Transluminal Stent-Graft Placements for the Treatments of Acute Onset and Chronic Aortic Dissections
Background Transluminal stent-graft placement (TSGP) for aortic dissection is a relatively new procedure. We performed TSGPs to seal the primary entry site to treat and prevent complications of aortic dissection. The early to mid-term outcomes were analyzed.
Methodsand Results—Thirty-seven patients with a primary intimal tear in descending aorta underwent TSPG. TSGP was performed in 16 acute onset dissections (AOD) with dissection-related complications instead of emergency surgery. Eight AOD without complications were treated to prevent aneurysmal enlargement. Thirteen chronic dissections were treated to prevent rupture. TSGP was technically successful in all cases. One patient with prehospital rupture died. The hospital mortality rate was of 2.7% overall, 6.3% in AOD with complications, 0% in AOD without complications and in chronic dissections. One persistent endoleak required open surgery, and 1 intimal tear was caused by the stent-graft, necessitating an additional TSGP. The primary success rate was 94.4% overall. After hospital discharge, no patient died or suffered aortic rupture during an average follow-up of 24.5 months. New intimal tears caused by the stent-grafts and a secondary endoleak developed in 3 AOD patients. One open procedure and 2 additional TSGPs were performed. Actuarial survival rate and cardiovascular event-free rate at 2 years are 97.3% and 78.3% overall, 93.8% and 48.0% in AOD with complications, 100% and 87.5% in AOD without complications, and both 100% in chronic dissections.
Conclusions TSGP is a reasonable treatment option for aortic dissection. However, delayed intimal tear formations caused by the stent-graft is a problem that requires further investigation.
Traditional management of patients with aortic dissection includes both medical and surgical. Most Stanford type B acute aortic dissections are managed medically. However, in complicated type B acute dissection such as those with rupture, impending rupture, end-organ ischemia, surgery is mandatory, although its outcome is poor.1–3 Moreover, aneurysmal enlargement in the chronic phase of dissection is another concern because it necessitates the use of extended graft replacements for thoracic and thoracoabdominal lesions. Surgical results of these procedures have improved markedly, but mortality and morbidity remain high. Even in recent series from major institutions, the mortality rare was reported to be 4.8% to 8.8%.2–4 These reports also indicated that extension of the aneurysm is a major risk factor for paraplegia.
The management of acute type A retrograde dissections with intimal tears in the descending aorta also is controversial. Originally, ascending aortic replacement alone was performed for this subtype.5 Graft replacement of just the ascending aorta for this subtype of dissection retains the origin of the dissection and the intimal tear in the arch and descending thoracic aorta, and the postoperative risk of rupture of the false lumen remains. On the other hand, the operative risk of extended aortic arch replacement is still high, despite recent advances in surgical technique.6
Some clinical experiences using transluminal stent-graft placement for descending aortic disease have been reported.7–13 However, this procedure is not a common treatment for aortic dissection, especially during the acute onset stage.
In our institute, we use transluminal stent-graft placement to seal the primary entry site in the descending thoracic aorta for both chronic dissections and acute onset dissections. We had 2 goals for stent-grafting for acute onset dissections. The first was to minimize surgical risk by decreasing the risk of complications of emergent operation, and the second was to prevent aneurysmal enlargement and necessity of extended replacement of aorta in the chronic phase. This report describes our early to mid-term results using transluminal stent-graft placement in acute onset and chronic aortic dissections. The feasibility of this treatment is discussed.
Between July 1997 and December 2000, 37 consecutive patients with aortic dissection because of an intimal tear in the descending aorta underwent transluminal stent-graft placement at Mie University Hospital. Approval by the Institutional Review Board and informed consent from each patient was obtained. There were 9 patients with acute type A retrograde aortic dissection and 28 patients with type B dissections. Thirteen patients had chronic dissections and 24 patients were acute onset dissections.
Patients with acute onset aortic dissection, who were hemodynamically unstable were excluded from this treatment. Inclusion criteria were follows: patients with dissection-related complications, such as rupture, impending rupture, and end-organ ischemia and patients at high risk for aneurysmal enlargement in the chronic phase, where the maximum aortic diameter was ≧ 40 mm on computed tomography (CT).14 Acute onset patients with complications who were hemodynamically stable underwent digital subtraction angiography (DSA) on admission using a caliber catheter to evaluate distance of landing zone, existence of multiple entries, blood supply to brunched arteries, existence of static obstruction, and feasibility of access route. When all findings seemed to be suitable, we subsequently performed emergent translumimal stent-graft placement.
