The History of the Transjugular Intrahepatic Portosystemic Shunt.

Few, if any, developments in the past three decades have advanced the field of portal hypertension more than the use of transjugular intrahepatic portosystemic shunts (TIPS). Initially pursued in animal studies more than 50 years ago, and discovered serendipitously, TIPS quickly became used clinically in the treatment of refractory esophageal hemorrhage. The technique is now used for many other clinical indications as well as to bridge patients to liver transplantation. Several technical advancements have improved short- and long-term outcomes of the procedure. This article will review the development of TIPS from its inception to the current state of care regarding this important minimally invasive option for patients with portal hypertension.

3D Printed Percutaneous Transhepatic Cholangiography and Drainage (PTCD) Simulator for Interventional Radiology.

PURPOSE: Learning how to perform percutaneous transhepatic bile duct drainage (PTCD) is challenging for interventional radiology (IR) trainees. Therefore, simulators are crucial for IR training and are being increasingly demanded in the evolving healthcare environment of value-based care. To facilitate interventional training, we tried to evaluate our newly developed liver phantom for further use in IR training. METHODS: We developed a liver phantom with a flexible hollow biliary tree, hydrogel-based liver parenchyma, plastic ribs, and silicone skin. The phantom was evaluated by 20 radiology residents from two hospitals. After an introduction, all participants tried to obtain biliary access by fluoroscopic guidance within 25 min. Puncture time, fluoroscopy time, and kerma area product were measured. After 7 days, the participants repeated the procedure on an altered and more difficult model. Additionally, a survey was handed out to every participant (20 residents, 5 experts, and 5 IR fellows) to evaluate the phantom in terms of accuracy and haptic feedback, as well as general questions regarding simulation. RESULTS: The residents performed significantly faster and were more self-confident on Day 7 than on Day 1, significantly decreasing puncture time, fluoroscopy time, and kerma area product (p ≤ 0.0001). The participants were very satisfied with their simulation experience and would trust themselves more in real-life scenarios. CONCLUSION: We were able to develop a phantom with high anatomical accuracy for fluoroscopy and ultrasound-guided interventions. The phantom successfully helped residents learn and improve their PTCD performance.

Interventional treatment of biliodigestive anastomosis leaks with a modified percutaneous transhepatic cholangiodrainage.

INTRODUCTION: Biliodigestive leaks are typically caused by an insufficiency at the surgical anastomosis. Biliodigestive anastomosis (BDA) insufficiencies can lead to bilomas, abscesses and vascular erosion in chronic conditions. MATERIAL AND METHODS: We performed a retrospective analysis of the medical and radiological records of all patients with biliodigestive insufficiency who received interventional treatment between July 2015 and February 2021. Nine patients (three with unilateral drainage and six with bilateral drainage) were treated with a modified percutaneous transhepatic cholangiodrainage (PTCD). Clinical success was considered after complete resolution of the peribiliary collections, absence of bile within the surgical drains, radiological patency of the BDA (contrast medium flowing properly through the BDA and no signs of leakage) and haemodynamic stability of the patient without signs of sepsis. RESULTS: Clinical success was achieved in all nine patients. No patients required revision surgery to repair their BDA. The mean indwelling drainage time was 34.8±16.5 days. The mean number of interventional procedures performed per patient was 6.6±2.0. CONCLUSION: Patients who present with BDA insufficiency may benefit from interventional radiological techniques. Our modified PTCD resolved the BDA leak in all nine cases and should be considered as a valuable option for the treatment of patients with this complication. Our technique demonstrated to be feasible and effective.

Percutaneous transhepatic biliary puncture simulator: a cord network prototype.

BACKGROUND: The aim of this study was to present a percutaneous transhepatic biliary puncture simulator that can be used without radiation exposure and that reflects the conventional anatomy of the biliary ducts and its vicinity structures. METHODS: An anatomically based model of the biliary tree was developed using a cord network fixed to a wooden frame. The skin, ribs, intercostal muscles, and right lower lobe pleura were simulated using foam sponge, plastic tubes, a polystyrene foam panel, and an air pad, respectively. For the puncture, we used a 20-G Chiba needle and a wire with distal double arches; these were used to troll a cord, simulating the successful puncture of a bile duct. A camera was also placed above the model to allow the trainees to train eye-hand coordination while viewing the image on a monitor in real time. The simulator was tested with 60 radiology residents to evaluate the confidence and skills transferability of the training model. RESULTS: After receiving an introduction of the system and 5 min of training under tutor surveillance, all participants were able to troll a cord of the biliary simulator by themselves in less than 4 min. Only one participant punctured the simulated pleura. The participants' evaluations showed positive results, with increased user confidence and skills transferability after the training session. CONCLUSIONS: This proposed simulator can be an effective tool to improve a trainee's confidence and competence while achieving procedural and non-procedural interventional radiology skills related to the liver. TRIAL REGISTRATION: Retrospectively registered.

