Acute type A aortic dissection intimal tears by 64-slice computed tomography: a role for endovascular stent-grafting?

AIM: The goal of this study was to identify physical characteristics of primary intimal tears in patients arriving to the hospital alive with acute type A aortic dissection using 64-multislice computerized tomography (MSCT) in order to determine anatomic feasibility of endovascular stent-grafting (ESG) for future treatment. METHODS: Radiology database was screened for acute type A aortic dissection since the time of acquisition of the 64-slice CT scanner and cross-referenced with surgical database. Seventeen patients met inclusion criteria. Images were reviewed for number, location, and size of intimal tears and aortic dimensions. Potential obstacles for ESG were determined. RESULTS: Ascending aorta (29%) and sinotubular junction (29%) were the most frequent regions where intimal tears originated. Location of intimal tears in nearly 75% of patients was inappropriate for ESG, and 94% of patients did not have sufficient proximal or distal landing zone required for secure fixation. Only 71% of patients underwent surgical aortic dissection repair after imaging and 86% of entry tears detected on MSCT were confirmed on intraoperative documentation. Only one patient would have met all technical criteria for ESG using currently available devices. CONCLUSION: Location of intimal tear, aortic valve insufficiency, aortic diameter>38 mm are major factors limiting use of ESG for acute type A dissection. Available stents used to treat type B aortic dissection do not address anatomic constraints present in type A aortic dissection in the majority of cases, such that development of new devices would be required.

Hemodynamic determinants of aortic dissection propagation by 2D computational modeling: implications for endovascular stent-grafting.

AIM: Aortic dissection is a life-threatening aortic catastrophe where layers of the aortic wall are separated allowing blood flow within the layers. Propagation of aortic dissection is strongly linked to the rate of rise of pressure (dp/dt) experienced by the aortic wall but the hemodynamics is poorly understood. The purpose of this study was to perform computational fluid dynamics (CFD) simulations to determine the relationship between dissection propagation in the distal longitudinal direction (the tearing force) and dp/dt. METHODS: Five computational models of aortic dissection in a 2D pipe were constructed. Initiation of dissection and propagation were represented in 4 single entry tear models, 3 of which investigated the role of length of dissection and antegrade propagation, 1 of which investigated retrograde propagation. The 5th model included a distal re-entry tear. Impact of pressure field distribution on tearing force was determined. RESULTS: Tearing force in the longitudinal direction for dissections with a single entry tear was approximately proportional to dp/dt and L2 where L is the length of dissection. Tearing force was much lower under steady flow than pulsatile flow conditions. Introduction of a second tear distally along the dissection away from the primary entry tear significantly reduced tearing force. CONCLUSION: The hemodynamic mechanism for dissection propagation demonstrated in these models support the use of ß-blockers in medical management. Endovascular stent-graft treatment of dissection should ideally cover both entry and re-entry tears to reduce risk of retrograde propagation of aortic dissection.

Endovascular repair of thoracoabdominal aneurysms: design options, device construct, patient selection and complications.

The aim of this paper was to describe the current status of endovascular thoracoabdominal aortic aneurysm (TAAA) repair. This is a comparative review of current device designs and implantation techniques. A literature review of all reported results of endovascular TAAA repair has also been carried out, together with a comparison of clinical outcomes achieved with endovascular TAAA repair and those achieved in current series of standard open TAAA repair. Endovascular TAAA repair has been performed with both unibody and modular devices, but modular devices currently predominate. In modular devices the aortic component provides access to the target visceral artery either through a fenestration or a cuff. Cuffs increase device profile and the length of aorta that is covered, but easily accommodate variations in deployment position and provide a good seal zone. Fenestrations do not affect device profile or add length to the device, but deployment position tolerates little deviation and the seal zone is tenuous. A covered stent is used to bridge the gap between the fenestration or cuff in the aortic component and the target visceral artery. Balloon-expandable covered stent branch extensions are delivered from the femoral approach when fenestrations are used. Self-expanding covered stents are delivered from either the brachial or femoral approach when cuffs are used, depending on the orientation of the cuff. Some groups reinforce the self-expanding covered stent with an uncovered self-expanding stent to enhance flexibility and stability. The majority of endovascular TAAA repairs have been performed in three centers, accounting for 84% of all reported cases. The treated TAAAs were Type 1 31.8%, Type 2 14.2%, Type 3 14.2% and Type 4 37.5%. Perioperative mortality is 6.9% (N. = 20), late mortality 13.6% (N. = 38), spinal cord ischemia (SCI) 14.9% (N. = 29) permanent in 6.7% (N. = 6), transient in 10.0% (N. = 9). Deterioration of renal function was reported in 9.8% (N. = 8), and required initiation of dialysis in 5.1% (N. = 5). Reintervention was required in 18 patients (20.0%) early in 8.9% and late in 11.1%. Branch occlusion developed in 3.5% (N. = 9) and stenosis in 0.85% (N. = 2). Current single-center series of open surgical TAAA repair report mortality rates of 5-16%, spinal cord ischemia rates of 3.8-15.5% and new onset dialysis between 2-16.2%. Population-based series of open surgical TAAA repair report mortality rates between 19.2-26.9%, spinal cord ischemia rates between 7.3-16.0% and new onset dialysis rates of 14.2-18.2%. Final status of SCI neurologic deficit, reintervention rates and branch occlusion or stenosis rates for open TAAA repair are inconsistently available, if at all. In conclusion, endovascular TAAA repair is an evolving technique that is developing increasing consistency in device design and implantation technique. It is effective in eliminating aneurysm flow and in preserving visceral branch perfusion. These early outcomes are better than the results achieved with open TAAA repair in population-based studies and are at least equal to the results of open TAAA repair reported from centers of focused expertise. These results support expanding the indications for endovascular TAAA repair to include standard risk patients.

