Article Text

Download PDFPDF
Aortic disease: thoracic endovascular aortic repair
  1. Colin Bicknell1,
  2. Janet T Powell2
  1. 1Department of Surgery & Cancer, Imperial College London, St Mary's Hospital, London, UK
  2. 2Department of Surgery & Cancer, Imperial College London, Charing Cross Hospital, London, UK
  1. Correspondence to Professor Janet T Powell or Colin Bicknell, Department of Surgery & Cancer, Imperial College London, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, UK; j.powell{at} or colin.bicknell{at}


Following the first reported case of thoracic endovascular aortic repair (TEVAR) in 1987, there has been rapid technological development and widespread uptake of this approach for thoracic aortic disease. TEVAR has particular advantages for acute thoracic syndromes and there is continuing development of TEVAR for use in the elective situation. This is still a young technology and many unknowns remain, including long-term outcomes and the relative advantages and disadvantages versus both conservative treatment and open surgical repair. The current uses of TEVAR in the treatment of aortic dissection, thoracic aortic aneurysm, penetrating aortic ulcer and trauma are discussed.

Statistics from

The first case of an aortic endograft being used was by Volodos et al to repair a traumatic thoracic aortic aneurysm in Ukraine in 1987.1 The first reported case of a thoracic endograft in the Americas was by Parodi several years later.2 Subsequently, in late 1994, Dake et al3 reported the successful placement of aortic endografts in 13 patients with descending thoracic aortic aneurysm, principally for degenerative disease but including two postdissection aneurysms, one post-trauma and one following repair of coarctation. All these early endografts were ‘homemade’ but commercial interest was awakened and by 2005 there were several approved endografts commercially available. Thoracic endovascular aortic repair (TEVAR) was born and enthusiastically adopted by interventionalists coming from different disciplines, interventional radiologists in the USA, vascular surgeons in the UK and cardiologists in some European countries for the treatment of aneurysms, dissection, penetrating ulcers and trauma in the descending thoracic aorta. Progressive improvements and technological advance now enable TEVAR to be used for more proximal (through to the arch of the aorta) and distal (thoracoabdominal) disease. Here, we discuss briefly the current indications for TEVAR focusing on its advantages, limitations and, since this is a relatively new technology, longer-term unknowns.


The most common form of dissection is type A, with the entry tear in the aorta proximal to the left subclavian artery (LSCA). TEVAR has only been used to manage acute type A dissection in specialised centres, in high risk patients, but with acceptable early and mid-term outcomes in small series.4 This series reports on nine patients who were not considered for open surgery (age 51–91 years, four men, seven dissections extending into the aortic root using mainly Gore or Cook endografts). There were three inhospital deaths (one aorto-atrial fistula, one following open conversion and one following endoleak and rupture in a patient who refused blood transfusion): this compares with an inhospital mortality from open surgery of 25% in better risk patients in the International Registry of Aortic Diseases. There were three late deaths (one possibly related to a postprocedural stroke) and three patients have fared well.

In contrast, TEVAR is increasingly being used to manage acute type B aortic dissection (distal to the LSCA), a disorder which appears to be increasing in incidence in the Western world.5 For type B dissection, stenting of the aorta at the level of the proximal entry tear may reduce flow in the false lumen, expand the true lumen diameter, and significantly improve visceral, renal, spinal and lower limb perfusion.

Uncomplicated patients (with no rupture, malperfusion or continuing pain and adequate blood pressure control) make up approximately three-quarters of patients after primary dissection6 and are managed medically as standard practice.7 Conversely, in type B dissection complicated by malperfusion, TEVAR is effective in immediate restoration of flow where there is dynamic occlusion of branch vessels seen clearly on digital subtraction imaging. For static and persistent occlusion, subsequent renal, visceral and iliac stenting also may be required.

There are significant advantages to TEVAR versus open surgery in the acute situation, which arise from the minimally invasive, image guided approach to reperfusion. Analysis of the International Registry of Aortic Dissection database and meta-analyses suggest that the mortality rate is significantly reduced with treatment by TEVAR, rather than open surgery.8 ,9 These data, of course, are in selected patients. However, a cautious presumption is made that the treated group are the complicated group, highlighting a significant advantage in endovascular treatment over open surgery.