In the treatment of chronic dissection, patients who had growing ulcer-like projections or whose maximal aortic diameter was ≧ 50 mm on CT, were enrolled to this study. All patients with chronic dissection underwent DSA to evaluate the blood supply of the major brunch vessels. Patients whose major brunch vessels were supplied from false lumen were excluded from this treatment.
The final set of inclusion criteria were:
Identification of a primary entry tear in the descending thoracic aorta
A primary entry tear located at least 1 cm distal to the orifice of the left subclavian artery
A tear located in the descending aorta proximal to Th10
Absence of severe dilatation (> 38 mm in diameter) or severe atherosclerotic change of landing zone for stent-grafting
No evidence of severe aortic regurgitation
No evidence of coronary artery or aortic arch branch ischemia
Femoral and iliac arteries of sufficient size and quality to permit passage of at least an 18Fr catheter
Among acute onset patients, 9 patients with acute type A retrograde dissection underwent this treatment in lieu of emergent operation to prevent rupture. Seven patients with acute type B dissection developed complications because of the dissection: persistent back pain despite medical therapy in 3 patients, rupture in 3 patients, ischemia of the kidneys and lower extremities in 1 patient. These patients also underwent stent-grafting instead of emergent operation. Eight patients with type B acute dissection at high risk for aneurysmal enlargement in the chronic phase, whose maximal aortic diameter was ≧ 40 mm on CT, underwent scheduled stent-graft placement to prevent the enlargement and rupture. Among chronic dissections, 4 patients with a type B dissection demonstrated a growing ulcer-like projection, and 9 patients with type B dissection had an aneurysm whose maximal aortic diameter was ≧ 50 mm on CT.
Table 1 shows the baseline demographics and clinical characteristics of the patients.
In cases of complicated acute onset dissection, emergent stent-graft placement was performed on the day of admission, following the diagnostic DSA. When complications were not present, multiplannar reconstruction and 3-D images were created from spiral CT better appreciate the configuration of the primary entry tear and the arch vessels. The stent-graft was placed electively in such cases.
In the acute phase, the long and short axis of the true lumen, and the diameter of entire aorta just proximal to the entry tear were measured by CT to determine which size stent-graft to use for the repair. In the chronic phase, the same measurements were obtained 2 cm distal to the primary entry tear. A stent-graft was chosen whose proximal diameter was 10% greater than the average of the maximum and minimum diameters of the true lumen, not exceeding the diameter of the entire aorta in the acute phase. A stent-graft was chosen whose proximal diameter was 20% greater than the average in the chronic phase. When the mean diameter of the true lumen at the proximal level was 25% larger than it was at the level 2 cm distal to the primary entry tear, a tapered stent-graft was used. The diameter of the distal end of the tapered stent-graft was designed by 20% over-sizing of the mean diameter of true lumen at the level 2 cm distal to the primary entry tear.
The stent-grafts were constructed of self-expanding, Gianturco stainless steel Z stent (Cook Critical Care, Inc, Bloomington, IN) covered with balloon expanded-expanded polytetra fluoroethylene (e-PTFE) grafts (Bard, Inc, Tempe, AZ). The procedure for stent-graft placement, has previously been described in detail.8,15 All implantations were performed under general anesthesia with tracheal intubation in the angiography suite. Coordinated teams of cardiovascular surgeons and interventional radiologists performed the procedures.
Follow-up CT performed within 4 weeks of the stent-graft implantation determined the extent of thrombosis in the aortic false lumen, the size of the true and false lumens, and blood flow through the branch vessels. Spiral CT was performed 3 and 6 months after graft placement, yearly thereafter to evaluate the thoracic and abdominal aorta. The diameter of the entire aorta and the false lumen just distal to the distal end of the stent-graft was measured by CT. Clinical information was compiled by retrospective review of medical records and by contacting their patients’ treating physicians. Current follow-up status was determined in June 2001 and was 100% complete. The mean follow-up period was 24.5±12.4 months (range, 4 to 45 months).
