The Frozen Elephant Trunk for the Treatment of Chronic Dissection of the Thoracic Aorta: A Multicenter Experience

Transcription

1 The Frozen Elephant Trunk for the Treatment of Chronic Dissection of the Thoracic Aorta: A Multicenter Experience Davide Pacini, MD,* Konstantinos Tsagakis, MD,* Heinz Jakob, MD Carlos-A. Mestres, MD, Alessandro Armaro, MD, Gabriel Weiss, MD, Martin Grabenwoger, MD, Michael A. Borger, MD, Friedrich W. Mohr, MD, Robert Stuart Bonser, MD, and Roberto Di Bartolomeo, MD Department of Cardiac Surgery, Sant Orsola-Malpighi Hospital, Bologna, Italy; Department of Thoracic and Cardiovascular Surgery, West German Heart Center Essen, Essen, Germany; Department of Cardiovascular Surgery, Hospital Clínico, University of Barcelona, Barcelona, Spain; Department of Cardiovascular Surgery, Hospital Hietzing, Vienna, Austria; Department of Cardiac Surgery, Leipzig Heart Center, Leipzig, Germany; and Department of Cardiothoracic Surgery, University Hospital Birmingham NHS Foundation Trust, Birmingham, United Kingdom ADULT CARDIAC Background. Because of the extensive involvement of the aorta, surgical treatment of its chronic dissection continues to represent a surgical challenge. We conducted a study of a multicenter experience to describe a multicenter experience in the treatment of this complex pathology, using the frozen elephant trunk (FET) technique. Methods. Between January 2005 and May 2010, 240 patients underwent treatment with the FET technique and had their clinical data collected in the International E-vita Open Registry. Ninety of the patients, who were the population in the present study, underwent operations for chronic dissection of the aorta (type A, 77%). The mean age of these 90 patients was years, and 72 (80%) of the patients were male. Sixty-two patients (69%) had undergone a previous aortic operation. All of the procedures in the study were performed with the aid of antegrade selective cerebral perfusion. Results. Total replacement of the aortic arch was done in 84 patients (93%). Cardiopulmonary bypass, myocardial ischemia, cerebral perfusion, and visceral ischemia times were , , 86 24, and minutes, respectively. In-hospital mortality was 12% (11 patients). One patient died from a stroke and 8 patients (9%) died from ischemic spinal cord injury. The false lumen (FL) in the patients aortae was evaluated with computed tomography after operation and during follow up. The rates of complete thrombosis of the FL around the elephant trunk were 69% and 79% at the first and last postoperative examinations, respectively. The rates of 4-year survival and freedom from aortic reoperation were 78% 5% and 96% 3%, respectively. Conclusions. The treatment of chronic aortic dissection (AD) with the FET technique is feasible, with respectable results. The rate of aortic reoperation with the use of this technique appears to be lower than that with a conventional approach to the repair of chronic AD. Ischemic spinal cord injury represents a concerning complication of the FET technique but seems to be unrelated to thrombosis of the FL. (Ann Thorac Surg 2011;92: ) 2011 by The Society of Thoracic Surgeons For related article, see page 1557 Chronic aortic dissection (AD) may affect the ascending aorta and parts of the aortic arch, such as the descending thoracic aorta. Most patients with chronic type A aortic dissection have had previous operations for acute dissection, usually receiving a simple supracoronary aortic replacement during the initial procedure. In Accepted for publication June 8, *Drs Pacini and Tsagakis equally contributed to the paper. Presented at the Forty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 31 Feb 2, Address correspondence to Dr Pacini, Department of Cardiac Surgery, Sant Orsola-Malpighi Hospital, Via Massarenti 9, Bologna, Italy dpacini@hotmail.com. such patients, dilation of the aortic false