Serious complications that require surgical interventions after endoluminal stent-graft placement for the treatment of infrarenal aortic aneurysms☆☆☆

Article Outline

• Abstract
• Methods
  • Patients
  • Description of the devices
  • Introduction of the devices
  • Explantation of the stent-graft prosthesis
• Results
• Discussion
  • Neck widening and stent dislocation
  • Stent implantation and infection
  • Complications associated with or despite the explantation
  • Limitations of the study
• Conclusions
• Acknowledgements
• References
• Copyright

Abstract 

Background: Endoluminal stent-graft placement for the treatment of infrarenal abdominal aortic aneurysms has gained widespread acceptance because it is associated with lower peri-interventional morbidity than conventional transabdominal surgery. In this study the long-term morbidity of the procedure was evaluated. Methods and Results: Between September 1994 and December 1998, 150 patients (age, 69.6 ± 8.49 years; 142 men, 8 women) with abdominal aortic aneurysms were treated with the placement of an intravascular nitinol stent-graft (Stentor [55] and Vanguard-System [95]); there were eight tubular grafts and 142 bifurcated grafts. Initial placement of the stent-graft was successful in 144 patients. Mean follow-up was 49 ± 25 months. In 13.3% of stent-graft placements we encountered the following complications: 4 cases of migration or dislocation of the prosthesis (30.5 ± 7.4 months after placement); 2 ruptures of the aorta (26.7 and 15.0 months after placement); 3 recurrent thromboses of the stent-graft (25.5 ± 5.3 months after placement); 3 endoleaks (27.5 ± 15.7 months after placement); and 5 infections of the prosthesis (26.6 ± 16.5 months after placement). There was no correlation between the complications and the type of stent used. All of these patients were treated by surgical replacement of the prosthesis with a Dacron graft. Conclusions: (1) The results suggest that most complications are due to a continuation of the disease process leading to loosening of the prosthesis. (2) Explantation of the prosthesis and surgical repair is feasible but bears additional risks. (3) Because the onset of reperfusion of the excluded aneurysm cannot be predicted, all patients with infrarenal aortic stent-grafts require frequent computer tomographic follow-up. (4) Finally, the results call for further improvements in the design of the stent-graft. (J Vasc Surg 2001;34:198-203.)

Endoluminal stent-graft placement has become an alternative therapy for the treatment of infrarenal abdominal aortic aneurysms (AAAs) in a number of medical centers. In these centers only those patients in whom stent-graft implantation is not possible are referred for conventional surgery. The change in therapeutic guidelines has occurred for the following three reasons: (1) endoluminal treatment is less invasive than conventional open surgery¹, ², ³, ⁴; (2) stent-grafts can be deployed with patients under local anesthesia, thereby offering a less risky treatment option in high-risk patients⁵; and (3) stent implantation has thus far been shown to yield excellent immediate results after implantation.⁶, ⁷, ⁸, ⁹ However, results on long-term follow-up are not yet available.

We started to implant stents 5 years ago, and we have recently encountered an increase of serious complications that require surgical explantation and replacement of the stent-graft with a polyester fiber (Dacron) graft in a substantial fraction of this patient population. In this study we summarize our experience with explantation of the stent-grafts.

Back to Article Outline

Methods 

Patients 

Between September 1994 and December 1998, 150 patients with AAA (mean age, 69.6 ± 8.49 years; 142 men, 8 women) were treated with endoluminal placement of polyester-covered nitinol stents. We implanted Stentor stent-grafts (Mialhe Stentor, Min Tec, Freeport, Bahamas) in 55 patients (5 tube and 50 bifurcated). The remaining 95 patients received implantations of Vanguard devices (Boston Scientific, Oakland, NJ) (3 tube and 92 bifurcated). The demographic data of the 150 patients receiving stent implantations are presented in Table I.

Table I. Characteristics of 150 patients treated with endoluminal aortic stents (September 1994 to December 1998).