The therapeutic definitions, which were described by Dake et al,9 were used to characterize how well the endovascular stent-graft worked.
Clinically important rates are reported with 95% confidence interval (95% CI), and continuous variables are expressed as means±SD. The Kaplan-Meier method was used to generate survival curves, which are reported as ±1 SE.
Stent-graft deployment in the true lumen of the aorta was technically successful in all cases (Table 2). Forty-nine stent-grafts were used in 37 patients (1.3 per patient). Thirty-five stent-grafts were the tube type and 14 were tapered. The diameter of the stent-graft was 34.1±3.3 mm (range, 30 to 42 mm) proximally and 31.9±3.1 mm (range, 28 to 38 mm) distally. The duration of procedure was 102±29 minute (range, 60 to 170 minute).
Early Results of Stent-Grafting
One patient with acute complicated aortic dissection died 9 days after the procedure. The procedura 30 day mortality in the patients with the acute complicated dissection is therefore, 6.3% (95% CI, 0.2 to 30.2%) with an overall mortality of 2.7% (95% CI, 0.1 to 14.2%). The patient who died was a 75-year-old man who had a ruptured Stanford type B dissection and mesenteric ischemia preoperatively. Pneumonia and multiple organ failure developed after the stent-graft procedure and he died 9 days after the procedure. Postmortem examination revealed that the primary intimal tear was completely covered and that the false lumen had thrombosed. There was no evidence of mesenteric infarction.
Two patients suffered major complications. One patient with acute onset type B dissection experienced rupture of a true abdominal aortic aneurysm, which was 55 mm in maximum diameter and separate from the dissection, 1 day post procedure. An emergent operation was performed successfully. On the intra-operative finding, there was no vascular injury caused by instrumentation from the stent-grafting. The other patient developed Methicillin-resistant Staphylococcus aureus pneumonia requiring tracheostomy, but he recovered. Minor complication included 1 minor stroke and 1 superficial wound infection. Therefore, the early morbidity rate was 10.8% (95% CI, 3.0 to 25.4%) overall, 18.8% (95% CI, 4.0 to 45.6%) in acute onset complicated dissections, 0% (95% CI, 0 to 36.9%) in acute onset uncomplicated dissections, and 7.7% (95% CI, 0.2 to 36.0%) in chronic dissections.
Initial endoleaks were documented by completion DSA in 5 cases 13.5% of overall, 20.8% in acute onset cases, and 0% in chronic cases.
However, thrombosis of the false lumen at the level of the stent-graft was documented in 35 out of 36 cases by the first postoperative CT scan. CT showed a persistent endoleak in 1 patient with acute onset type B uncomplicated dissection, who had an initial endoleak. There were no endoleaks in the remaining 4 patients who presented with initial endoleaks. The patient with the persistent endoleak underwent an open procedure 4 months after graft placement because of continuing enlargement of the aortic diameter. A new intimal tear because of injury caused by the distal end of the stent-graft and blood flow into the false lumen were identified in 1 patient by follow-up CT scan 4 weeks after the procedure without symptoms. Additional transluminal stent-graft placement was successfully performed without complication in this patient (Figure 1). Therefore, the primary success rate were 94.4% (95% CI, 81.3 to 99.3%) overall, 93.3% (95% CI, 68.1 to 99.8%) in acute onset complicated dissections, 87.5% (95% CI, 47.3 to 99.7%) in acute onset uncomplicated dissections, and 100% (95% CI, 75.3% to 100%) in chronic dissections (Table 3).
Late Survival and Complications
No patient died during long-term follow-up. Neither was there any case of documented aortic rupture.
Follow-up imaging studies after discharge identified new intimal tears at the proximal end of the stent-graft 3 months after placement in 2 patients with acute dissection. A local dissecting aneurysm occurred in 1 patient, and conventional graft replacement was performed successfully 4 months later. In the other case, additional stent-graft placement was successful and the new tear was covered completely. A secondary endoleak developed 6 months after the initial placement because of a change in the stent-graft configuration. The stent-graft required the additional placement, after which the endoleak was controlled and the size of false lumen decreased (Figure 1).