TEVAR is certainly not without risks. These are summarised in table 1. Paraplegia rates are largely dependent on the length of aorta covered, indication for repair and patient stability and may reach 10% in the highest risk patients. Equally, neurological injury from stroke is a major cause of adverse outcomes. The risks of neurological injury are known to be lower than that in aneurysmal disease—with wire manipulation in a non-atherosclerotic aortic arch the embolic risks may be reduced.10 One great concern following TEVAR for type B dissection is proximal extension (type A progression). This complication is rare; systematic reviews of the literature suggest that the risk of type A progression is about 2% with increased risk in those with stent oversizing.11

Table 1

Outcomes following thoracic endovascular aortic repair for non-acute type B dissection

The current debate in this area is largely centred on strategies to reduce long-term sequelae of dissection. Conservative treatment for uncomplicated dissection is successful in reducing early mortality when compared with open surgery in the acute setting.12 ,13 However, if the risk of acute intervention was reduced sufficiently with minimally invasive therapy, early TEVAR may reduce long-term dilatation and rupture or need for more extensive (hybrid surgical visceral revascularisation and stenting or open thoracoabdominal surgery).

In those patients with uncomplicated dissection, TEVAR before the chronic phase (within about 6 weeks following the index dissection) appears safe. The mortality rate from stenting uncomplicated stable patients in the ADSORB trial was 0% with significant early remodelling of the dissection after stenting.14 This is much more impressive than stenting in chronic dissections15 adding to the case for early stenting of uncomplicated type B dissection.

In the only randomised controlled trial comparing TEVAR in the subacute period versus conservative treatment for uncomplicated type B dissection using mortality as a primary endpoint,16 2-year data show no difference in outcome. Much commentary was directed at potential design flaws in this trial, not least pointing out that many patients were stented more than 3 months after the index event (dissection) in the chronic phase when aortic remodelling was less likely to take place and that aneurysmal dilatation often occurs later than 2 years in conservatively managed patients. The study was therefore extended to 5 years.17 The data at this point suggest there may be an advantage to stenting early due to a number of late, sudden deaths in the conservative treatment group. Again, there has been scepticism of these results focusing on the methods of landmark analysis used, but the enthusiasm for further investigation of early stenting remains.

Aneurysmal disease

For aneurysmal disease in the arch and thoracic aorta, the advantages of TEVAR over open surgery are easy to understand. Ever since the revolutionary publication of 13 successful cases by Dake and colleagues,3 there has been an exponential rise in the volume of stenting in the UK18 and elsewhere. Patients undergoing TEVAR are significantly older than patients with abdominal aortic aneurysm and may have significant age-related comorbidity. The open ‘alternative’ to stenting involves thoracotomy and left heart bypass. Therefore, even the most ‘endo-resistant’ of cardiovascular surgeons have to admit that thoracic stenting is an attractive alternative. There certainly is evidence that many aortic-related emergencies are improved by transfer and management in specialist, high-volume centres, where TEVAR is available.

It is difficult to imagine that there will ever be a randomised controlled trial of elective TEVAR versus surgery for descending thoracic aortic aneurysm. Systematic reviews have attempted to compare these treatment strategies and have shown benefit, but in selected patients. At present, a large scale extended UK registry may provide the best data on the comparative results

Despite the enthusiasm for TEVAR of degenerative thoracic aneurysms in patients with atherosclerotic (non-genetic) disease, a number of large-scale registry and National data sources highlight the potential morbidity and mortality in those undergoing TEVAR (table 2). The results highlight two things of interest. First, for these aneurysms, the risk stroke/paraplegia and death that is far in excess of that associated with TEVAR for other pathologies. The complications of TEVAR appear to be related to the underlying pathology.10 Second, it seems that longer-term survival is poor for patients with thoracic aneurysms. Examination of large-scale datasets that have followed TEVAR aneurysm patients for 1, 3 and 5 years have highlighted a mortality that is far in excess of the age matched population or of those who have had open surgical aneurysm repair.19

Table 2

Outcomes following thoracic endovascular aortic repair for descending thoracic aortic aneurysms

Some of these observations are likely to result from the selection of frailer and higher risk patients for TEVAR rather than open surgery or the treatment of more extensive aneurysms as the technology has improved.