lumen (FL) distal to the site of graft insertion is not uncommon and requires reoperation in up to 30% of cases [1, 2]. Such patients represent a clinical challenge for the surgeon because of the reoperative nature of the procedure for treating such luminal dilation and the extensive involvement of the aorta. Various approaches with acceptable results have been proposed for the treatment of such patients [3 6]. The aim of the present study was to examine a multicenter experience in the treatment of this complex pathology with the frozen-elephant trunk (FET) technique. Dr Jakob discloses that he has a financial relationship with JOTEC GmbH by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 ADULT CARDIAC 1664 PACINI ET AL Ann Thorac Surg FET FOR CHRONIC THORACIC AORTIC DISSECTION 2011;92: Material and Methods Patients Between January 2005 and May 2010, 240 patients underwent complex repair of the thoracic aorta with the frozenelephant trunk (FET) technique and were enrolled in the International E-vita Open Registry [7, 8]. The selection of these 240 patients was approved by the Institutional Review Board of the University of Essen, Germany, and individual consent for the study was waived. Of the 240 patients chosen for enrollment, 166 patients underwent operations for AD, and 90 of these, who constituted the population for the present study, had operations for chronic AD in 6 European centers (Barcelona, Spain; Bologna, Italy; Essen, Germany; Birmingham, United Kingdom; Hietzing, Vienna, Austria; and Leipzig, Germany). The mean age of these 90 patients was years, and 72 (80%) were male. Sixty-two of the 90 patients (69%) had undergone previous surgical repair of the thoracic aorta, and 5 (5%) had had endovascular aortic repair (3 abdominal and 2 thoracic). Six patients (7%) had Marfan syndrome. The preoperative clinical details of the patients conditions are shown in Table 1. Type A AD was documented in 69 patients (77%) and type B in 21 (23%). Patients with AD of type B underwent FET because of concomitant aneurysm of the aortic arch and the ascending aorta, anatomic contraindications to thoracic endovascular aortic aneurysm repair, or associated cardiac diseases requiring additional procedures (coronary artery bypass grafting [CABG], valve surgery, or both). The extension of the aortic FL was limited to the thoracic aorta in 6 patients (7%), whereas the remaining 93%, or vast majority of the patients, had involvement of the entire thoracoabdominal aorta. The mean diameter of the descending aorta at the level of the pulmonary artery bifurcation was mm (Table 1). Surgical Technique No standard surgical protocol for stent-graft placement and arch replacement was enforced in the study. Therefore, the use of a guide-wire for stent-graft insertion, oversizing of the endoprosthesis (by more than 10%), and the operative technique used for arch replacement were based on the surgeons preferences. All patients underwent median sternotomy. Cardiopulmonary bypass (CPB) was instituted with arterial inflow via the right subclavian artery in 62 patients (69%). Other sites of arterial cannulation were the ascending aorta in 9 patients (10%), previous aortic prostheses in 11 patients (12%), the femoral artery in 2 patients (2%), and other sites in 6 patients (7%). Antegrade selective cerebral perfusion (ASCP) was used in all patients. Cerebrospinal fluid drainage was used in 33 patients (37%). Total replacement of the aortic arch was done in 84 patients (93%) and subtotal replacement of the arch in 6 patients (7%). In 39 patients (43%), supra-aortic vessels were reimplanted separately with the use of an additional branched prosthesis, and in the remaining 51 patients (57%) these vessels were reimplanted with the en bloc technique. In 9 patients (10%) the left subclavian artery was not felt to be salvageable, and was therefore oversewn at its origin. The