Temperature-induced configuration changes in hydrogel-coated coils and their relevance in embolization procedures.

BACKGROUND: The present study attempted to demonstrate how the configuration of hydrogel-coated coils is influenced by different temperature exposures. Thirty detachable hydrogel-coated coils were evaluated in an in vitro water immersion test under five different temperature ranges (22.6 °C, 37 °C, 40-50 °C, 50-60 °C, and 60-70 °C). The configuration changes were classified (configuration I, configuration II, and configuration III) according to the curling that occurred during 30 min of immersion. Configuration stability of five Hydrogel-coated coils was also evaluated in a two-step temperature immersion test. RESULTS: All hydrogel-coated coils showed some configuration changes during water immersion. However, a logarithmic transformation of the time and temperature data showed a significant (p < 0.05) negative linear correlation between time and temperature for all coil configurations (configuration I: R = 0.97, configuration II: R = 0.98, configuration III: R = 0.97). The time needed to reach configuration III (complete coiling) was 160.4 ± 41.9 s at 37.5 °C (range: 100-205 s), 45.7 ± 22.2 s at 47.5 °C (range: 23-70 s), 20.2 ± 7.2 s at 57.5 °C (range: 14-32 s), and 10.3 ± 2.4 s at 67.5 °C (range: 7-13 s). CONCLUSIONS: Temperatures above 55 °C induced immediate configurational changes in the hydro-coated coils, achieving complete curling within less than 30 s. Temperatures near 36 °C (normal body temperature) require more time to reach optimal coil curling (configuration III). The optimization of HydroCoil preparation can reduce interventional procedural time and improve clinical results.

Coil embolization of a thoracic aorta hematoma with branch artery pseudoaneurysm - case report.

BACKGROUND: A thoracic aorta hematoma with branch artery pseudonaneurysm is a very rare complication of thoraric blunt trauma. The standard treatment of this type of injury is aortic endograft placement. CASE PRESENTATION: We present a case in which a thoracic aorta hematoma with branch artery pseudoaneurysm was treated with coil embolization instead of endografting. CONCLUSIONS: Coil embolization of aortic injuries may be a safe and definitive treatment alternative in selected cases. This technique has the potential to reduce the risk of procedure-related complications.

Computer Tomography-Guided Percutaneous Indocyanine Green Injection for Intraoperative Mapping of Metastatic Suspected Lesions.

Introduction: Surgical treatment in oncology is one of the main part concerning the surveillance rate of the patient in case of tumor recurrence. Metastatic suspected lesions are mostly located in the abdomen or pelvis and are diagnosed by PET, MRI, or CT scan. Especially surgery of small lesions in recurrent disease for diagnostic or therapeutic purpose is often challenging. Material and Methods: We report a case series of 3 patients who were treated in our department due to a metastatic suspected lesion in PET-CT in follow up. For histological confirmation we performed a laparoscopy using a near infrared camera (NIR) for an improved visualization of the metastatic suspected lesion during surgical treatment. Previously the lesion was marked with an amount of Indocyanine Green (ICG) via computer tomography-guided percutaneous injection. The lesion was identified via NIR camera. While changing the camera in NIR mode, it show up as a blue spot due to the fluorescent signal. After correct identification it was removed and send to pathology. Results: In all 3 cases they confirmed the diagnosis of a metastatic lesion. Complication occur in just one case, where the metastatic lymph node infiltrated the external iliac vein, which led to a high blood loss. In this case a vascular interposition had to be done. Conclusions: Because of separate wavelengths, which are used for illumination and recording, only the marked area is visible, not the background.Due to correct identification, resection of the lesion was improved and healthy surrounding tissue could have been spared.

New Interventional Treatment Model for Pancreatic Neoplasms Using Gemcitabine-Eluting Hydrogel Devices: In Vitro and In Vivo Results.

