European Journal of Cardio-Thoracic Surgery Advance Access published February 10, 2015

ORIGINAL ARTICLE

European Journal of Cardio-Thoracic Surgery (2015) 1–8 doi:10.1093/ejcts/ezv044

Evaluation of the downstream aorta after frozen elephant trunk repair for aortic dissections in terms of diameter and false lumen status† Gabriel Weissa,*, David Santera, Julia Dumfarthb, Harald Pisarika, Marie Luise Harrera, Sandra Folkmanna, Markus Macha, Reinhard Moidla and Martin Grabenwogera a b

Department of Cardiovascular Surgery, Hospital Hietzing, Vienna, Austria Department of Cardiac Surgery, University Hospital Innsbruck, Innsbruck, Austria

* Corresponding author. Department of Cardiovascular Surgery, Hospital Hietzing, Wolkersbergenstrasse 1, 1130 Vienna, Austria. Tel: +43-699-11996622; fax: +43-1-801102729; e-mail: [email protected] (G. Weiss). Received 14 September 2014; received in revised form 28 December 2014; accepted 12 January 2015

Abstract OBJECTIVES: To analyse the clinical outcomes of surgical repair of DeBakey type I and III aortic dissection (AD) by using the frozen elephant trunk (FET) technique, and to evaluate the postoperative behaviour of the residual aorta. METHODS: In total, 27 consecutive patients underwent treatment of the thoracic aorta for AD with the FET technique in a tertiary-care hospital in Vienna/Austria between 2005 and 2012, and were enrolled in this case series study. All operations were performed under circulatory arrest and bilateral antegrade cerebral perfusion. During the follow-up, a clinical examination was performed as well as aortic diameters and false lumen (FL) patency evaluated by computed tomography (CT) imaging at following levels: pulmonary bifurcation, diaphragm and coeliac trunk. RESULTS: The mean age of the patient cohort was 56 ± 12 years; 21 patients were male. Twenty-two (82%) and 5 (18%) patients presented with DeBakey type I and type III AD, respectively. The hospital mortality rate was 7% (2/28); 2 patients died due to non-aortic-related reasons during a follow-up period of 48 ± 26 months. Three (11%) patients had a stroke, and 2 (7%) a spinal cord injury. The follow-up CT scans revealed FL thrombosis in 96% of the patients at the level of the pulmonary bifurcation (P < 0.001). Distal to the stent graft, at the level of the diaphragm and coeliac trunk, FL patency was observed in 52% (P = 0.1) and 78% (P = 0.6) of the patients, respectively. The true lumen of all analysed aortic segments increased significantly while the mean aortic diameter remained stable. CONCLUSIONS: Compared with conventional surgery for extensive ADs, the FET technique provides a high rate of FL thrombosis of the thoracic aorta. Keywords: Aorta • Aortic dissection • Frozen elephant trunk technique

INTRODUCTION Due to improvements in surgical techniques and perioperative care, the outcome of surgical repair for aortic dissection (AD) has improved substantially over the last decades. However, most survivors who have undergone conventional surgery have a residual dissected aorta, often with a patent false lumen (FL) in up to 90% [1–3]. Residual dissection in the descending aorta substantially influences the long-term prognosis of these patients. Fattouch et al. [1] found that a patent FL associates with significantly lower freedom from reoperation and survival rates. Another series evaluating the effect of FL patency after conventional surgery also observed that patients with a persistent FL had an increased risk of late aortic growth [4–6]. Therefore, an alternative operative strategy that promotes the obliteration of the FL without increasing †

Presented at the 28th Annual Meeting of the European Association for CardioThoracic Surgery, Milan, Italy, 11–15 October 2014.

the operative risk is needed. Feasible and safe therapeutic options may be the new hybrid procedures that combine open antegrade stent grafting with conventional aortic arch surgery; this approach may be particularly suitable for younger patients. One promising approach is the frozen elephant trunk (FET) procedure. The aim of the present case series study was to analyse the clinical outcomes of surgical repair of DeBakey type I and III AD via the FET technique, and to assess the behaviour of the FL and the diameter in the residual aorta after surgery.

MATERIALS AND METHODS Patients Between October 2005 and December 2013, 61 patients underwent thoracic aorta treatment for extensive aortic pathologies with the FET technique. In all cases, the E-vita open stent-graft

© The Author 2015. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

AORTIC SURGERY

Cite this article as: Weiss G, Santer D, Dumfarth J, Pisarik H, Harrer ML, Folkmann S et al. Evaluation of the downstream aorta after frozen elephant trunk repair for aortic dissections in terms of diameter and false lumen status. Eur J Cardiothorac Surg 2015; doi:10.1093/ejcts/ezv044.