Double-barrel technique for preservation of aortic arch branches during thoracic endovascular aortic repair.

Thoracic endovascular aortic repair (TEVAR) may involve either planned or inadvertent coverage of aortic branch vessels when stent grafts are implanted into the aortic arch. Vital branch vessels may be preserved by surgical debranching techniques or by placement of additional stents to maintain vessel patency. We report our experience with a double-barrel stent technique used to maintain aortic arch branch vessel patency during TEVAR. Seven patients underwent TEVAR using the double-barrel technique, with placement of branch stents into the innominate (n = 3), left common carotid (n = 3), and left subclavian (n = 1) arteries alongside an aortic stent graft. Gore TAG endografts were used in all cases, and either self-expanding stents (n = 6) or balloon-expandable (n = 1) stents were utilized to maintain patency of the arch branch vessels. In three cases the double-barrel stent technique was used to restore patency of an inadvertently covered left common carotid artery. Four planned cases involved endograft deployment proximally into the ascending aorta with placement of an innominate artery stent (n = 3) and coverage of the left subclavian artery with placement of a subclavian artery stent (n = 1). TEVAR using a double-barrel stent was technically successful with maintenance of branch vessel patency and absence of type I endoleak in all seven cases. One case of zone 0 endograft placement with an innominate stent was complicated by a left hemispheric stroke that was attributed to a technical problem with the carotid-carotid bypass. On follow-up of 2-18 months, all double-barrel branch stents and aortic endografts remained patent without endoleak, migration, or loss of device integrity. The double-barrel stent technique maintains aortic branch patency and provides additional stent-graft fixation length during TEVAR to treat aneurysms involving the aortic arch. Moreover, the technique uses commercially available devices and permits complete aortic arch coverage (zone 0) without a sternotomy. Although initial outcomes are encouraging, long-term durability remains unknown.

Endovascular treatment of thoracoabdominal aortic aneurysms.

Open surgical repair of thoracoabdominal aortic aneurysms (TAAA) bridges the aneurysm with a large, conventional, unstented graft and restores flow to the visceral arteries through short grafts or direct sutured connections between the visceral arterial orifices and the primary conduit. The combination of retrograde visceral bypass and endovascular aneurysm exclusion substitutes an endovascular stent-graft for a standard graft, stented overlaps for sutured anastomoses, and transluminal insertion for direct aortic exposure. Compared to open surgery, the combination treatment requires less dissection, and causes less hemodynamic instability, and lower complication rates, particularly paraplegia. The multi-branched stent-graft substitutes endovascular visceral bypass through branches of the stent-graft for surgical visceral bypass through branches of a conventional extraluminal graft, which has the potential to further reduce surgical dissection, hemodynamic instability, and complication rates. We favor a modular approach in which short, axially oriented cuffs are extended into the visceral arteries, using self-expanding covered stents. In the past year, we have used this approach to implant multi-branched thoracoabdominal stent-graft in 16 patients. In our opinion, this approach will eventually assume a prominent role in the management of TAAA.