OBJECTIVES: In vitro and in vivo evaluation of fast- and slow-release gemcitabine-eluting hydrogel (GEH) devices. METHODS: For in vitro elution, the GEH devices were placed in phosphate-buffered saline at 37 °C. Periodically, the solution was analyzed for gemcitabine. The devices consisting of fast release (n = 8), slow release (n = 6), or bland (n = 4) were delivered through a 5-Fr catheter into the gastroduodenal artery of a pig. Additionally, four pigs were treated with intravenous (IV) injection of gemcitabine. Pigs were killed at day 1 (n = 9), day 7 (n = 11), or day 21 (n = 2). Gemcitabine concentrations in the plasma and tissues were determined. RESULTS: In vitro, gemcitabine was completely eluted within 6 h or 30 days for the fast- and slow-release devices, respectively. All 22 pigs were treated without morbidity or mortality. Gemcitabine plasma concentrations peaked at about 105,000 ± 30,000, 252 ± 101, 22 ± 29, and 0 ± 0 ppb for the IV, fast-release, slow-release, and bland treatments, respectively. At days 1 and 7, gemcitabine concentrations were higher in the pancreas for the GEH devices than IV. Gemcitabine delivery to the pancreas was sustained over 21 days in the slow-release group. CONCLUSIONS: Treatment with GEH devices resulted in at least equivalent gemcitabine concentration in the pancreas and reduced concentration in the plasma, heart, liver, and duodenum, at least equivalent to IV injection and reduced concentrations elsewhere. These results show the potential of sustained local delivery of gemcitabine to treat pancreatic neoplasms with reduced side effects.

Disastrous Portal Vein Embolization Turned into a Successful Intervention.

Portal vein embolization (PVE) may be performed before hemihepatectomy to increase the volume of future liver remnant (FLR) and to reduce the risk of postoperative liver insufficiency. We report the case of a 71-year-old patient with hilar cholangiocarcinoma undergoing PVE with access from the right portal vein using a mixture of n-butyl-2-cyanoacrylate and ethiodized oil. During the procedure, nontarget embolization of the left portal vein occurred. An aspiration maneuver of the polymerized plug failed; however, the embolus obstructing portal venous flow in the FLR was successfully relocated into the right portal vein while carefully bypassing the plug with a balloon catheter, inflating the balloon, and pulling the plug into the main right portal vein.

Regional Radiation Pneumonitis After SIRT of a Subcapsular Liver Metastasis: What is the Effect of Direct Beta Irradiation?

We herein present a patient undergoing selective internal radiation therapy with an almost normal lung shunt fraction of 11.5%, developing histologically proven radiation pneumonitis. Due to a predominance of pulmonary consolidations in the right lower lung and its proximity to a large liver metastases located in the dome of the right liver lobe a Monte Carlo simulation was performed to estimate the effect of direct irradiation of the lung parenchyma. According to our calculations direct irradiation seems negligible and RP is almost exclusively due to ectopic draining of radioactive spheres.

The "embowire" technique: concept and description of a novel embolization technique for narrow vessels.

We describe a new, useful embolization technique applied to occlude narrow vessel branches (≤ 1.5 mm (0.06") in diameter) by deployment of one hydrocoil, through a microcatheter, in a way similar to the way in which one might navigate through the vascular lumen with a guidewire.

Hepatic arterial infusion enhances DOTATOC radiopeptide therapy in patients with neuroendocrine liver metastases.

Intravenously administered radiolabeled peptides targeting somatostatin receptors are used for the treatment of unresectable gastroenteropancreatic neuroendocrine tumors (GEP-NETs). Recently, we demonstrated a high first-pass effect during intra-arterial (i.a.) administration of positron emission tomography (PET) labeled (68)Ga-DOTA(0)-d-Phe(1)-Tyr(3)-octreotide (DOTATOC). In this pilot study, we investigated the therapeutic effectiveness of arterial administered DOTATOC, labeled with the therapeutic ß emitters (90)Y and (177)Lu. (90)Y- and/or (177)Lu-DOTATOC were infused into the hepatic artery of 15 patients with liver metastases arising from GEP-NETs. Response was assessed using DOTATOC-PET, multiphase contrast enhanced computed tomography, magnetic resonance imaging, and the serum tumor marker chromogranin A. Pharmacokinetic data of the arterial approach were assessed using (111)In-DOTATOC scans. With the treatment regime of this pilot study, complete remission was achieved in one (7%) patient and partial remission was observed in eight (53%) patients, six patients were classified as stable (40%; response evaluation criteria in solid tumors criteria). The concomitant decrease of elevated serum tumor marker confirmed the radiologic response. Median time to progression was not reached within a mean follow-up period of 20 months. Receptor saturation and redistribution effects were identified as limiting factors for i.a. DOTATOC therapy. The high rate of objective radiologic response in NET patients treated with arterial infusion of (90)Y-/(177)Lu-DOTATOC compares favorably with systemic chemotherapy and intravenous radiopeptide therapy. While i.a. DOTATOC therapy is only applicable to patients with tumors of limited anatomic distribution, the results of this pilot study are a promising development in the treatment of GEP-NET and warrants further investigation of this novel approach.

