Introduction

Aortic dissection is common pathology of thoracic aorta estimating 2.9 per 100,000 year incidence [1]. Open surgery requires thoracotomy, lung collapse, aortic cross-clamping and intercostal arteries reimplanta­-tion usually in hypothermia. Moreover temporary multiorgan ischemia and significant blood loss is common [2]. Since 1999, when Dake et al. reported stent-graft deployment as a new alternative for open surgery in thoracic aorta dissection [3], numerous authors shown that results of stent-graft placement are encouraging and satisfactionary [4–7]. However the technique is not free from complications: acute or retrograde type A dissection (6–8%), stroke (3%), paraplegia (2%) [8], access-related complications (3.3%) [9], endoleaks (4%), bowel infarction, limb ischemia, wound infection [8] and collapse of the prosthesis [10]. Nevertheless, complication rate is lower compared to open surgery [2].

Stent-graft collapse has been firstly reported by Mellisano [11]. It is known as exceedingly rare event, not generally observed in most studies [4, 8, 9, 12]. With collected database, Jonker et al. shown that this kind of complication occurs within first month after implantation, is associated with graft oversizing and traumatic aortic injury as an etiology. Collapse was most common in case of young age and placement of the endograft in the aortic arch [10].

The aim of the paper was to present the case of stent-graft collapse and highlight the interval to the adverse event, the management and also the weight of post-operation surveillance.

Case report

A 44-year old male with resistant hypertension and chronic kidney disease (stage 4) was admitted in February 2012 to the emergency ward at the local hospital presenting chest pain and bilateral 50 meters claudication distance since previous year. Computed tomography at admission revealed collapsed proximal part of Zenith stent-graft in descending thoracic aorta and fully patent false lumen of dissection. Entry tear was on the anterior wall of aorta at the level of brachiocephalic trunk and there was no dissection below stent-graft (fig. 1). At admission ankle-brachial index equaled 0.5 on the right side and 0.43 on the left side.

In anamnesis there was Bentall operation in 2002 (due to type A dissection) proceeded with diagnosis of chro­nic aortic dissection in control CT in 2004. Primary entry tear was at the level of brachiocephalic trunk and secondary 8 cm above celiac trunk. Primitively pathology was treated pharmacologically for two years, however in consideration of another hospitalization due to chest pain we decided to perform endovascular repair by implantation of Zenith TX2 (32 x 147 mm) endograft at the level of secondary tear. Post-implantation angiography and CT revealed no atypical placement of the stent-graft (fig. 2). Chest pain resolved and patent false lumen was left for further observation and staged stent-grafting in case of enlargement. However, patient did not undergo CT surveillance and 6 years later the abovementioned collapse occurred.

To treat collapsed prosthesis endovascular management was performed. In consideration of inability of the closure of primary entry tear with standard device, firstly Zenith TX2 stent-graft (28 mm × 127 mm) was deployed on the verge of ostium of left subclavian artery. Afterwards second Zenith TX2 stent-graft (32 mm ×127 mm) was placed connecting the prior with the collapsed, re-expanding latter. Control angiography revealed proper flow through aorta and still patent false lumen. During hospital stay symptoms of claudication resolved completely with improvement of ankle-brachial index to 0.81 on the right and 0.69 on the left side, no re-collapse or neurological complications were encountered, nor thrombosis, kinking or migration of stent-grafts seen. At 3 months after operation control CT shown full thrombosis of false lumen and patient remained asympthomatic for over one year (fig. 3).

Discussion

Since more than decade endovascular treatment is valuable alternative for open surgery in the management of thoracic aorta diseases [3]. With efficacy, less mortality and complications, mini-invasive approach, stent-grafts are widely used in various indications [13, 14]. On the other hand, different complications could occur including endograft collapse.

Collapse of the stent-graft most often occurs in patients with small radius curvature of aorta, steeply angulated, with poor apposition stent-graft to the aortic wall [15–18]. The last was described as ‘bird’s beak’ deformity [19], increasing instability of proximal part. In severely angulated proximal aortic neck, stent-grafts are unable to conform to the inner curvature of the arch and protrudes into the lumen. This lack of device wall apposition exposes the device to repetitive pullout strengths that can lead to stent fractures and/or collapse [20]. Steinbauer et al. [16] first observed lack of sufficient attachment of the device to the small curvature with tight aorta and a short radius, hence the endograft was towering up against the high hemodynamic forces in the area and might have resulted proximal collapse. For the prevention in these difficult circumstances, prophylactic placing balloon-expandable stents have been proposed [21, 22].

Plenty of the studies highlights that excessive oversize of the stent-graft is the risk factor of prosthesis collapse. According to IFU (instruction for use) in aneurysm treatment as well as in the traumatic aortic rupture it should be 10–15% [23].

Sample of collapsed stent-grafts analyzed by Jonker et al. shows that 67% were oversized ≥ 20% and 16% of prostheses were oversized ≥ 40% in the cohort [10]. It has to be admit, that part of excessively oversized stent-grafts were intentionally used. Lack of all possible and adequate dimensions in the stock and emergency of situation forces to proceed as the only available salvage option [18, 20, 24]. However, based on multifactorial etiology, stent-graft collapse should not be linked with oversizing only. Available are several reports of the device failure with absence of excessive oversizing [16, 20].