The choice of stent-graft for endovascular repair of abdominal aortic aneurysm.

The aim of this study was to explain variations in the results of endovascular aneurysm repair as a consequences of device design. Low profile, trackable systems, such as Zenith and Excluder, rarely fail to traverse the iliac arteries, even in the presence of iliac tortuosity or stenosis. In most patients, optimal sizing is only possible with systems, such as Zenith, Talent, and Quantum lp, that have a wide range of diameters. Short, angulated necks call for a high degree of flexibility and secure, barb-enhanced proximal fixation, which are features of Excluder, Zenith and Ancure. The main risk factors for rupture are migration, type III endoleak, and aneurysm dilatation. Migration rates are high for devices, such as AneuRx, that have neither barbs nor suprarenal stents. Aneurysm shrinkage occurs at high rates with non-porous stent-grafts, such as Zenith, Talent, and Ancure, but at far lower rates with porous stent-grafts, such as Excluder and AneuRx. Type III endoleak, due to fabric failure or component separation, was a common failure mode for the Vanguard device, but is rare with newer devices. Suture breakage, barb separation and stent breakage occur frequently, yet clinical consequences, such as endoleak or rupture, are rare. Graft thrombosis is also unusual when the prostheses is fully-stented. In conclusion, modern devices are more versatile, more effective, and more durable than their first generation counterparts.

Endoleak after endovascular repair of abdominal aortic aneurysm.

PURPOSE: We sought to assess the role of endovascular techniques in the management of perigraft flow (endoleak) after endovascular repair of an abdominal aortic aneurysm. METHOD: We performed endovascular repair of abdominal aortic aneurysm in 114 patients, using a variety of Gianturco Z-stent-based prostheses. Results were evaluated with contrast-enhanced computed tomography (CT) at 3 days, 3 months, 6 months, 12 months, and every year after the operation. An endoleak that occurred 3 days after operation led to repeat CT scanning at 2 weeks, followed by angiography and attempted endovascular treatment. RESULTS: Endoleak was seen on the first postoperative CT scan in 21 (18%) patients and was still present at 2 weeks in 14 (12%). On the basis of angiographic localization of the inflow, the endoleak was pure type I in 3 cases, pure type II in 9, and mixed-pattern in 2. Of the 5 type I endoleaks, 3 were proximal and 2 were distal. All five resolved after endovascular implantation of additional stent-grafts, stents, and embolization coils. Although inferior mesenteric artery embolization was successful in 6 of 7 cases and lumbar embolization was successful in 4 of 7, only 1 of 11 primary type II endoleaks was shown to be resolved on CT scanning. There were no type III or type IV endoleaks (through the stent-graft). Endoleak was associated with aneurysm dilation two cases. In both cases, the aneurysm diameter stabilized after coil embolization of the inferior mesenteric artery. There were two secondary (delayed) endoleaks; one type I and one type II. The secondary type I endoleak and the associated aneurysm rupture were treated by use of an additional stent-graft. The secondary type II endoleak was not treated. CONCLUSIONS: Type I endoleaks represent a persistent risk of aneurysm rupture and should be treated promptly by endovascular means. Type II leaks are less dangerous and more difficult to treat, but coil embolization of feeding arteries may be warranted when leakage is associated with aneurysm enlargement.

An endovascular system for thoracoabdominal aortic aneurysm repair.

PURPOSE: To describe a stent-graft system for endovascular repair of thoracoabdominal aortic aneurysm (TAAA) that preserves side branch perfusion. TECHNIQUE: The modular endograft system includes 3 components. The primary stent-graft is custom-made from conventional graft fabric and Gianturco Z-stents. Covered nitinol Smart Stents are used for the visceral and renal extensions, and the distal extension is made from a modified Zenith system. With the supine patient under general anesthesia, the components are delivered sequentially through surgically exposed femoral and right brachial arteries in an operation that requires prolonged periods of magnified high-resolution imaging. This system was first used in a 76-year-old man with a contained rupture of a supraceliac ulcer and a large abdominal aortic aneurysm ending proximally at the celiac artery. The endograft was implanted successfully, but the patient developed paraplegia on day 2; imaging documented an excluded aneurysm and excellent flow through the endograft and all prosthetic branches. DISCUSSION: Endovascular repair of TAAA appears to be feasible. If there are no serious, specific, unavoidable complications, the potential advantages are enormous.