ascending aorta was replaced in 63 patients (70%). Procedures associated with the FET procedure in the study are shown in Table 2. Mean CPB, myocardial ischemia, ASCP, and visceral ischemia times were , , 86 24, and minutes, respectively (Table 2). E-vita Open Graft The E-vita open stent graft (Jotec, Hechingen, Germany) consists of a 15-cm long, self-expandable, nitinolcovered stent graft with an integrated proximal nonstented vascular prosthesis. Stent-graft sizing was performed according to the dimension of the true aortic lumen. In 15 patients (17%), the stent graft was oversized by more than 10%. The mean stent-graft diameter was 30 4 mm. The stent graft was implanted antegradely Table 1. Patient Characteristics Variable No. (%) Number of patients 90 (100%) Age (years), mean SD Age 70 years 13 (14) Male 72 (80) BMI, mean SD 27 4 Emergency 24 hours 2 (2) Marfan syndrome 6 (7) Aortic valve insufficiency 2 29 (32) Malperfusion 7 (8) CAD 8 (9) EF 60% 41 (46) Previous aortic repair for AAD 62 (69) Previous EVAR Thoracic 2 (2) Abdominal 3 (3) COPD 10 (11) PVD 8 (9) Creatinine 2 mg/dl 8 (9) History of stroke 3 (3) AD characteristics Stanford classification Type A 69 (77) Type B 21 (23) Extension of false lumen Thoracic aorta 6 (7) Thoracic abdominal 84 (93) Descending aorta a (mm), mean SD TL collapse 12 (13) Aortic calcification 26 (29) a Level of the pulmonary bifurcation. AAD acute aortic dissection; AD aortic dissection; BMI body mass index; CAD coronary artery disease; COPD chronic obstructive pulmonary disease; EF ejection fraction; EVAR endovascular aortic repair; PVD peripheral vascular disease; TL true lumen.

3 Ann Thorac Surg PACINI ET AL 2011;92: FET FOR CHRONIC THORACIC AORTIC DISSECTION Table 2. Intraoperative Data Variable No. (%) Arterial cannulation sites Right subclavian artery 62 (69) Ascending aorta 9 (10) Aortic prosthesis 11 (12) Femoral artery 2 (2) Others 6 (7) CPB time (min), mean SD ASCP time (min), mean SD Visceral ischemic time (minutes), mean SD Myocardial ischemic time (minutes), mean SD Arch replacement Total 84 (93) Subtotal 6 (7) Reimplantation of supra-aortic vessels Island 51 (57) Separate 39 (43) Sacrifice of LSA 9 (10) Ascending aorta replacement 63 (70) Aortic valve intervention Repair 4 (4) Isolated valve replacement 9 (10) Bentall procedure 13 (14) CABG 7 (8) MV repair/replacement 3 (3) E-vita open stent graft data E-vita open diameter (mm), mean SD 30 4 Use of guide wire 77 (86) Oversizing 10% 15 (17) Level of proximal stent graft anastomosis Distal to LSA 89 (99) Between LCA and LSA 1 (1) Distal landing zone T6 T7 3 (3) T8 T9 59 (66) T10 or lower 28 (31) Arch replacement with integrated graft 45 (50) ASCP antegrade selective cerebral perfusion; CABG coronary artery bypass grafting; CPB cardiopulmonary bypass; LCA left carotid artery; LSA left subclavian artery; MV mitral valve replacement. into the descending thoracic aorta through the opened aortic arch during circulatory arrest. In 77 patients (86%) the stent-graft was positioned with a stiff guidewire. After deployment, the stent was sutured to the aortic stump and the vascular prosthesis was pulled back into the aortic arch. Stent-graft deployment inside the true lumen of the aorta was successful in all but one case, in which the distal stent ended in the FL. In this patient the guidewire was not used. The proximal anastomosis of the stent graft was done beyond the left subclavian artery in all but one patient, in whom the anastomosis was done between the left subclavian and left common carotid arteries. The distal end of the stent graft was deployed at the level of T6 T7 in 3 patients (3%), at the level of T8 T9 level in 59 patients (66%), and the level of T10 or lower in 28 patients (31%). The stent-graft data are shown in Table 2. Follow-Up During follow-up, the patients were examined by physicians and the behavior of the aortic FL was evaluated