the systolic pressure within the aneurysm sac after exclusion was reduced to 0.28 ± 0.12 compared with 1.0 for systemic pressure. Aneurysms excluded with Dacron stent grafts, had significantly lower intra-aneurysmal sac systolic pressure of 0.11 ± 0.02 (1.0 for systemic pressure compared with ePTFE stent grafts, p 2 In the multi-center trial of the original Excluder, only 19% of patients had AAA sac regression, while 14% had AAA sac enlargement at 2 years.3 A single institute study of patients treated with the original Excluder found that the probability of freedom from sac growth or re-expansion at 4 years was only 43%.4 AAA sac hygromas with the appearance of viscous, gelatinous sac content have been described with ePTFE grafts in several case reports following open and endovascular repair. The formation of these “seromas” around prosthetic (PTFE) grafts used for hemodialysis access and subclavian artery-pulmonary artery shunts has long been known and reported since the early 1980’s.11 Williams described a 12 cm cystic mass filled with a seroma like fluid surrounding a patent PTFE graft 21/2 years after open AAA repair and was one of the first to hypothesize ultrafiltration through the graft as a potential cause.11 Risberg described 3 cases of aneurysm sac hygromas after repair of AAA using PTFE grafts; 2 occurred after endovascular repair, one after open repair. Analysis of the fluid drained from these hygromas indicated activation of both the coagulation and fibrinolytic systems.12 An aneurysm sac hygroma was discovered 18 months after thoracic endograft repair (ePTFE stent graft) when the patient became dyspneic.5 A rupture related to an hygroma was described in case series of five patients with symptomatic sac enlargement after open AAA repair with PTFE grafts. During laparotomy, a seroma with gelatinous material was discovered that histologically revealed degenerate red blood cells and amorphous eosinophilic material.6 In this study, AAA excluded with ePTFE stent grafts resulted in thrombi that histologically were consistent with an acute thrombus with red blood cell fragments and disorganized fibrin, while thrombi of AAA excluded with Dacron stent grafts were mostly composed of collagen or granulation tissue consistent with a chronic thrombus. In the aneurysm sac, thrombus formation is influenced by stasis of blood that occurs after the aneurysm is excluded from antegrade flow. This stasis leads to fibrin deposition and also prevents the inflow of inhibitors of clotting factors, which helps allows the formation of an organized thrombus. Without the inflow of fibrinolytic factors, the thrombus in the aneurysm sac is eventually converted into a subendothelial mass of connective tissue that can potentially recanalize.12 MRI has been used to help identify the stages of thrombus formation in both animal models and human disease.13 This knowledge is based on studies of acute cerebral hemorrhage, which demonstrated that the changing appearance of MRI signal intensity is secondary to the interactions of the molecules of hemoglobin degradation (oxyhemoglobin, deoxyhemoglobin, methemoglobin, ferritin and hemosiderin) and the changing hydration state of the hemorrhage.14,15 A swine model of acutely formed carotid thrombi found that the evolution of MRI appearance correlated to the histological progression of the thrombus (red blood cells and fibrin —> cellular debris —> connective tissue —> organized collagen). The study found that the MRI signal intensity demonstrated a rapid increase followed by a decline as thrombus became stable.16 In this study, a similar pattern of increase in signal intensity followed by decline over time was seen in those aneurysms excluded with Dacron stent grafts, while the signal intensity remained elevated in aneurysms excluded with ePTFE stent grafts which correlated with the different histological findings. AAA excluded by ePTFE stent grafts were found to have significantly higher aneurysm sac pressures than those that were excluded with Dacron stent grafts without any evidence of endoleaks. Endotension is proposed to account for continued aneurysm expansion after endovascular repair in the absence of aneurysm perfusion.1 In a small case series, endotension was found to be a non-pulsatile sac pressure that was approximately one third of systemic pressure.7 Potential causes of endotension include degradation of the aneurysm sac thrombus and ultrafiltration through the graft material. Sjoren et al. postulated that the pressure in aneurysm sacs may trigger an up-regulation of tissue plasminogen activator, increasing the fibrinolytic activity and resulting in the lysis of the thrombus.16 The increased pressure in the aneurysm sac from endotension has the potential to result in an increase in the aneurysm sac size, which in multiple studies has been seen to more commonly occur after repair with the original Excluder than grafts made of Dacron, including the AneuRx stent graft.2–4 Reports of rare ruptures after endograft repair have been associated with most stent grafts, and are usually secondary to fixation of the aortic neck or iliac artery or patient refusal to have treatment for known endoleaks.17 Graft material and porosity have significant implications regarding the endovascular repair of AAA. The porosity of the prosthetic graft material is critical to its ability to create a hemostatic seal and prevent hemorrhage through the graft. Previous experimental research utilizing a canine model of AAA evaluated the effect of stent graft porosity on pressure transmission to the aneurysm sac after treatment with an endovascular stent graft.10 In this study, non-porous stent grafts were demonstrated to reduce pressure transmission to the aneurysm sac to less than 10% of systemic pressure. In contrast, the highly porous co-knit graft resulted in transmission of pressure that was 80% of systemic pressure.19 Increased transudation of fluid through the ePTFE stent is caused by characteristics specific to the material. Normally, ePTFE is hydrophobic, but may become “wettable” if exposed to certain agents, such as betadine or blood.12 Finally, some studies have shown that transudation of fluid through PTFE may occur because of failure of the fibroblasts to incorporate the graft.20 In this animal model of AAA, we found that endovascular treatment with either Dacron or ePTFE stent grafts reduces the mean intra-aneurysmal pressure to a non-pulsatile pressure that is less than 30% of systemic pressure. However, there was significantly greater pressure present in aneurysms treated with porous ePTFE stent grafts than Dacron grafts. Histology and MRI suggest transudation of serous blood components into aneurysm sacs treated using ePTFE stent grafts that may contribute to increased intra-aneurysmal pressure. This transudation of serous blood components may explain the clinical finding of expansion of the aneurysm sac in the absence of an endoleak after treatment with ePTFE stent grafts. However, further studies will be necessary to completely characterize the clinical significance of this endotension.