An update of the Zenith endovascular graft for abdominal aortic aneurysms: initial implantation and mid-term follow-up data.

PURPOSE: To evaluate the initial and mid-term results of the Zenith endovascular grafting system for infrarenal abdominal aortic aneurysms. METHODS: Prospective databases at seven centers were used to assess a cohort of patients that underwent treatment for aortic, aortoiliac, or iliac aneurysms since 1995. Data were analyzed to yield descriptive characteristics that pertained to the patients, the aortic morphologic features, the graft configuration, and the complications. Follow-up imaging data were used to determine size changes of the aneurysm sac, endoleak rates, and further complications. Finally survival data were expressed with a Kaplan-Meier analysis. RESULTS: A total of 528 patients were treated with the Zenith endograft. Most of the patients (66%) were considered to be at a high physiologic risk for open repair. Successful graft implantation was accomplished in all but four patients. An overall endoleak rate of 15% was noted, of which 4% was treated urgently because they were thought to represent attachment site faults. The mean follow-up period was 18 months. A total of eight endograft migrations were detected after 2 years of follow-up with an early version of the system. There were three late conversions; two ruptures occurred during the follow-up period. CONCLUSION: This early and mid-term data support the use of the Zenith endovascular graft for the treatment of aortic and aortoiliac aneurysms in properly selected patients. The risks of significant complications or aneurysm rupture are low.

Endovascular repair of a presumed aortoenteric fistula: late failure due to recurrent infection.

PURPOSE: To describe a case of presumed aortoduodenal fistula that was treated by endovascular implantation of a stent-graft. METHODS AND RESULTS: A 76-year-old man was transferred from another hospital where he had been treated for upper gastrointestinal hemorrhage over a 2-month period. Ten years previously, he had undergone aortobifemoral bypass, the right limb of which recently thrombosed. At the time of transfer, computed tomographic scanning showed a large false aneurysm between the aorta and the duodenum. Endoscopy disclosed mucosal erosions in the fourth portion of the duodenum. Following implantation of 2 overlapping stent-grafts, the bleeding ceased and the false aneurysm disappeared. At no time did the patient have a fever. The patient initially did well, but 8 months after treatment, he presented with fever and chills. Recurrent infection had caused erosion of the aorta so that a large portion of the stent-graft was visible from the duodenum. The infected graft and stent-grafts were removed in a two-part operation, from which the patient recovered satisfactorily. CONCLUSIONS: Endovascular stent-grafts may have a role to play in the management of aortoduodenal fistula, if only as a temporary measure to control bleeding.

Endovascular aneurysm repair in high-risk patients.

PURPOSE: The purpose of this study was to evaluate the role of endovascular aneurysm repair in high-risk patients. METHODS: The elective endovascular repair of infrarenal aortic aneurysm was performed in 116 high-risk patients with either custom-made or commercial stent grafts. The routine follow-up examination included contrast-enhanced computed tomography (CT) before discharge, at 3, 6, and 12 months, and annually thereafter. Patients with endoleak on the initial CT underwent re-evaluation at 2 weeks. Those patients with positive CT results at 2 weeks underwent endovascular treatment. RESULTS: Endovascular repair was considered feasible in 67% of the patients. The mean age was 75 years, and the mean aneurysm diameter was 6.3 cm. The American Society of Anesthesiologists grade was II in 3.4%, III in 65.5%, IV in 30.1%, and V in 0.9%. There were no conversions to open repair. Custom-made aortomonoiliac stent grafts were implanted in 77.6% of the cases, custom-made aortoaotic stent grafts in 11.2%, and commercial bifurcated stent grafts in 11.2%. The 30-day rates of mortality, major morbidity, and minor morbidity were 3.4%, 20.7%, and 12%, respectively, in the first 58 patients and 0%, 3.4%, and 3.4%, respectively, in the last 58. The late complications included five cases of stent graft kinking, two cases of femorofemoral graft occlusion, and three cases of proximal stent migration, one of which led to aneurysm rupture. At 2 weeks after repair, endoleak was present in 10.3% of the cases. All the type I (direct perigraft) endoleaks underwent successful endovascular treatment, whereas only one type II (collateral) endoleak responded to treatment. The technical success rate at 2 weeks was 86.2%, and the clinical success rate was 96.6%. The continuing success rate was 87.9%. Seventeen patients died late, unrelated deaths. CONCLUSION: Endovascular aneurysm repair is safe and effective in patients at high risk, for whom it may be the preferred method of treatment.