with computed tomography (CT) or magnetic resonance imaging (MRI). The follow-up examinations were done before discharge, at 6 and 12 months postoperatively, and annually thereafter. Statistical Analysis Statistical analysis was done with SPSS version 18.0 (SPSS, Chicago, IL). Continuous variables are expressed as mean standard deviation throughout this report, and were compared through use of the unpaired twotailed Student s t test. The paired two-tailed Student s t test was used to compare the diameters of the descending aorta and those of the false and true aortic lumina preoperatively, postoperatively, and during follow-up. Categorical variables are expressed as percentages throughout this report and were compared through use of the 2 test or Fisher s exact test where appropriate. A two-tailed value of p 0.05 was considered to indicate statistical significance. Standard Kaplan-Meier actuarial techniques were used to analyze survival data and freedom from aortic reintervention. Results Early Results The overall in-hospital mortality in the study was 12% (11 patients) (Table 3). The cause of death was cardiac failure in 2 patients, severe bleeding in 2, abdominal aortic rupture in 1, sepsis in 1, cholecystitis in 1, pancreatitis in 1, malperfusion in 1, and multiorgan failure in 2. Univariate analysis revealed a significant association between Table 3. Postoperative Data Variable No. (%) In-hospital mortality 11 (12) Low-output syndrome 7 (8) Intubation 72 hours 28 (31) Rethoracotomy for bleeding 12 (13) Dialysis Permanent 4 (4) Temporary 14 (16) Gastrointestinal complications 4 (4) Stroke 1 (1) TND 6 (7) Spinal cord injury Paraplegia 4 (4) Paraparesis 4 (4) TND transient neurologic dysfunction ADULT CARDIAC

4 ADULT CARDIAC 1666 PACINI ET AL Ann Thorac Surg FET FOR CHRONIC THORACIC AORTIC DISSECTION 2011;92: in-hospital mortality and a serum creatinine concentration above 2 mg/dl, stent-graft placement without a guide wire, an intubation time exceeding 72 hours, and a need for permanent dialysis (Table 4). Twenty-eight patients (31%) required ventilation for more than 72 hours. A total of 12 patients (13%) underwent repeat thoracotomy for bleeding. Acute renal failure requiring dialysis occurred in 18 patients, in 14 (16%) of whom it was temporary and in 4 (4%) of whom it was permanent. Postoperative cerebral complications included stroke in 1 patient (1%) and transient neurologic dysfunction in 6 patients (7%). The overall incidence of spinal-cord injury (SCI) was 9% (8 patients), with the injury being complete (paraplegia) in 4 patients and incomplete (paraparesis) in 4. In one case paraplegia occurred after a stent-graft was implanted in the distal descending aorta via the femoral artery in a single procedure. A Larger preoperative diameter of the false aortic lumen and a ratio of 1:4 of the diameter of the true lumen to that of the complete descending aorta were the only variables associated with SCI in a univariate analysis (Table 5). In 83 patients (92%), a postoperative, predischarge CT examination was done and was used for evaluating the Table 5. Univariate Analysis of Spinal Cord Injury Variable No. of Patients (%) No. of Injuries (%) p Value TL/descending aorta (mm) 0.25 a No 39 (63) 1 (3) Yes 23 (37) 5 (22) Preoperative FL diameter (mm), a mean SD Landing zone T6 T7 No 87 (97) 8 (100) Yes 3 (3) 0 Landing zone T8 T9 No 31 (34) 4 (13) Yes 59 (66) 4 (7) Landing zone T10 No 62 (69) 4 (7) Yes 28 (31) 4 (14) FL thrombosis at stent level b No 25 (30) 3 (12) Yes 58 (70) 5 (9) FL patency at stent level b No 74 (89) 6 (8) Yes 9 (11) 2 (22) Table 4. Univariate Analysis of In-Hospital Mortality Variable No. of Patients (%) No. of Deaths (%) p Value Creatinine 2 mg/dl No 82 (91) 8 (10) Yes 8 (9) 3 (38) Oversizing 10% No 75 (83) 7 (9) Yes 15 (17) 4 (27) Stent graft placement without guide wire No 77 (86) 4 (5) Yes 13 (14) 7 (54) Sacrifice