AAA, Abdominal aortic aneurysm; ASA, American Society of Anesthesia physical status; type A, involvement of the aorta only; type B/C, involvement of the aortic bifurcation or the bifurcation and the iliac arteries, respectively.

Fourteen (9.3%) of the 150 patients were classified as class 4 according to the American Society of Anesthesia physical status. All patients included in the study had asymptomatic aneurysms. At least 1 week before the procedure, a spiral computer tomography (CT) and a multiplane angiography of the abdominal aorta were performed in all patients considered for stent-graft implantation. Patients with inflammatory abdominal aneurysms, as revealed with CT, clinical investigations, and inflammatory parameters, did not receive stent implantations. Aneurysms were classified as type A (n = 8, involvement of the abdominal aorta only) or type B or C (n = 142, involvement of the aortic bifurcation or the bifurcation and iliac arteries, respectively). Patients with excentric or saccular aneurysms were excluded from the study. Patient inclusion criteria for endoluminal treatment of AAA have been described in detail before⁶, ⁷, ⁸ and were as follows: (1) minimal diameter of aortic aneurysm greater than 50 mm and (2) proximal aortic neck not less than 10 mm in length and not more than 27 mm in diameter.

All stent-graft implantations were performed on an elective basis. The procedures were performed with 121 patients (80.7%) under general anesthesia. Twenty-nine interventions (19.3%) were performed with patients under local anesthesia (10-20 mL of 1% lidocaine injected at the groin cut-down) and sedation (remifentanil 0.05 μg/kg per minute and midazolam 0.5-1.0 mg, given intravenously). All stents were implanted by the same interventional radiologist assisted by vascular surgeons who performed the arterial access. All stent-graft implantations were done in the angiographic suite with the patient prepared as for a conventional, transabdominal procedure. Before the intervention all patients received a single dose of broad spectrum antibiotics (cephalosporin; 2 g, given intravenously) and an additional dose 8 hours thereafter. Follow-up with CT was scheduled at 1, 6, and 12 months after implantation and each year thereafter.

Description of the devices 

The Min Tec Stentor graft and the Vanguard device are flexible, self-expanding endoprostheses that are made from nitinol wire stent frames covered with a polyester fabric. The Min Tec Stentor graft has been superseded, has been redesigned, and has evolved into the Vanguard device. Additionally, the introducer system for delivery of the stent-grafts has been modified. Before the Vanguard prosthesis could undergo further technical modifications, its realization was stopped at the end of 1999 after leaks of the prosthesis with subsequent rupture of the aneurysmal sac had been reported. The stent-grafts, as used in the current study, were available in straight or bifurcated designs and selected according to the anatomy of the aneurysms. The devices were commercially available in standard sizes and were selected individually for each patient.¹⁰ Both devices have been extensively described before.⁶, ⁷, ⁸

Introduction of the devices 

The introduction and deployment of the devices have been described in detail elsewhere.⁶, ⁷, ⁸ Briefly, after induction of anesthesia, heparin (100 IU/kg) was given. The femoral arteries were exposed by a standard groin incision on the left or right side, and the aortic section, plus the ipsilateral iliac limb of the stent prosthesis, was implanted under fluoroscopic guidance with the introduction catheter. If a bifurcated graft was needed, contralateral percutaneous femoral puncture was performed, and the introducer containing the second limb of the stent-graft was inserted. After exclusion of the AAA, the arteriotomy and the groin incision were closed.

If the procedure was performed with patients under general anesthesia, the patients were intubated and transferred to the intensive care unit. After routine extubation, the patients were observed while in the intensive care unit for 12 hours before they were transferred to the ward. If the procedure was performed with patients under local anesthesia, patients were transferred immediately after the intervention to the ward.

All patients received heparin intravenously for 48 hours after the intervention (partial thromboplastin time, 40-60 seconds) followed by long-term treatment with aspirin 100 mg/d.