Aortomonoiliac endovascular grafting combined with femorofemoral bypass: an acceptable compromise or a preferred solution?

The relative merits of aortomonoiliac and bifurcated stent-graft configurations depend on the patient's arterial anatomy and clinical status. Aortomonoiliac stent-grafts are simple to make, simple to insert, and versatile. They are most useful when the iliac artery anatomy is severely distorted and the patient is old, sick, and inactive. The main problems with this approach are all consequences of femorofemoral bypass. The bifurcated stent-graft is the preferred alternative in healthy patients, because it ensures flow to both common iliac arteries, thereby eliminating the need for femorofemoral bypass. However, bifurcated stent-grafts and their delivery systems are difficult to make and difficult to deploy, especially when the iliac anatomy is distorted or emergency circumstances preclude preoperative sizing. This article addresses the advantages and disadvantages of the aortomonoiliac graft.

Abdominal aortic aneurysm in high-risk patients: short- to intermediate-term results of endovascular repair.

PURPOSE: To assess the safety and efficacy of endovascular repair of abdominal aortic aneurysm in high-risk patients during the short to intermediate term. MATERIALS AND METHODS: Endovascular aneurysm repair was performed in 50 patients considered too high risk for conventional repair. Stent-grafts were inserted through surgically exposed femoral arteries with fluoroscopic guidance. The anesthetic technique was epidural in 36 patients, general in 12, and local in two. Aortouniiliac stent-grafts were inserted in 42 patients and aortoaortic in eight. RESULTS: There were no deaths and no conversions to open surgical repair. The primary success rate (complete aneurysm exclusion according to CT criteria) was 88% (44 of 50). The secondary, clinical, and continuing success rates were all 98% (49 of 50). Surgical time was 196 minutes +/- 67 (mean +/- SD), blood loss was 284 mL +/- 386, and volume of contrast material administered was 153 mL +/- 64. The time from the end of the surgery to resumption of a normal diet was 0.58 days +/- 0.56, to ambulation was 1.22 days +/- 0.77, and to discharge from the hospital was 3.63 days +/- 1.60. Wound problems accounted for the majority of complications. There were no instances of pulmonary failure, renal failure, stent-graft migration, or late leakage. CONCLUSION: Endovascular repair of abdominal aortic aneurysm is feasible in two-thirds of high-risk patients, with a low mortality and high success rate during the short to intermediate term.

Endovascular repair of abdominal aortic aneurysm using a pararenal fenestrated stent-graft.

PURPOSE: To report an unusual case of endovascular abdominal aortic aneurysm (AAA) exclusion in which a fenestrated stent-graft was used to seal a proximal Type I endoleak. METHODS AND RESULTS: An 84-year-old man with a 6.0-cm AAA underwent an aortomonoiliac aneurysm exclusion procedure that was complicated by a proximal endoleak. Because the patient had no right kidney, an additional stent-graft was designed to cover the right renal artery stump while preserving left renal perfusion through a fenestration in the graft material. This approach was successful in obliterating the endoleak around the proximal attachment site, but flow through the lumbar arteries remained. CONCLUSIONS: The use of a fenestrated stent-graft is feasible, but the type of fenestration in this case has limited applicability owing to the rarity of patients with suitable anatomy.

Femoral artery exposure for endovascular aneurysm repair through oblique incisions.

PURPOSE: To offer an alternative technique for accessing the femoral artery prior to endovascular grafting. TECHNIQUE: An oblique incision is made over the medial half of the inguinal ligament and continues to the femoral sheath, which is opened longitudinally. The distal external iliac artery and proximal common femoral artery are isolated. A tiny stab wound is made distal to the primary wound for femoral artery puncture and catheter access. CONCLUSIONS: Using an oblique incision at the level of the inguinal ligament optimizes exposure for endograft insertion and may minimize the frequency of serious wound complications.

Endoleakage after stent-graft treatment of abdominal aneurysm: implications on pressure and imaging--an in vitro study.