of LSA No 81 (90) 8 (10) Yes 9 (10) 3 (33) Intubation 72 hours No 62 (69) 4 (7) Yes 28 (31) 7 (25) Postoperative dialysis No 72 (80) 6 (8) Yes 18 (20) 5 (28) Permanent postoperative dialysis No 86 (96) 8 (9) Yes 4 (4) 3 (75) Gastrointestinal complication No 86 (96) 8 (9) Yes 4 (4) 3 (75) LSA left subclavian artery. a 62 patients with available preoperative aortic dimensions. b 83 patients with available postoperative findings on computed tomography. FL false lumen; TL true lumen. false aortic lumen. The rate of complete thrombosis of the FL was higher at the level of the stent (70%) than in the distal thoracic aorta (28%) or abdominal aorta (21%) (Table 6). Midterm Results During a mean follow up period of months, there were 7 late deaths (9%). The causes of death were unrelated to the aorta in 4 patients (2 cardiac, 1 pulmo- Table 6. Behavior of Aortic False Lumen as Assessed With Computed Tomography PreDischarge No. (%) Follow-Up No. (%) Stent graft level Complete thrombosis 58/83 (70) 60/65 (92) Partial thrombosis 16/83 (19) 4/65 (6) Patent 9/83 (11) 1/65 (2) Distal descending thoracic aorta Complete thrombosis 23/83 (28) 31/65 (48) Partial thrombosis 26/83 (31) 21/65 (32) Patent 34/83 (41) 13/65 (20) Abdominal aorta Complete thrombosis 10/79 (13) 4/62 (6) Partial thrombosis 3/79 (4) 8/62 (13) Patent 66/79 (83) 50/62 (81)

5 Ann Thorac Surg PACINI ET AL 2011;92: FET FOR CHRONIC THORACIC AORTIC DISSECTION Fig 1. Overall actuarial survival rate in chronic aortic dissection after frozen elephant trunk surgery with the E-vita open-stent graft nary thromboembolism, and 1 pancreatitis) and related to it in 3 patients. The deaths in these 3 cases were from aortic rupture during endovascular repair of an abdominal aortic aneurysm in 1 patient at 207 days after the initial FET procedure; an aortotracheal fistula at 321 days postoperatively in 1 patient; and sudden death from what was strongly suspected to be an aortic rupture at 11 months postoperatively in 1 patient. The actuarial survival rates were 80% 5% and 78 5 at 12 and 48 months postoperatively, respectively (Fig 1). During follow-up, CT images for evaluation of the thoracic and abdominal aorta were available for 65 and 62 patients, respectively. In the latest CT control examination, complete thrombosis of the false aortic lumen at the level of the perigraft was documented in 60 patients (92%), partial thrombosis in 4 patients (6%), and patency in 1 patient (2%). At the distal thoracic aortic level, thrombosis of the FL was documented in 31 patients (48%), a partial thrombosis in an additional 21patients (32%), and patency of the FL in 13 patients (20%). In the abdominal aorta, most of the patients (81%) had patency of the FL (Table 6). Aortic dimensions were available preoperatively for 71 of the 90 patients (79%) in the study, for 62 of 79 patients (discharged from the hospital) (78%) prior to hospital discharge, and for 40 of 72 patients (alive at last follow-up control) (56%) at the last follow-up examination. The overall diameter of the aorta at the level of the pulmonary artery bifurcation, us well us the diameter of the FL, decreased significantly with time after the FET procedure (Fig 2), whereas the diameter of the true aortic lumen increased steadily during follow-up (Table 7). ADULT CARDIAC Fig 2. Multidetector computed tomographic scans of a male patient during 4 years of follow-up after frozen elephant trunk surgery with the E-vita open-stent graft for chronic type A aortic dissection. (A, D) Computed tomographic scans at 6 years after acute dissection of the aorta and proximal aortic repair. In the covered thoracic aorta, a stable thrombosis of the false lumen, combined with shrinkage of the false lumen, was observed after 1 year (B) and 4 years (C). In the abdominal aorta the false lumen remained patent, with a stable aortic diameter after 1 year (E) and after 4 years (F).