Three new aortic dissections developed in different segments than the original lesion in 2 patients. In the first patient, chronic aortic dissection at the level of proximal to original stent-grafting segment was observed by follow-up CT, 6 months later. Primary intimal tear of the chronic dissection was located on 3 cm distal to origin of left subclavian artery and additional stent-grafting was performed and the new entry tear was sealed. The second patient developed a new ulcer-like projection distal to original stent-grafting segment. Because dilatation of the dissecting aorta and growth of the new ulcer-like projection were not identified, medical therapy and imaging follow-up was continued. A Stanford type A acute dissection suddenly occurred 22 months after the stent-grafting. Emergent extended aortic arch replacement was performed successfully. The entry tear located in the aortic arch was unrelated to the stent-grafting. One patient developed an acute myocardial infarction 14 months after stent-graft placement and was treated medically. All these cardiovascular events occurred inpatients with acute onset dissection (Table 4).
The overall actuarial cardiovascular event-free survival was 85.8%±5.9% at 1 year and 78.3%±7.4% at 2 years. The rate was 72.0±12.0% at 1 year and 48.0%±16.0% at 2 years inpatients with acute onset complicated dissections, 87.5%±11.7% at 1 year and 87.5%±11.7% at 2 year with uncomplicated complications. The rate was 100%±0% both at 1 and 2 years in chronic dissections (Figure 2A). The overall actuarial survival rate was 97.3%±2.7% at 1 year and 97.3%±2.7% at 2 years. The rate was 93.8%±6.1% at 1 year and 93.8%±6.1% at 2 years inpatients with acute onset complicated dissections and 100%±0% at 1 year and at 2 years with acute onset uncomplicated dissections. The rates were also 100%±0% at 1 and 2 years inpatients with chronic dissection (Figure 2B).
Remodeling of the Aorta
Complete or partial thrombosis of the false thoracic aortic lumen was identified in 34 of 36 patients (94.4%, 95% CI, 81.3 to 99.3%) by last available CT from 6 to 36 months after (average±SD, 21.5±1.1 month). The diameter of the entire aorta and the false lumen was measured by CT. Decrease (a reduction of at least 5 mm compared with the size before stent-graft placement) in the size of the false lumen just distal to the distal end of the stent-graft was observed in 32 of 36 cases overall (88.9%, 95% CI, 73.9 to 96.9%), in 14 out of 15 patients with acute onset complicated dissections (93.3%, 95% CI, 68.1 to 99.8%), in 7 out of 8 patients with acute onset uncomplicated dissections (87.5%, 95% CI, 47.3 to 96.8%) and in 11 out of 13 patients with chronic dissection (84.6%, 95% CI, 54.6 to 98.1%). Complete obliteration of the false lumen (≦5 mm) was achieved in 21 out of 36 cases overall (58.3%, 95% CI, 40.8 to 74.5%), in 10 of 15 patients with acute onset complicated dissections (66.7%, 95% CI, 38.9 to 88.2%), in 6 of 8 patients with acute onset uncomplicated dissections (75%, 95% CI, 34.9 to 96.8%) and in 5 out of 13 patients with chronic dissection (38.5%, 95% CI, 13.9 to 68.4%) (Figure 3). Complete obliteration of the false lumen of the ascending aorta was confirmed in all patients with type A dissection on the CT.
In 1994, Dake et al reported the feasibility of endovascular stent-grafting for descending thoracic aneurysms.7 Recently, the Stanford University group and our group have reported favorable preliminary clinical results using stent-grafting to treat acute aortic dissection.10 We also have reported our clinical experience with stent-grafting for some subtypes of acute dissections, Stanford type A retrograde dissection and type B dissection with true lumen stenoses.15,16 Sealing the primary entry tear by stent-grafting was very effective in both subtypes of dissection.
Based on this experience, we initiated a clinical trial of transluminal stent-graft placement to seal the primary entry site in the descending thoracic aorta, not only in chronic aortic dissection, but also in cases of acute onset. We had 2 goals of treatment in cases of acute onset. The first was the management of any complication attributable to the dissection as an alternative to emergent surgery. The second was to prevent aneurysmal enlargement in the chronic phase, thus obviating the eventual need for extended aortic replacement.