Collapse of the prosthesis according to the literature occurs during procedure or shortly after deployment, typically in following few days [10, 16, 18, 20–22, 25]. The critical period of highest risk seems to be within first month after implantation [20]. However, numerous of papers report delayed collapse nearly year after implantation or more [5, 24, 26]. That led Lee [21] to assumption that there are two possible periods of the failure ¾ early associated with delivery, deployment and conformation to the local anatomy and late ¾ where wireform fractures and component separations responses for endograft collapse. Similar observation showed Nienaber et al. [2], where mechanical failure of stent-graft was observed on average 19 months. Such anomaly was not observed in our case. However, the most fragile part seems to be proximal part of the prosthesis and collapse of that part was mostly observed [10, 12].

In the Department of General, Vascular and Transplant Surgery of Warsaw Medical University since March 1998 to April 2013, 1638 patients underwent endovascular treatment of aorta. Among these 359 had implanted stent-graft to thoracic aorta for various causes with 114 due to type B aortic dissection. Along these years there was observed only one case of collapsed endograft, primitively treated due to aortic dissection. Atypically, it occurred in descending thoracic aorta and 71 months post deployment. To our knowledge, presented case of stent-graft collapse in 71^st month after implantation is one of the most delayed symptomatic collapse in the literature. Such slowly arising evaluation of symptoms of the sequelae is unique, as well.

After endovascular repair, a strict imaging follow-up is essential to detect endoleaks, dissection progression, development of proximal or distal aorta enlargement, but also to detect abovementioned ‘bird’s beak’ deformity leading to endograft collapse [20, 27]. Stent-graft collapse can be entirely asymptomatic, but rather causes pseudocoarctation syndrome with no femoral pulse, visceral malperfusion and hypertensive crisis [10, 18, 20, 26, 28, 29]. Relatively more severe occlusion was observed, if symptoms were present at admission and was associated with higher, statistically significant 30-day mortality compared to asymptomatic patients [10]. Our patient did not attend for control CT protocol post hospital stay, so it was impossible to detect any radiological abnormalities, even though he suffered from decreasing claudication distance for one year.

The role of Thoracic Endovascular Repair (TEVAR) in chronic type B aortic dissection remains controversial and its mid-term success as an alternative to open repair or best medical therapy remains unknown. In Investigation of Stent-Grafts in Aortic Dissection trial (INSTEAD) statistically significant difference was reached for thrombosis of false lumen in dual therapy (medical treatment combined with stent-graft implantation) with simultaneous shrinkage of the false lumen. On the other hand, TEVAR failed to improve two-year survival and adverse event rates [29]. Entry tear closure seems to be the key for healing process of false lumen [8, 19, 29, 30]. Gaxotte et al. [5] observed that complete false lumen thrombosis is difficult because of the multiplicity of re-entry sites and proximity of the visceral branches. Thus, thoracic stent-graft does not appear to be a definitive treatment of dissection [4]. Indication for TEVAR in patients with type B chronic aortic dissection is its complication including visceral or leg ischemia, aortic rupture, refractory hypertension, and refractory pain, patients with a descending thoracic aorta of 40 mm or larger in diameter at the onset of aortic dissection, patients with a descending thoracic aorta 50 mm or larger in diameter at any time and thoracic aortic aneurysm [31].

In our case suggested entry closure on the level of brachiocephalic trunk was unavailable. It could be only treated with aortic debranching or branched/fenestrated stent graft. However it is much more complicated and we have not seen indication to perform it. The first CT after implantation was calm and patient has become asymptomatic, so we did not decide about next intervention. As the primary entry tear was left open the probable cause of the collapse was enlargement of false canal which constricted proximal part of the prosthesis from out of lumen.

In the treatment of collapsed stent-graft several therapeutic options can be proposed: it can be removed via an open thoracotomy or be re-expanded by an endoluminal approach. The first solution stays in opposite the idea of mini-invasiveness of endovascular repair, however the advantage is impossibility of recurrent collapse [22]. Most reported cases of prosthesis failure have been treated by the implantation of bare stent or by the implantation of another stent-graft within the collapsed one [11]. First time performed by Idu et al. [15], balloon re-expansion, now is obsolete, since fair outcome and recollapse occure even intraprocedurally. Kasirajan and colleagues calculated that the similar success rate was achieved using stent (86%) and performing open surgery (86%) higher than, when ballooning was proceed to repair (50%) [32]. Failure of endovascular salvage of collapsed endografts and late complications related to graft can be effectively and safely managed with open repair [24]. Our management was firstly concentrated to restore flow in the true lumen with first stent-graft and secondary to re-expand collapsed prosthesis.

Conclusions

The cause of stent-graft collapse is not only exceed oversizing but also false lumen dilatation and compression of the endoprosthesis. Therefore postoperative surveillance is crucial to detect first signs of the collapse. The first choice treatment should be endovascular.

References