6 ADULT CARDIAC 1668 PACINI ET AL Ann Thorac Surg FET FOR CHRONIC THORACIC AORTIC DISSECTION 2011;92: Table 7. Aortic Dimensions at the Level of the Pulmonary Artery Bifurcation Variable Preoperative (71 patients) Postoperative (62 patients) Follow-Up (40 patients) p Value Descending aorta, mm True lumen, mm False lumen, mm Secondary aortic repair of the distal aorta was done in 20 (25%) of the 90 patients in the study, of whom 18 patients underwent endovascular repair (17 on the thoracic aorta and 1 on the abdominal aorta) and 2 underwent open surgery (1 on the thoracic aorta and 1 on the abdominal aorta). The actuarial freedom from secondary endovascular repair of the distal aorta was 78% 5% and 69% 8% at 12 and 48 months, respectively, and the actuarial freedom from secondary open surgical repair was 96% 3% at both 12 months and 48 months (Fig 3). Comment The intermediate and long-term prognoses of patients with chronic AD remain inferior to what has been expected. Medical therapy is the treatment of choice for most patients with stable chronic dissection of the aorta. However, a careful follow-up of the condition of the aorta is an essential component of planning the management of this condition and of minimizing late morbidity and mortality from residual aortic disease. Despite optimal medical treatment, the aortic disease frequently evolves with dilatation of the FL in chronic AD. In such cases, surgical repair is indicated and usually consists of a complex staged treatment, the so-called elephant-trunk technique [3]. This technique facilitates construction of the distal anastomosis used in the technique during the initial operation, and avoids hazardous dissection of the distal aortic arch during the second procedure. However, even if the single stages in which the technique is accomplished provide good results in terms of mortality and morbidity, the overall mortality with the FET technique is disturbingly high, at up to 50% of the patients in whom it is used [9]. The FET technique with a hybrid stent-graft prothesis has been developed as a means of treating extensive disease of the thoracic aorta. The results described thus far with this technique are acceptable, particularly in view of the extent of aortic disease and the comorbidities of the patients in whom the technique has been used [7, 8, 10, 11]. In the current study, the overall in-hospital mortality rate was 12% and was comparable to that with conventional aortic arch repair with or without an elephant trunk procedure. A serum creatinine level above 2 mg/dl was the only preoperative variable associated with early mortality by univariate analysis. Early morbidity in the present study, including stroke, transient neurologic dysfunction, and renal failure, did not exceed that reported for the conventional elephant trunk procedure [12 14]. However, these operations are complex and time-consuming, and therefore require good strategies for myocardial, cerebral, and visceral protection. With regard to cerebral protection, we recommend bilateral ASCP as the best method for ensuring an adequate cerebral blood supply and preserving the metabolism and function of brain tissue [15, 16]. Spinal cord injury was a not infrequent complication in the present study, occurring in 8 patients in total (in 4 of whom it was complete and in then other 4 incomplete). A large preoperative diameter of the false aortic lumen and a ratio of less than 1:4 of the diameter of the true aortic lumen to that of the descending aorta were the factors that we found to be associated with SCI by univariate analysis. Spinal cord injury was not related to thrombosis of the false aortic lumen in the present study. On the contrary, this complication occurred more often in patients with a patent FL than in those without patency during follow-up (22% vs 8%, respectively). Fig 3. Kaplan-Meier curves demonstrate freedom from (A) secondary surgical and (B) endovascular intervention along the downstream aorta after frozen elephant trunk surgery with the E-vita open-stent graft for chronic aortic dissection.