For treatment of aneurysms that extend into the aortic arch, hybrid grafting techniques have found favour as endovascular specialists have considered this problem. Extra-anatomical bypass surgery techniques such as carotid–subclavian bypass and carotid–carotid cross overgrafting allow stent graft coverage of supra-aortic branch vessels and avoid thoracocotomy or sternotomy in, often elderly, patients. For extensive aneurysms of the aortic arch, sternotomy and revascularisation of all vessels from the ascending aorta are possible with stent coverage of the entire arch landing the proximal stent in the ascending aorta (figure 1). These techniques have been successful as they are versatile and avoid circulatory arrest or left heart bypass. There are some custom made devices and devices that can be applied to avoid extra-anatomical bypass. The expense and the production time at present have limited large-scale use of these grafts.

Figure 1

Extra-anatomical approaches to extending the landing zone of thoracic aortic stents. (A) Carotid–subclavian bypass graft. A left common carotid artery (LCCA) to left subclavian artery (LSCA) bypass graft with occlusion of LSCA origin can be performed to perfuse the left arm and left vertebral artery as the thoracic stent is landed between the LCCA and LSCA. (B) Carotid–carotid bypass and carotid–subclavian bypass grafting to allow stent graft placement between the innominate artery and LCCA. (C) A ‘full arch hybrid approach’ with grafts from the ascending aorta to the right common carotid artery, LCCA and a carotid–subclavian bypass graft to allow stent graft placement in the ascending aorta.

Similar techniques can be used for distally extending aneurysms—into the visceral aortic segment. Visceral hybrid techniques (retrograde revascularisation of the renal and mesenteric arteries from the aorta or iliac arteries and subsequent stenting of the aneurysm) have allowed these patients to be treated without thoracoabdominal incision, reducing the risk of respiratory compromise and reduced visceral ischaemic times known to be crucial in determining mortality.20 Early reports from hybrid techniques were exceptionally promising. However, as the technique was widened to a greater number of patients, a much better idea of the mortality and morbidity of this procedure was understood.21 This procedure has its place in the current practice, but as experience has been gained we know that the re-intervention rate is high, particularly in younger patients and those with connective tissue disorders.

Treatment of thoracoabdominal aneurysms with custom made fenestrated and branched stent designs has become more widespread with good results reported from highly-specialised endovascular units.22 Newer stent graft designs have allowed more flexibility and repositionable fenestrations; improved techniques and greater knowledge have tackled previously unsuitable cases including those with dissection associated aneurysmal degeneration, improved navigation using CT onlay techniques and improved catheter technologies such as robotic platforms may further improve the suitability of patients for this technique.

The development of a totally endovascular approach to treatment of arch aneurysms, with branched endografts, is the current focus of many industry run R&D departments (figure 2). All of the major stent graft companies have tackled this problem. While avoiding sternotomy and either full arch replacement or debranching and stenting is certainly an attractive prospect, with the prospect of a fast recovery, there may be significant risk involved. The risk of stroke may be significant, with manipulation of stent grafts and catheters in the arch that contains complicated plaque or a thrombus-lined aneurysm. Early results are difficult to put into context as the numbers are low in each series and experience in this area needs to grow before an informed choice can be made by surgeon and patient.

Figure 2

The arch branched device available from Bolton Relay (with company permission). The branches are designed to be introduced from the left carotid and innominate arteries in turn. A postimplantation image is shown in the right hand panel.

The place of TEVAR for young patients, especially those with connective tissue disorders, is uncertain. With the long-term risk of further aortic dilatation or lengthening and subsequent endoleak or rupture, many suggest that open surgery is the treatment of choice in the younger patients. In contrast, in the emergency situation of thoracic aneurysm rupture, TEVAR may offer advantages over open surgery.18

PAU and local pathologies

While dissection of the aorta (and intramural haematoma) is responsible for most presentations of acute aortic syndromes, penetrating atherosclerotic ulcer (PAU) is an important condition, especially when considering the treatment of aortic pathology with TEVAR. This and other localised aortic pathologies (such as aneurysm after coarctation patch repair and false aneurysms after open surgery) lend themselves very well indeed to treatment with short thoracic stents. The early morbidity and mortality following TEVAR for PAU can be high (10%–20%) and many consider that TEVAR should be reserved for patients with saccular aneurysm, rupture or intractable pain.23 Incidental ulceration found on imaging for other reasons can most often be managed conservatively.