Explantation of the stent-graft prosthesis 

Explantation was indicated for reasons stated below, if additional stent-graft placement was unsuccessful, or if not possible. The following was the explanation procedure: After transverse laparotomy and exposure of the aneurysm, the left renal vein was mobilized, and the aorta was clamped as proximal as possible but just distally to the renal arteries. The aneurysm was opened longitudinally, and the stents were released from the surrounding vascular tissue without a reopening of the aortic cross-clamp. A dislocated prosthesis could be removed from the aneurysmal sac manually without tractive force. In all other cases, the body of the stent-graft was grasped with a vascular clamp, and the anchoring part of the stent was released from its position. After the stent prosthesis was removed, the iliac arteries were blocked with balloon catheters, and a Dacronbifurcated graft was placed in conventional manner.¹¹

In all cases a distinct ring formed by the neck of the aneurysm remained inferior to the renal arteries, which could be used for the proximal suture line of the Dacron prosthesis. Pledgets of polytef (Teflon) felt were routinely incorporated into the proximal suture line. The Dacron prosthesis was routinely soaked with an antibiotic solution (rifampin 600 mg) before its implantation in patients with infected stent-grafts. If the new graft could not be covered with the remaining aortic wall tissue, the omentum was used to achieve total coverage. The placement of omentum was especially important in those cases where infection of the prosthesis required excision of large parts of the surrounding tissue.

Back to Article Outline

Results 

Successful deployment of the stent-grafts was achieved in 144 patients (96%). In three patients (2%) stent implantation was unsuccessful because of insertion failure of the introducer delivery system (Stentor [1]; Vanguard [2]). In another three patients conversion to open procedure had to be performed immediately because of rupture of the iliac artery by the introduction system in one patient (Stentor) and proximal displacement of the endovascular prosthesis with subsequent refilling of the aneurysm in two patients (Vanguard). Sixteen patients were lost for further follow-up studies. The remaining 128 patients were followed up between 11 and 60 months (mean, 49 ± 25 months) after initial successful stent-graft placement (Fig 1).

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1. 

  Outcome of 150 patients after endoluminal stent-graft placement for the treatment of infrarenal aortic aneurysms.

We encountered complications that required explantation of the stent-graft within the first 4 years in 13.3% (20/150) of the cases (during the peri-intervention period [3]; during follow-up period [17]). Table II shows the different complications that require explantation of the stent-graft during the follow-up period. Stents were found to be dislocated, with leakage, thrombosed, ruptured, or infected. Dislocation of the stent-graft with distal displacement of its body in the direction of the blood flow occurred in four patients (Table II). There was no correlation between the length of the aortic neck and dislocation of the endoprosthesis (Table II). Contrary to expectations, in only one of four patients who required explantation of the graft because of stent dislocation, the length of the aortic neck was shorter than 15 mm at the time of the primary intervention. Interestingly, widening of the aortic neck was diagnosed in three of four patients with dislocated stent-grafts during follow-up (Table II). Additionally, two of these patients presented with an angulation of the aortic neck, which hindered a reintervention by deployment of an endovascular “cuff” between the renal arteries and displaced graft. Fig 2 shows an example of a dislocation of the stent-graft prosthesis with displacement of its body in the direction of the blood flow.

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 2. 

  Arterial subtraction angiography of a 69-year-old patient with dislocation of infrarenal aortic stent-graft prosthesis 42 months after successful implantation. Arrow is pointing to stent inside aorta. Stent is dislocated at the proximal neck. See text for further details.

Leaks requiring explantation of the stent were detected in three patients (Table II) at 14.0 months since 49.6 months after initially successful exclusion of the aneurysm.

Table II. Complications that require explantation of the stent-graft during follow-up

AAA, Abdominal aortic aneurysm.

Thrombosis of a limb of the prosthesis occurred in three patients (Table II). Two of these patients presented with recurrent thrombosis despite local and systemic thrombolytic therapy. In all three patients with thrombosis of the stent, the follow-up aortography of the aorta detected an acute angulation between the body of the bifurcated graft and its modular, iliac component.

Despite an implanted stent, rupture of the AAA occurred in two patients, 26.7 and 15.0 months after stent implantation (Fig 3, Table II).

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 3. 