Early treatment of acute dissection with complication resulted in 1 hospital death (6.3%). However, this patient had severe metabolic acidosis at the time of admission because of rupture and intestinal ischemia. His preoperative condition determined the postoperative course, even though the procedure itself was successful. Considering the patient’s clinical situation, we feel the outcome was not unexpected and the overall results were acceptable.
The cardiovascular event-free rate is low for acute stage dissections with complications. The main reason is the necessity for additional stent-grafting or open surgery in cases of new intimal tears. The exact mechanism responsible for the formation of these intimal tears is not clear. It is possible that local forces directed against the intima are produced by bending of the stent-graft and friability of the intima in acute stage dissection. More flexible stent-grafts should be developed to reduce the risk of this delayed complication.
The results of stent-grafting to prevent aneurysmal enlargement for acute onset dissection are acceptable. No patient died, and only 1 patient suffered aneurysmal enlargement because of persistent endoleak, probably because of the large diameter of the proximal landing zone. We consider the cause of this complication to be physician error: inappropriate patient selection. We currently consider a landing zone larger than 38 mm to be an exclusion criterion. No other patients has experienced aneurysmal change during follow-up. The false lumen in the thoracic aorta became smaller or disappeared in 93.3% of patients with complications and in 87.5% patients without complications.
These preliminary data suggest that endovascular stent-grafting is a viable treatment for acute onset aortic dissection.
The clinical results of the stent-grafting for chronic dissection are also acceptable, as previously reported.9,11 There were no deaths or ruptures in our series. The primary entry site was sealed and thrombosis of the false lumen in the thoracic aorta was accomplished in all patients. The size of false lumen was reduced in 84.6% patients and did not subsequently enlarge in any case. These results of endo-grafting for chronic dissections are better than those for true thoracic aortic aneurysms.9,11–13 We believe that our results are better because the characteristics of the landing zone were relatively good. Most true aneurysms have severe atherosclerotic changes in the aorta and access route arteries. Atheromatous plaques carry a high risk of distal embolism, the incidence of which ranges from 6.3% to 8%.9,12,13 Conically shaped landing zones in true aneurysms also have a high risk of endoleak, the incidence of which ranges from 6.3% to 24%.9,12,13 The incidence of embolism associated with stent-grafting and the incidence of persistent endoleak were quite low in our dissection series, compared with the incidence in true aneurysms. A distance of landing zone is very important issue to prevent endoleak. Many dissections have primary entry tears cross to the orifices of the left subclavian arteries. We excluded the patients whose primary entry tears did not locate at least 1 cm distal to the orifice of the left subclavian artery. Among the patients with tears located at least 1 cm distal to the orifice, existence of entry tear located at small curvature of aortic arch shows high risk of endoleak. We intentionally occluded an orifice of left subclavian artery in such a case. One patient was occluded the orifice with the stent-graft but he did not show ischemia of his upper extremity and brain.
Stent-grafting decreased the maximum size of the thrombosed false lumen without endoleak in all patients with acute onset dissection, but had not decreased the size in 15.4% of chronic dissections as determined by the follow-up CT. A similar phenomenon already has been reported for stent-grafting of true aneurysm, and its incidence is very high. The size of thrombosed aneurysmal sacs without endoleak did not decrease in 62% to 82% patients with true aneurysms of the descending thoracic or abdominal aorta.17–19 The concept of endotension because of unobservable endoleak and pressure transmittance through thrombus has recently been suggested as a cause of this phenomenon.20 There is a possibility that retrograde flow into the false lumen of thoraco-abdominal or abdominal aorta through re-entry tears retains the size of the thrombosed false lumen just distal to the distal end of the stent-graft. Effectiveness of treatments for dissections are affected by multiple entries or re-entries sites into false lumens. Although enlargement of the false lumen was not observed in this study, careful follow-up is mandatory.
The relatively small number of patients and brief follow-up period are insufficient to make unqualified recommendations. However, these data suggest that transluminal stent-graft placement is an attractive treatment option for aortic dissection. It is possible that the strategy for managing acute dissection will evolve away from open procedures to less invasive therapies, such as stent-grafting.
Correspondence to Takatsugu Shimono, MD., Department of Thoracic and Cardiovascular Surgery, Mie University, School of Medicine, 2–174 Edobashi, Tsu, Mie 514-8507, Japan. E-mail firstname.lastname@example.org