2

G. Weiss et al. / European Journal of Cardio-Thoracic Surgery

Table 1:

Patient characteristics of the entire cohort

n (%)

Overall, n = 27

Acute AD, n = 23 (85)

Chronic AD n = 4, (15)

Age Male Female DeBakey type I DeBakey type III Emergency < 24 h Arterial hypertension Diabetes mellitus COPD Creatinine > 2 mg/dl Marfan Previous cardiovascular surgery Previous EVAR AV regurgitation Malperfusion Neurological symptoms preoperatively

56 ± 12 21 (78) 6 (22) 22 (81) 5 (22) 17 (63) 17 (63) 6 (22) 2 (7) 3 (11) 1 (4) 2 (7) 3 (11) 12 (45) 5 (19) 2 (7)

58 ± 11 19 (83) 4 (17) 21 (91) 2 (9) 17 (74) 13 (57) 5 (22) 2 (9) 3 (13) 0 (0) 0 (0) 3 (13) 11 (48) 5 (22) 2 (9)

47 ± 14 2 (50) 2 (50) 2 (50) 2 (50) 0 (0) 4 (100) 1 (25) 0 (0) 0 (0) 1 (25) 2 (50) 0 (0) 1 (25) 0 (0) 0 (0)

AD: aortic dissection; AV: aortic valve; COPD: chronic obstructive pulmonary disease; EF: ejection fraction; EVAR: endovascular aortic repair. Categorical values are n (%).

Figure 1: Computed tomography scan of a complicated DeBakey type III aortic dissection, which was not considered suitable for thoracic endovascular aortic repair (TEVAR) due to an acute angled aortic arch.

hybrid prosthesis ( Jotec GmbH, Hechingen, Germany) was used [7–9]. Of these 61 patients, 27 underwent surgery for an AD with a dissection membrane that extended into the descending aorta diagnosed by computed tomography (CT) scan. The DeBakey classification for AD was used to provide the extension of FL, which was evaluated by at least one aortic imaging modality. These patients were included in the present study. The indication for an FET repair was a DeBakey type I or type III dissection. In the case of DeBakey type III dissection, the primary indication for FET repair was a complicated course combined with an aortic morphology that was not considered suitable for endovascular therapy (a short landing zone in the aortic arch, an acute angled aortic arch, an ascending aorta of >40 mm or tortuous/calcified access vessels) (Fig. 1). All operations were performed under circulatory arrest in moderate hypothermia and bilateral antegrade cerebral perfusion. Informed consent was obtained retrospectively from all surviving patients. The diagnosis of AD was confirmed by a CT scan in all cases. The decision as to which surgical technique should be used was left to the treating surgeon.

Clinical details and patient characteristics The mean age of the study population was 56 ± 12 years, and 21 patients were male (78%). Four (15%) had a chronic dissection, defined as a dissection that was surgically treated 14 days after the initial event. Two patients (7%) had a history of previous surgical replacement of the ascending aorta, and in 3 patients (11%), the

descending aorta had been previously treated by endovascular therapy. Marfan syndrome was documented in 1 patient (4%). Twenty-two (82%) and 5 (19%) patients presented with DeBakey type I and III dissections, respectively. The primary entry tear was located in the ascending aorta or aortic arch in 21 patients (78%) and in the descending aorta distal to the left subclavian artery (LSA) artery in 6 patients (22%). One patient (4%) developed a retrograde DeBakey type I dissection after stent-graft implantation for DeBakey type III dissection. In 3 cases (11%), the dissection re-entered the mid-descending aorta. In the remaining patients, the dissection membrane extended to the abdominal or iliac level. The preoperative clinical details and patient characteristics are given in Table 1.

Surgical technique In all patients, surgical access was obtained by a full median sternotomy. Extracorporeal circulation was established by cannulation of the right axillary artery (n = 26) or the right femoral artery (n = 1). In all cases, antegrade selective cerebral perfusion (ASCP) was performed. After establishing the cardiopulmonary bypass (CPB), the patient core body temperature was reduced to moderate hypothermia (25–28°C) [10]. Once the target temperature had been reached, the CPB was disconnected and the aortic arch was incised. Bilateral ASCP with a total flow rate of 10 ml/kg body weight was achieved through the arterial cannula by clamping the brachiocephalic trunk directly distal its offspring of the aortic arch and also by inserting a perfusion line in the left carotid artery. The implantation of the stent graft was performed over a guide wire, which was antegradely placed through the opened aortic arch into the true lumen (TL) of the dissected descending aorta. Twenty-five (89%) patients underwent concomitant replacement of the aortic arch and ascending aorta, whereas in 2 patients (7%)