  CT image of the abdomen of a 76-year-old patient with rupture of an AAA despite an implanted stent-graft. Arrow is pointing to site of rupture. See text for further details.

In both patients a follow-up CT scan 3 months or 6 weeks before the rupture of the AAA did not show any endoleak. However, there was an increase in diameter of the aneurysm of 17 mm and 20 mm.

Graft infection that required explantation of the stent was diagnosed in five patients (Table II). Patients with infected stent-grafts were admitted between 6.7 and 55.8 months after initial stent-graft placement. In three of the cases further interventions (thrombolysis, stent) had been necessary 26.7 ± 23.3 months after initial implantation (ie, 6.7 ± 4.6 months before explantation). All patients presented with recurrent fever, elevation of C-reactive peptide, leukocytosis, and persistent back or abdominal pain and were treated with broad spectrum antibiotics intravenously as soon as they were admitted to the hospital. Repeated blood cultures before explantation of the stent-graft could not recover organisms in any patient. The diagnosis of graft infection was supported with CT findings, which showed air entrapment around the stent-graft prosthesis in three patients (Fig 4) and a thickened aneurysm wall that was in direct contact with an intestinal loop in two patients.

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 4. 

  CT image of the abdomen of a 69-year-old patient with graft infection. Arrow is pointing to air inclusion around stent-graft prosthesis. See text for further details.

Although the CT findings suggested the presence of aortoduodenal fistulas, no gastrointestinal bleeding was encountered in any of the patients. However, after the adherent intestinal loops were released from the aneurysm wall, the prosthesis was directly visible in one patient, which suggested the presence of a lumen connecting the aneurysm and the enteral cavity. At operation a dense, shiny, and highly vascular reaction of the aneurysmal sac was diagnosed in all five patients with graft infection. Cultures of the aortic wall and the surface of the stent-graft were taken from all patients. One patient, who presented with a lumen in the aneurysmal wall at the time of stent-graft explantation, was positive for bacteria culture (Escherichia coli ), suggesting the presence of an aorto-enteral fistula. The cultures in the remaining patients were negative for bacteria.

All explantation procedures were performed with patients under general anesthesia. Explantation of the stents had to be performed on an emergency basis in three patients (rupture of AAA [2]; rupture of iliac artery [1]). In 17 patients the prosthesis was explantated electively. In all cases a transperitoneal approach was used, and infrarenal cross-clamping of the aorta was possible. In five of the 20 patients, we encountered complications during or after the operation. In one patient the thrombosis of the graft resulted in limb ischemia. Explantation of the graft and implantation of a Dacron graft did not rescue the leg. One patient died of multiorgan failure 6 days after stent explantation. In one patient acute renal failure developed perioperatively; in another, a wound infection developed. A third patient had respiratory failure and required temporary support.

Back to Article Outline

Discussion 

We demonstrate that stent-graft placement for treatment of AAA is associated with a considerable risk of serious complications during long-term follow-up. We encountered serious complications in 13% of the patients during the first 4 years, which were independent of the type of the implanted device. The origin of the encountered complications can be characterized by two pathomechanisms: (1) widening of the proximal neck of the aorta leading to dislocation of the stent-graft with subsequent thrombosis or leakage and (2) infection associated with the placement of a foreign body.

Neck widening and stent dislocation 

Dislocation of the stent-grafts from the proximal fixation site caused most of the complications that required stent explantation. Dislocation of the stents led to refilling of the AAA, which resulted in rupture despite an implanted stent in two of six patients and caused thrombosis of the stent because of damage to the nitinol wire frame. The pathomechanism of stent dislocation can be characterized either by progression of the disease process and subsequent widening of the aortic neck or by failure of the fixation of the stent-graft prosthesis in the first place. Because the initial implantation of those stents that had to be explantated during follow-up was without complications, in our hands, it is reasonable to conclude that failure of stent fixation is not responsible for the dislocations. However, we found widening of the aortic neck of about 25% in three of four patients with dislocations of the stent. There was no correlation between preinterventional diameter or length of the aortic neck and the dislocation. Thus, it appears that a continuation of the disease process may be responsible for most of the complications encountered and that the fixation mechanism of the stents is not designed to prevent this scenario. Although these findings are in contrast to a recent study by Walker et al,¹² who did not find any evidence for proximal neck widening and subsequent dislocation after endoluminal repair of AAA in 112 patients during a 4-year period, the presence of further neck widening has to be kept in mind because the consequences can be devastating (as documented by the patient who died after stent explantation). Our results suggest that widening of the aortic neck is a crucial part in the pathomechanism of dislocation and must be addressed in the design of stent-grafts in the future.