Insertion of totally implantable venous access devices: an expertise-based, randomized, controlled trial (NCT00600444).

OBJECTIVE: Comparison of two different insertion techniques for implantation of totally implantable access ports (TIAP). BACKGROUND: TIAP are introduced through different open and closed cannulation strategies and by various medical experts. The aim of this expertise-based randomized trial was to compare venous cutdown approach with puncture of subclavian vein. METHODS: One hundred and ten patients scheduled for primary implantation of a TIAP were randomly assigned to either open insertion technique performed by surgeons or puncture of the subclavian vein under fluoroscopic guidance by radiologists at an outpatient single university center. The primary endpoint was the primary success rate of the cannulation strategy. A logistic regression model was used for analysis adjusting for age, Karnofsky index, body mass index and surgeons', and the radiologists' experience. RESULTS: Percutaneus cannulation was not superior to surgical venous cutdown in the intention-to-treat analysis (odds ratio, 0.37; 95% CI, 0.07; 2.15) and the as-treated analysis (odds ratio, 0.16; 95% CI, 0; 1.28). The procedure was shorter with surgery (median, 21 minutes; 95% CI, 14; 30) than with radiology (median, 45 minutes; 95% CI, 43; 50) (P < 0.001), and the dose of radiation was lower with surgery (median, 37 cGy/cm(2); 95% CI, 26; 49) than with radiology (200 cGy/cm(2); 95% CI, 200; 300) (P < 0.001). CONCLUSION: Central venous cannulation for insertion of TIAPs can be performed safely and effectively with both approaches. The open direct surgical access requires further strategies for successful placement of a TIAP, and percutaneous Seldinger technique requires more time and a higher dose of radiation and is associated with risk of pneumothorax.

Lung sequestration: a rare cause for pulmonary symptoms in adulthood.

BACKGROUND: Lung sequestration is a rare congenital pulmonary disorder and is usually diagnosed in children with recurrent pulmonary infections. Lung sequestrations are not commonly found to be a cause of respiratory symptoms in adults. OBJECTIVES: It was the aim of this study to show that pulmonary sequestration is rare in advanced age and can be accompanied by severe pulmonary symptoms. METHODS: We conducted a case series analysis of patient characteristics, symptoms, diagnosis and treatment of 11 adults with a lung sequestration at the Thoraxklinik Heidelberg between 2001 and 2009. RESULTS: From 2001 to 2009, intralobar lung sequestration was diagnosed and treated in 11 adults aged 19 to 58 years with an average age of 39.9 ± 11.3 years and a male:female distribution of 5:6. In 3 patients (27.3%), the predominant symptom was hemoptysis. Recurrent pulmonary infections occurred in 1 patient (9.1%); pneumonia and lung abscess were detected in 2 patients (18.2%). In 3 cases (27.3%), dry cough was the predominant symptom, and in only 2 cases (18.2%), lung sequestration was asymptomatic. Eight patients (72.7%) were diagnosed by imaging techniques prior to surgery. In 3 cases (27.3%), diagnosis was made intraoperatively and by pathological examination. Surgical intervention included 7 lobectomies (63.6%), 3 wedge resections (27.3%) and 1 (9.1%) segmentectomy. CONCLUSION: Lung sequestration in adults is rare, but it can cause severe pulmonary symptoms. In cases of recurrent pulmonary infections of identical localization or recurrent hemoptysis, lung sequestration should be considered in order for the diagnosis to be made rapidly. Surgical resection is the treatment of choice.

Paclitaxel-induced arterial wall toxicity and inflammation: tissue uptake in various dose densities in a minipig model.