7 Ann Thorac Surg PACINI ET AL 2011;92: FET FOR CHRONIC THORACIC AORTIC DISSECTION It is well known that the subclavian arteries participate in perfusion of the spinal cord, but sacrifice of the left subclavian artery was not associated with either paraplegia or paraparesis. None of the 9 patients in whom the left subclavian artery was sacrificed developed SCI. Moreover, we did not find any correlation between SCI and longer times of CPB, visceral ischemia and ASCP. Flores and colleagues [17] demonstrated that extensive coverage of the descending aorta with the graft region of the FET prosthesis, with excessive sacrifice of intercostal arteries, represents a strong risk factor for SCI in patients undergoing FET surgery. Flores and colleagues suggested that another possible mechanism for SCI could be thromboembolism in the presence of severe atherosclerosis at the distal landing zone of the FET prosthesis, which was associated with SCI in 36% of patients with such atherosclerosis [17]. We failed to find a significant correlation between extensive coverage of the descending aorta and SCI in the present study, although the rate of SCI was slightly higher in patients in whom the distal landing zone of the stent was at T10 or lower (14% vs 6%, respectively). Moreover, none of the 3 patients who had stent coverage at levels up to T8 experienced paraparesis or paraplegia. Cerebrospinal fluid drainage, even in isolated endovascular surgery, has been shown to be an effective means of preventing SCI [18, 19]. In the present study, drainage of CSF was used in one-third of the patients without any obvious effect on the incidence of SCI. The role of stent-grafting in chronic AD remains uncertain. The ability of the stent-graft to induce complete thrombosis of the FL of the dissected aorta remains in question. The progressive thickening and stiffening of the dissected membrane, and the presence of multiple reentry sites distal to the stent itself, may play a role against lumen thrombosis in the FL. The reported rate of complete thrombosis in the FL is significantly higher in patients with acute than in those with chronic AD (70% vs 38%, respectively) [20, 21]. However, markedly different results have been reported in a small group (n 19) of patients with chronic AD of type B treated with FET, in which the rate of complete thrombosis of the FL was 94% [22]. Our results confirmed this higher rate of thrombosis of the FL in patients given the FET prosthesis. Moreover, we demonstrated an increasing rate of thrombosis with time, from 70% immediately postoperatively to 92% during follow-up. The proximal fixation of the stent to the native aorta, thereby avoiding proximal endoleakage and stent migration, is probably the main factor responsible for these improved results as compared with those in isolated endovascular therapy. Thrombosis of the FL of the dissected aorta is well correlated with reverse remodeling of the aorta and with subsequent reduction of aortic diameters and volumes [23, 24]. We showed that the overall aortic diameter at the level of the pulmonary artery bifurcation decreased from mm before FET surgery to mm during follow-up, although this difference was not statistically significant. The reduction was due to a decrease in diameter of the FL from mm preoperatively to mm during follow-up (p 0.05), and to a simultaneous dilation of the true lumen from 15 6mm preoperatively to 27 6 mm during follow-up (p 0.05). Complete thrombosis of the FL in the abdominal aorta can result in visceral ischemia if the abdominal arteries arise from the false lumen. This is a very rare eventuality because there are nearly always re-entry sites in the abdominal aorta or in the iliac arteries. However, careful evaluation of the thoracoabdominal aorta must be done before FET surgery, and we strongly believe that the FET procedure should not be done if re-entry sites are not visualized in the distal descending thoracic or abdominal aorta or both, and the visceral arteries arise from the FL of the dissected aorta. We did not have any case of documented visceral malperfusion from occlusion of the FL