Stent implantation and infection 

Graft infection was the second most frequent reason for graft explantation. Two mechanisms have to be considered for the occurrence of graft infection. First, the infection may be related to the implantation procedure,¹³ especially in those cases where further interventions were required. Clearly, every intervention poses a risk for infection by contamination of the devices used or other nonspecific mechanisms. Although reinterventions were necessary in 30% of all cases (data not presented), we found no correlation between the time or the number of the interventions and the onset of symptoms. Because of the late onset of the stent-graft infection, it is likely that the pathomechanism of infection of the stent prosthesis is not related to intraoperative contamination but rather to distant septic foci or infection from intestinal organisms. The preoperative treatment with broad spectrum antibiotics might explain that blood cultures were negative for bacteria in most cases.

The second mechanism may be related to mechanical irritation of the surrounding tissue by the stent-graft causing erosion of the aortic wall and the development of an aorto-enteral fistula.¹⁴ Although we did not find any acute bleeding at the time of operation, we found a hole in the aneurysm after removal of the adhering intestines, suggesting the presence of a lumen. This mechanism would also be in agreement with the lack of microorganisms in the blood cultures. Because no bleeding was observed during explantation, one might speculate that the stents were infected from intestinal organisms through a fistula, and the inflammatory reaction of the body preceded the septic component of this process. If this mechanism holds true, the long-term durability of these stents may be rather grim.

Complications associated with or despite the explantation 

Explantation of the devices was easy to perform in each of our patients, although we encountered perioperative complications in 25% of our patients, including one death.¹⁵ Infrarenal clamping of the aorta was possible in every patient because all stent-grafts had been fixated below the renal arteries. Other investigators recommend suprarenal deployment of the stents to prevent dislocation.¹⁶ If stent-graft explantation becomes necessary in patients with suprarenal fixation, suprarenal clamping will be necessary during explantation. This procedure will be associated with ischemia to the kidneys, which increases the perioperative risk substantially. In our patients, dislocation was unrelated to the diameter or the length of the aneurysm neck at the time of implantation. For these reasons, we advise caution for the liberal use of suprarenal fixation of aortic stent-grafts.

Limitations of the study 

The accuracy of the data presented may be limited by an incomplete follow-up. It is most likely that we underestimated rather than overestimated the true number of complications. Although it is reasonable to conclude that most patients with serious complications and symptoms that require explantation are included in the study, it is conceivable that more patients with stents already require reintervention or explanation. Because of the symptom-free nature of the disease and the excellent immediate results of stent implantation, it may be difficult to improve follow-up compliance of the patients. However, regular CT scans (every 6-12 months) appear to be an absolutely necessary component of the follow-up process to guarantee patient safety. In addition, criteria have to be defined when explantation may be indicated, even in the absence of leaks or thrombosis. The two cases where rupture of the aneurysm occurred shortly after an uneventful CT scan support this notion.

Back to Article Outline

Conclusions 

We conclude that

Back to Article Outline

Acknowledgements 

We wish to thank Dr Koppany Sarai for surgical guidance and the many helpful discussions, and Heinrich Taegtmeyer, MD, DPhil, for valuable suggestions.