Series containing patients with other localised pathologies alone are difficult to find; the results are usually amalgamated with other TEVAR patients. However, this group of patients may benefit the most from stenting due to increased risks of open surgery, and the long-term effects of treatment are determined by age and type of pathology rather than treatment method.23 Often, these patients have had surgery before or have pathology in anatomically difficult areas such as the distal arch, as in Ortner's description of the cause of recurrent laryngeal nerve paralysis (figure 3). Given that there is a local pathology, which requires sealing with a short stent body, the risks of paraplegia should be low. One also could presume that there is often normal calibre healthy aorta either side of the pathological defect in the wall reducing the risk of stent migration.

Figure 3

This 50-year-old man had an aortic patch repair as a young adult, which has undergone aneurysmal change 25 years after; A, innominate artery; B, left common carotid artery; C, left subclavian artery; and D, false aneurysm at coarctation repair site (left hand image). He underwent carotid–subclavian bypass grafting and thoracic stent graft repair with good results and thrombosis of the aneurysm sac (right hand image).

Although these patients often require extra anatomical bypass grafting, these endografts tend to perform well in the long term.


This group of patients may have the most to gain from thoracic stent grafting. They are critically unwell with life-threatening comorbidity, typically in more than one organ system, with a life-threatening disruption of the aorta. Many do not survive open surgery but stent grafting can significantly reduce the risk of treatment.24 ,25

The development of smaller size, conformable stents has allowed the treatment of many more patients than those in earlier series.

For young patients, there are concerns over long-term sequelae and in terms of follow-up regimes. However, this cohort seems to have done exceptionally well, perhaps due to the fact that it is a localised pathology. TEVAR is now the cornerstone of treatment for this condition in many guidelines.

The future

Extending the applicability of TEVAR to more patients is one of the key areas of development. Developments in branched arch devices and evidence of effectiveness are the key to extending the indications for TEVAR and a focus of interest in industry development teams. Expansion of endovascular treatments to patients with pathology in the ascending aorta is also a developing field. This applies to acute type A dissection or PAU, where a localised tear can be successfully managed with a stenting approach, and to patients with PAU or suture line disruption after surgery. There have been a number of successful reports, but large series are lacking, as many patients are anatomically unsuitable. Short stents for this purpose have been used up to this point, such as the proximal extension devices for the infra-renal segment introduced through the subclavian artery (using devices with short nose cones). A number of ‘off the shelf’ devices are now available for aneurysms including the visceral aortic segment—in elective situation, custom made devices may be ordered but this luxury is not available for the acute situation. However, many unknowns need to be answered before their widespread use. One pertinent issue is that of anatomical ‘tolerance’ for malalignment; we do not know what will occur if the stent is not custom made for each individual patient.26

Further research is required to identify whether a dissection specific stent graft could be of advantage. Certainly, the prospect is an attractive one: to reduce risk of type A dissection, to effectively cover a proximal entry tear with less radial force and better conformability as well as improving remodelling distally with reduced spinal cord risk using uncovered stenting.27 Series of patients are currently small but the evidence of favourable aortic remodelling throughout the aorta is evident on analysis of detailed imaging.

Developments must be directed towards improving the robustness of treatment, perhaps by reducing endoleaks by active fixation (such as the Aptus endoanchor device for TEVAR). In addition, reducing the ‘footprint’ of intervention by using better navigation techniques and intelligent catheter design such as robotics may well be important in reducing the risks (especially of embolic disease).28


We now know a great deal about the risks and benefits of TEVAR even if a direct comparison with surgery has not been made formally. TEVAR has many advantages for acute aortic syndromes. However, there is still much controversy and much to learn. Even though the last 20 years have taught us much and we have seen a massive expansion of techniques and stent graft designs, there are just as many again in research and development pipelines. It is not unreasonable to suggest that TEVAR is here to stay and will continue to improve for many years to come.


We thank T Troeng and M Bjorck for preiliminary data from the Swedvasc registry



  • Contributors This manuscript was developed jointly by the authors, with CB responsible for the figures and JP responsible for the tables. Both authors approved the final draft.

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.