PURPOSE: Paclitaxel is an antiproliferative agent in drug-eluting stents with largely unknown tissue interaction. Toxicity might result from overdosage and/or accumulation. Part 1 of this two-step study investigated how paclitaxel uptake depends on dose density, coronary drug transfer kinetics, and elution efficacy. MATERIALS AND METHODS: With cobalt chromium stents and Polyzene-F nanoscale coating, low, intermediate, and high paclitaxel dose densities (25 microg, 50 microg, and 150 microg per stent) were investigated in porcine right coronary arteries (RCAs). Coronary and myocardial tissue concentration measurements and determination of on-stent paclitaxel and plasma concentrations were performed at 2, 8, 24, and 72 hours. RESULTS: For all stents, uptake was similar at all time intervals (paclitaxel RCA concentration range, 1,610-33,300 ng). Low- and intermediate-dose stents showed similar RCA concentrations, but those for high-dose stents were three times greater. Residual on-stent paclitaxel concentration was not time-dependent, at 33.3% on low-, 30.6% on intermediate-, and 17.4% on high-dose stents. Paclitaxel was measurable in only the plasma immediately after stent placement, with a linear dose relationship and a timely regression: measurements in high-dose stents were 0.0454-0.656 ng/mL at 1 minute and 0.0329-0.0879 ng/mL at 5 minutes. Untreated control samples of the left coronary artery showed a linear dose-dependent concentration (12.6 ng/g, 21.2 ng/g, and 85.2 ng/g). CONCLUSIONS: Overall coronary paclitaxel uptake is fairly independent from the baseline overall dose density and, hence, depends on immediate binding mechanisms of the arterial wall. This is supported by the fact that, regardless of the applied dose density, the kinetics of paclitaxel uptake did not follow an exposure time pattern.

Intraindividual comparison of selective arterial versus venous 68Ga-DOTATOC PET/CT in patients with gastroenteropancreatic neuroendocrine tumors.

PURPOSE: Therapy with the somatostatin analogue DOTA-(0)-Phe(1)-Tyr(3)-octreotide (DOTATOC) labeled with a beta-(DOTA-Phe-Tyr-Octreotide) emitter such as 90Y or 177Lu is accepted for the palliative treatment of unresectable neuroendocrine cancer. However, the optimal route of administration has not been determined. Using positron-emission tomography (PET)-labeled 68Ga-DOTATOC, we compared selective tumoral uptake on PET/computed tomography (CT) after arterial or venous administration of the agent in patients with gastroenteropancreatic neuroendocrine tumor. EXPERIMENTAL DESIGN: Fifteen patients with neuroendocrine cancer were examined with 68Ga-DOTATOC PET/CT after intravenous (i.v.) and intraarterial (i.a.) administration within 4 weeks of each other and without any intervening therapy. Eleven patients had multifocal metastases, six were considered to have unresectable primary tumor. The intraarterial catheter was placed in the vessel supplying the main tumor burden. The standard uptake value (SUV) was used to compare intratumoral concentrations of 68Ga-DOTATOC. RESULTS: Compared with i.v. infusion, the i.a. infusion resulted in an increased SUV in 117 of 122 (96%) liver metastases. The average increase in SUV was 3.75-fold higher with i.a. administration. The increase in uptake for the primary tumors was dependent on the selectivity of the catheter placement, resulting in variable increases in SUV after i.a. injection (1.44- to 7.8-fold higher). CONCLUSIONS: This study showed that uptake of DOTATOC is commonly several fold higher after selective i.a. administration in comparison with i.v. injection in both the primary tumor as well as in liver metastases of neuroendocrine cancer. Therefore, intraarterial DOTATOC is a promising drug for regionally intensified radiopeptide therapy.

Arterial distribution characteristics of Embozene particles and comparison with other spherical embolic agents in the porcine acute embolization model.

PURPOSE: To determine the arterial distribution pattern of the embolic agent Embozene within the porcine kidney and compare it with those of other spherical embolic agents. MATERIALS AND METHODS: Embozene, Embosphere, Bead Block, and Contour SE in size classes of 100-300 microm, 500-700 microm, and 700-900 microm and Embozene and Embosphere in the size class of 40-120 microm were used for total arterial occlusion in minipig kidneys. Organs were evaluated microscopically regarding vascular distribution of the different embolic agents and particle sizes. RESULTS: The following variations of arterial distribution were identified. In the 40-120-microm size class, Embosphere particles penetrated significantly deeper compared with Embozene (P = .04). In the 100-300-microm size class, Bead Block showed a significantly deeper distribution as microscopy identified particles in arteries much smaller than their nominal size. In the 500-700-microm size class, Embosphere and Contour SE showed a deeper distribution. The most uniform arterial distribution was observed in the 700-900 microm size class,. However, few Embosphere and Contour SE particles were found in arcuate arteries, also indicating a distal distribution. CONCLUSIONS: Throughout the four most-used size classes, from very small (40-120 microm) to large (700-900 microm), the distribution characteristics of the four tested materials vary substantially. Particularly, small Embosphere particles and small Bead Block particles showed a more distal distribution, as did medium-sized Embosphere and Contour SE particles. In the largest investigated size class, the distribution was more uniform. In general, the Embozene particles are very uniform in size, and they seem to reach vessels closely corresponding to their nominal size.