G. Weiss et al. / European Journal of Cardio-Thoracic Surgery

3

AORTIC SURGERY

Table 2: Intraoperative data (n = 57) n (%) Time of surgery (min), mean ± SD CPB time (min), mean ± SD ASCP time (min), mean ± SD Visceral ischaemia time (min), mean ± SD MHCA time (min), mean ± SD Cardiac ischaemia time (min), mean ± SD Arch replacement Total Subtotal None Reimplantation of the supra-aortic vessels Island Separate None Level of proximal stent-graft anastomosis Zone 2 Zone 3 Zone 4 Distal landing zone ≤T5 T6–7 T8–9 ≥T10 Ascending aorta replacement Yes No CABG Aortic valve/root intervention Repair Replacement None E-vita open parameter E-vita open diameter (mm), mean ± SD

255 ± 102 204 ± 46 54 ± 12 58 ± 12 4±2 95 ± 34 10 (37) 15 (56) 2 (7) 22 (82) 3 (11) 2 (7) 1 (4) 24 (89) 2 (7) 2 (7) 10 (37) 11 (41) 4 (15) 23 (85) 4 (15) 0 (0)

Figure 2: The Jotec E-vita Open hybrid prosthesis.

the hybrid graft was implanted into the descending aorta over a partial incision of the aortic arch. After deployment and fixation of the FET prosthesis, the aortotomy was directly closed with a running suture. The mean durations of CPB, myocardial ischaemia, ASCP, visceral ischaemia and moderate hypothermic circulatory arrest were 204 ± 46 min, 95 ± 34 min, 54 ± 12 min, 58 ± 12 min and 4 ± 2 min, respectively. The duration of circulatory arrest was defined as the duration where the patient had no extracorporeal circulation at all. This exclusively includes the time where the brain was not perfused. The intraoperative data and procedures associated with the FET repair are given in Table 2.

3 (11) 5 (18) 19 (70)

Hybrid stent graft 31 ± 4

ASCP: antegrade selective cerebral perfusion; CABG: coronary artery bypass grafting; CPB: cardiopulmonary bypass; MHCA: moderate hypothermic circulatory arrest; TL: true lumen. Continuous values are mean ± SD; categorical values are n (%).

presenting with a DeBakey type III dissection, the FET prosthesis was implanted into the descending aorta without replacing the aortic arch. Eight patients (30%) required an additional aortic valve intervention. The mean stent-graft diameter of the implanted prosthesis was 31 ± 4 mm. The distal end of the stent graft was located proximal to the level of TH 8 in 12 patients (44%) and below TH 8 in 15 patients (56%). The stent-graft data are given in Table 2. Aortic arch replacement (total or subtotal) and reimplantation of the supra-aortic vessels (island or separate) were performed depending on the morphological involvement of the dissected aorta and the surgeon’s preference. Total aortic arch replacement with the conventional FET technique was performed in 10 patients (37%), whereas 15 (56%) received a subtotal or so-called ‘arch-light’ repair [8]. With this technique, the supra-aortic branches are not isolated from the descending aorta, thus preserving a narrow connection between the LSA and the downstream aorta, similar to a hemiarch repair. After deploying the stented part of the FET prosthesis, the vascular graft of the E-vita open is pulled out of the stent graft, and shortened to a rim of
2 cm, which is thereafter sutured to the aortic wall with a continuous polypropylene running suture supported by Teflon felts. The remaining arch and ascending replacement are then completed with a separate vascular prosthesis. In 2 patients (7%), arch replacement was not required. In these cases,

The details of the Jotec E-vita Open hybrid prosthesis has been described elsewhere (Fig. 2) [9]. In brief, the stent graft was implanted antegradely into the descending thoracic aorta through the opened aortic arch during circulatory arrest. In all patients, the stent graft was positioned over a stiff guide wire. The size and length of the stent graft was selected according to the preoperative CT scans and by intraoperative measurement of the true aortic lumen under direct vision.

Follow-up All patients underwent a clinical follow-up examination and aorta evaluation by CT scan or magnetic resonance imaging before discharge, at 6 and 12 months postoperatively, and annually thereafter. The mean follow-up duration of the 25 surviving patients was 48 ± 26 months. The follow-up time was calculated with the date of the most recent CT scan and health check at our department. The aortic diameter and FL patency at the pulmonary bifurcation, diaphragm and coeliac trunk levels were evaluated.

Statistical analyses All continuous data are expressed as mean ± standard deviation or median and range. Patient groups were compared in terms of continuous and categorical variables by two-tailed paired t-test, the χ 2-test (Yates’ continuity corrected) and McNemar’s test (for paired variables), respectively. Survival was calculated using the

G. Weiss et al. / European Journal of Cardio-Thoracic Surgery

4

Kaplan–Meier method, and survival curves were compared using the log-rank test. The threshold of significance was set at P < 0.05. SPSS (Version 21, IBM, Armonk, New York, NY, USA) was used for all statistical analyses.