BACKGROUND: Endoleakage is a fairly common problem after endovascular repair of abdominal aortic aneurysm and may prevent successful exclusion of the aneurysm. The consequences of endoleakage in terms of pressure in the aneurysmal sac are not exactly known. Moreover, the diagnosis of endoleakage is a problem because visualization of endoleaks can be difficult. METHOD: With an ex vivo model of circulation with an artificial aneurysm managed by means of a tube graft, studies were performed to evaluate precisely known diameters of endoleaks with both imaging techniques (computed tomography and digital subtraction angiography) and pressure measurements of the aneurysmal sac. The experiments were performed without endoleak (controls) and with 1.231-French (0.410 mm), 3-French (1 mm), and 7-French (2.33 mm) endoleaks. Pressure and imaging were evaluated in the absence and presence of a simulated open lumbar artery. The pressure in the prosthesis and in the aneurysmal sac were recorded simultaneously. Digital subtraction angiography with and without a Lucite acrylic plate, computed tomographic angiography, and delayed computed tomographic angiography were performed. For the first experiments, the aneurysmal sac was filled with starch solution. All tests were repeated with fresh thrombus in the aneurysmal sac. RESULTS: Each endoleak was associated with a diastolic pressure in the aneurysmal sac that was identical to diastolic systemic pressure, although the pressure curve was damped. At digital subtraction angiography without a Lucite acrylic plate, the 1.231-French (0.410 mm) endoleak was visualized without an open lumbar artery. When a Lucite acrylic plate was added, the endoleak was not visible until a lumbar artery was opened. In the presence of thrombus within the aneurysmal sac, all endoleaks were not visualized at digital subtraction angiography. At computed tomographic angiography, all endoleaks were not visualized in the absence of a thrombus mass in the aneurysmal sac. In the presence of thrombus within the aneurysmal sac, the 1.231-French (0.410 mm) endoleak became visible after opening of a simulated lumbar artery. At delayed computed tomographic angiography, all endoleaks were visualized without and with thrombus. CONCLUSION: Every endoleak, even a very small one, caused pressure greater than systemic diastolic pressure within the aneurysmal sac. However, small endoleaks were not visualized with digital subtraction angiography and computed tomographic angiography, whereas all endoleaks were visualized with a delayed computed tomographic angiography protocol. We believe that follow-up examinations after stent graft placement for aortic aneurysms should focus on pressure measurements, but until this is clinically feasible, delayed computed tomographic angiography should be performed.

Endovascular repair of abdominal aortic aneurysm: getting out of trouble.

The standard techniques of endovascular aneurysm repair sometimes fail to produce atraumatic stent-graft delivery of hemostatic implantation, and additional maneuvers are required to avoid conversion to open repair. Between June 1996 and May 1997 elective endovascular aneurysm repair was performed in 33 high risk patients, using a Z-stent-based prosthesis. Challenging anatomic features included: short neck (< 15 mm) in four cases, angulated neck (> 60 degrees) in seven, iliac aneurysm in six, and iliac tortuosity (> 80 degrees) in 24. There were no deaths, no renal failure, no pulmonary failure, no graft thrombosis, no migration, and no conversions to open surgery. Deviations from standard technique were required to treat iliac artery dissection, iliac artery stenosis, and leaks resulting from proximal stent malalignment, proximal stent malposition, and distal stent malposition. The necessary adjunctive maneuvers included: additional stent placement, additional stent-graft placement, and balloon dilatation. Mean operating time was 191 +/- 72 min, mean contrast volume was 148 +/- 76 ml, and mean blood loss was 314 +/- 427 ml. Mean time from operation to discharge from the hospital was 3.5 +/- 1.67 days. These short-term results demonstrate that endovascular aneurysm repair is safe and effective in high risk patients, only if adjunctive maneuvers are available to supplement standard technique.

Bifurcated stent-graft for abdominal aortic aneurysm.

The purpose of this study was to examine bifurcated stent-graft implantation for abdominal aortic aneurysm. Fifty-seven patients were treated and followed with serial computed tomography scans for up to 3 years. Patients were allocated to three groups (first 20, second 20, last 17) according to when the repair was performed. Successful treatment is defined as exclusion of the aneurysm from the circulation, based on contrast computed tomography. Success rates in the three groups were 55%, 70% and 100%. Perigraft leak (endoleak) was present on initial assessment in 4/20, 2/20 and 1/17. Two of these aneurysms ruptured early in the postoperative period. Thereafter, leaks were sought and treated aggressively. Kinking and thrombosis occurred in six of the first 20 patients, but did not occur in any of the last 37 patients, in whom the graft limbs were routinely stent-supported throughout. Infrarenal implantation in very a short neck (< 10 mm), or a thrombus-lined neck was associated with proximal stent migration. In conclusion, changes in patient selection and technique have led to a steady rise in the short-term success rate of the stent-grafted implantation.