Paclitaxel-induced arterial wall toxicity and inflammation: part 2--long-term tissue response in a minipig model.

PURPOSE: In part 1 of the present study, the authors demonstrated that coronary paclitaxel uptake from drug eluting stents (DESs) was not dependent on exposure time and dose. In this second part, the effect of the different paclitaxel dose densities on long-term biologic behavior was evaluated. MATERIALS AND METHODS: In 40 minipigs, (with 4- and 12-week follow-up), identical stents with the same three paclitaxel dose densities as in part 1 were implanted in the right coronary artery. Minipigs implanted with Polyzene-F nanocoated stents served as the control group. Quantitative angiography measuring average luminal diameter (from three in-stent reference points), minimal luminal diameter (from the point of maximum in-stent stenosis), average late loss, maximum late loss, and binary stenosis rate was performed, as was microscopy to determine neointimal thickening, injury score, and inflammation. RESULTS: All three DESs were associated with a high average late loss, binary stenosis rate, and neointimal thickening, without significant differences. Drug-free stents had significantly less late in-stent stenosis: there was an average late loss of 0.3 mm +/- 0.3 in drug-free stents versus 0.8 mm +/- 0.2 in intermediate-dose stents and 1.5 mm +/- 0.6 in high-dose stents (P = .04). DES-associated inflammation was high in all DESs and six times higher as in the drug-free stents (Kornowski scores of 0.2 +/- 0.1 in drug-free stents, 1.3 +/- 0.9 in low-dose stents, 1.7 +/- 0.8 in intermediate-dose stents, and 1.3 +/- 1.0 in high-dose stents; P = .04). It worsened with time in all DESs, as did late in-stent stenosis. CONCLUSIONS: The extensive and long-term retention of paclitaxel even in a low-dose formulation, at least according to the present labeling of DESs, might be associated with negative long-term results with regard to inflammation and late in-stent stenosis.

Comparison of a microporous thermoplastic polyurethane-covered stent with a self-expanding bare nitinol stent in a porcine iliac artery model.

PURPOSE: To compare outcomes with a thermoplastic polyurethane (TPU)-covered self-expanding nitinol stent-graft (TPU graft) with those of a bare self-expanding nitinol stent in a porcine model. MATERIALS AND METHODS: Fourteen TPU grafts and 14 commercially available bare nitinol stents were implanted, one each, in the iliac arteries of 14 minipigs. Follow-up was performed at 1 week (six animals), 4 weeks (four animals), and 12 weeks (four animals). The primary study endpoint was in-stent stenosis assessed with quantitative angiography and microscopy. Secondary endpoints were injury, inflammation, and endothelialization. RESULTS: After 1 week, the maximum percentage luminal loss was significantly greater in TPU grafts (average, 16.2%; range, 0.0%-35.8%) than in bare nitinol stents (8.2%; 0.0%-17.3%) (P = .04). Three of the four TPU grafts were occluded after 4 weeks, and all four TPU grafts were occluded after 12 weeks. Binary stenosis was seen in three of four bare nitinol stents after both 4 and 12 weeks. At 4-week follow-up, the average percentage luminal loss was significantly greater in TPU grafts (85.2%; 40.8%-100%) than in bare nitinol stents (49.5%; 37.9%-62.4%) (P = .003). The difference in neointimal height and percentage average stenosis between TPU grafts (1,028.7 microm and 68.4%) and bare nitinol stents (1,033.6 microm [918.0-1,118.40 microm] and 68.1% [60.44%-71.99%]) was not statistically significant. After 12 weeks, the average percentage luminal loss was 100% in TPU grafts due to occlusion of all stent-grafts and 24.9% (8.0%-63.9%) in bare nitinol stents (P = .011). CONCLUSIONS: TPU grafts failed to provide improved patency compared with the bare nitinol stents because of excessive neointimal growth and subsequent occlusion. In addition, the bare nitinol stents showed considerable in-stent stenosis at angiography and microscopy.