Back to Article Outline

References 

1. Podore PC, Throop EB. Infrarenal aortic surgery with a 3-day hospital stay: a report on success with a clinical pathway. Vasc Surg. 1999;29:787–792
   • View In Article
2. Moore WS, Vescera CL. Repair of abdominal aortic aneurysm by transfemoral endovascular graft placement. Ann Surg. 1994;220:331–339
   • View In Article
   • MEDLINE
   • CrossRef
3. Moore WS, Rutherford RB. Transfemoral endovascular repair of abdominal aneurysm: results of the North American EVT phase 1 trial. J Vasc Surg. 1996;23:543–553
   • View In Article
   • Abstract
   • Full-Text PDF (2853 KB)
   • CrossRef
4. May J, White GH, Yu W, Ly CN, Waugh R, Stephen MS, et al.  Concurrent comparison of endoluminal versus open repair in the treatment of abdominal aortic aneurysms: analysis of 303 patients by life table method. J Vasc Surg. 1998;27:213–220
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (81 KB)
   • CrossRef
5. Henretta JP, Hodgson KJ, Mattos MA, Karch LA, Hurlbert SN, Sternbach Y, et al.  Feasibility of endovascular repair of abdominal aortic aneurysms with local anesthesia with intravenous sedation. J Vasc Surg. 1999;29:793–798
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (296 KB)
   • CrossRef
6. Blum U, Langer M, Spillner G, Mialhe C, Beyersdorf F, Buitrago-Tellez C, et al.  Abdominal aortic aneurysms: preliminary technical and clinical results with transfemoral placement of endovascular self-expanding stent-grafts. Radiology. 1996;198:25–31
   • View In Article
   • MEDLINE
7. Blum U, Voshage G, Lammer J, Beyersdorf F, Tollner D, Kretschmer G, et al.  Endoluminal stent-grafts for infrarenal abdominal aortic aneurysms. N Engl J Med. 1997;336:13–20
   • View In Article
   • MEDLINE
   • CrossRef
8. Blum U, Voshage G, Beyersdorf F, Tollner D, Spillner G, Morgenroth A, et al.  Two-center German experience with aortic endografting. J Endovasc Surg. 1997;4:137–146
   • View In Article
   • MEDLINE
   • CrossRef
9. Becquemin JP, Lapie V, Favre JP, Rousseau H. Mid-term results of a second generation bifurcated endovascular graft for abdominal aortic aneurysm repair: the French Vanguard trial. Vasc Surg. 1999;30:209–218
   • View In Article
10. Allenberg JR, Schumacher H. Endovascular reconstruction of infrarenal aortic aneurysm (AAA). Chirurg. 1995;66:870–877
    • View In Article
    • MEDLINE
11. Hagenmüller GW, Hold M, Ptakovsky H. Surgery of infrarenal abdominal aortic aneurysm. Chirurg. 1995;66:857–869
    • View In Article
    • MEDLINE
12. Walker SR, Macierewicz J, Elmarasy NM, Gregson RH, Whitaker SC, Hopkinson BR. A prospective study to assess changes in proximal aortic neck dimensions after endovascular repair of abdominal aortic aneurysms. J Vasc Surg. 1999;29:625–630
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (210 KB)
    • CrossRef
13. Heikkinen L, Valtonen M, Lepantalo M, Saimanen E, Jarvinen A. Infrarenal endoluminal bifurcated stent graft infected with Listeria monocytogenes. J Vasc Surg. 1999;29:554–556
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (64 KB)
    • CrossRef
14. Sergeant PT, Derom F. Aorto-enteric fistula as clinical entity. Acta Chir Belg. 1978;77:187–194
    • View In Article
    • MEDLINE
15. Jacobowitz GR, Lee AM, Riles TS. Immediate and late explantation of endovascular aortic grafts: the endovascular technologies experience. J Vasc Surg. 1999;29:309–316
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (74 KB)
    • CrossRef
16. MacIerewicz J, Walker SR, Vincent R, Wastie M, Elmarasy N, Hopkinson BR. Vascular surgical society of great britain and ireland: perioperative renal function following endovascular repair of abdominal aortic aneurysm with suprarenal and infrarenal stents. Br J Surg. 1999;86:696
    • View In Article