RESULTS Postoperative outcome Hospital mortality (defined as death in the hospital during recovery from surgery) was 7% (2/27). The cause of death was visceral ischaemia with severe consecutive gastrointestinal bleeding in 1 patient (4%) and hepatic failure after a prolonged stay in the intensive care unit (ICU) in the other patient (4%). Two patients (7%) died during the follow-up period due to non-aortic-related reasons. Univariate analysis revealed haemodynamic instability at admission and a history of renal insufficiency (defined as serum creatinine levels of >2 mg/dl) as significant risk factors for in-hospital mortality. The acute and chronic dissection patients did not differ significantly in terms of survival. The median durations of ICU and in-hospital stay were 4 (range, 1–44 days) and 18 days (range, 9–88 days), respectively. Stroke occurred in 3 (11%) patients and spinal cord injury followed by hemiparesis occurred in 2 (7%) patients. Of these 5 patients, 2 patients (7% of all 27 patients) were discharged with a permanent

Table 3: Postoperative outcome Aortic disease, n (%)

Overall (n = 27)

Acute AD (n = 23)

Chronic AD (n = 4)

In-hospital mortality Intubation time > 72 h ICU stay (days), median (range) Hospital stay (days), median (range) Dialysis permanent Stroke Spinal cord injury Sepsis

2 (7) 6 (22) 4 (1–44) 18 (9–88)

2 (9) 6 (26) 4 (1–44) 15 (10–88)

0 (0) 0 (0) 3 (1–4) 17 (9–31)

2 (7) 3 (11) 2 (7) 2 (7)

2 (9) 2 (9) 2 (7) 2 (7)

0 (0) 1 (25) 0 (0) 0 (0)

AD: aortic dissection; ICU: intensive care unit. Continuous values are median ± SD; categorical values are n (%).

neurological impairment. Preoperative significant risk factors for the occurrence of a neurological event could not be identified in the present study. Acute renal failure occurred in 8 patients (30%), of whom 2 (7%) required permanent dialysis. Postoperative outcome data are given in Table 3. Seven patients (26%) required a secondary endovascular intervention after the FET procedure. Two patients (7%) received a stent-graft extension into the aortic arch due to type la endoleak formation. In 1 patient, the endoleak was immediately excluded on postoperative day 1 by endovascular therapy. The other patient was kept under close surveillance. Finally, 4 years after the primary FET procedure, due to late aortic growth, this patient received a stent-graft extension into the aortic arch by sacrificing the LSA. The remaining 5 patients underwent secondary endovascular treatment at the following locations: the common iliac artery, the renal artery, the coeliac trunk, the LSA and the brachial artery. Since the target location was not located in the aorta and these secondary interventions were not associated with the primary FET repair, we did not include these procedures in the freedom from aortic reintervention rate.

Postoperative aortic imaging In 26 patients, aortic diameter and FL patency at the pulmonary bifurcation, diaphragm and coeliac trunk levels were evaluated by CT scan within 14 days of the primary FET procedure. At the level of the pulmonary bifurcation, 16 (62%) and 7 (27%) patients showed no or only partial FL patency 14 days after surgery, respectively. Compared with the FL patency rate in this level that was seen during the preoperative CT scan, there was a significant increase in both complete and partial thrombosis after surgery (both P < 0.001). Three patients (11%) were discharged with a patent FL in the perigraft space. Distal to the stent graft, significant changes in the FL lumen status were not observed. At the level of the diaphragm, complete and partial FL thrombosis occurred less commonly than at the pulmonary bifurcation: 17 and 22%, respectively. More distally, at the level of the coeliac trunk, the FL patency rate remained at 82% (Table 4). Further follow-up CT scans were available for 25 patients during a mean follow-up period of 48 ± 26 months. With regard to the final follow-up CT scans, the rate of FL thrombosis at the level of the pulmonary bifurcation increased to 96% (P < 0.001). Distally, at the level of the diaphragm and coeliac trunk, the FL remained

Table 4: False lumen status before surgery versus predischarge and the follow-up CT scan Aortic level Pulmonary bifurcation Patent Thrombosed Diaphragm Patent Thrombosed Coeliac trunk Patent Thrombosed

Preoperative, n (%)

Predischarge, n (%)

24 (89) 3 (11)

10 (38) 16 (62)

19 (83) 4 (17)

19 (83) 4 (17)

19 (86) 3 (14)

19 (86) 3 (14)

P-value

Follow-up, n (%)