in the present study. Moreover, gastrointestinal complications and postoperative dialysis, both strongly associated with in-hospital mortality, had no correlation with thrombosis of the abdominal false lumen. The present multicenter study has several limitations, including a lack of randomization that may have led to a selection bias, as well as the lack of a control group and the use of different operative protocols among the surgeons participating in the study. In conclusion, the treatment of chronic AD with the FET is feasible, with respectable mortality and morbidity. In addition, the procedure results in a low rate of reoperation on the downstream aorta, which is probably lower than the rate with a conventional two-stage surgical approach to the treatment of AD. Moreover, although ischemic SCI represents a not negligible complication of the FET procedure, it seems to be unrelated to thrombosis of the FL. We believe that the current coverage of the descending aorta with the FET stent graft may play a role in SCI, even though we were unable to demonstrate any correlation of the two; however, for this reason the length of the stent-graft will be reduced from 15 to 13 cm. Beyond this, follow-up surveillance is crucial to define long-term effectiveness of the FET technique. References Heinemann M, Laas J, Karck M, et al. Thoracic aortic aneurysms after acute type A aortic dissection: necessity for follow-up. Ann Thorac Surg 1990;49: Fattori R, Bacchi-Reggiani L, Bertaccini P, et al. Evolution of aortic dissection after surgical repair. Am J Cardiol 2000;86: Borst HG, Walterbusch G, Schaps D. Extensive aortic replacement using elephant trunk prosthesis. Thorac Cardiovasc Surg 1983;31: Kouchoukos NT, Masetti P, Mauney MC, et al. One-stage repair of extensive chronic aortic dissection using the archfirst technique and bilateral anterior thoracotomy. Ann Thorac Surg 2008; 86: Beaver TM, Martin TD. Single-stage transmediastinal replacement of the ascending, arch, and descending thoracic aorta. Ann Thorac Surg 2001;72: Karck M, Chavan A, Hagl C, et al. The frozen elephant trunk technique: a new treatment for thoracic aortic aneurysms. J Thorac Cardiovasc Surg 2003;125: Tsagakis K, Pacini D, Di Bartolomeo R, et al. Multicenter early experience with extended aortic repair in acute aortic ADULT CARDIAC

8 ADULT CARDIAC 1670 PACINI ET AL Ann Thorac Surg FET FOR CHRONIC THORACIC AORTIC DISSECTION 2011;92: dissection: is simultaneous descending stent grafting justified? J Thorac Cardiovasc Surg 2010;140:S Tsagakis K, Pacini D, Di Bartolomeo R, et al. Arch replacement and downstream stent grafting in complex aortic dissection: first results of an international registry. Eur J Cardiothorac Surg 2011; 39: Safi HJ, Miller CC III, Estrera AL, et al. Optimization of aortic arch replacement: two-stage approach. Ann Thorac Surg 2007;83:S Karck M, Chavan A, Khaladj N, et al. The frozen elephant trunk technique for the treatment of extensive thoracic aortic aneurysms: operative results and follow-up. Eur J Cardiothorac Surg 2005;28: Di Bartolomeo R, Pacini D, Savini C, et al. Complex thoracic aortic disease: single-stage procedure with the frozen elephant trunk technique. J Thorac Cardiovasc Surg 2010;140: S LeMaire SA, Carter SA, Coselli JS. The elephant trunk technique for staged repair of complex aneurysms of the entire thoracic aorta. Ann Thorac Surg 2006;81: Svensson LG, Kim KH, Blackstone EH, et al. Elephant trunk procedure: newer indications and uses. Ann Thorac Surg 2004;78: Etz CD, Plestis KA, Kari FA, et al. Staged repair of thoracic and thoracoabdominal aortic aneurysms using the elephant trunk technique: a consecutive series of 215 first stage and 120 complete repairs. Eur J Cardiothorac Surg 2008;34: Pacini D, Di Marco L, Leone A, et al. Cerebral functions and metabolism after antegrade selective cerebral perfusion in aortic arch surgery. Eur J Cardiothorac Surg 2010;37: Griepp RB. Cerebral protection during aortic arch surgery. J Thorac Cardiovasc Surg 2001;121: Flores J, Kunihara T, Shiiya N, et al. Extensive deployment of the stented elephant trunk is associated with an increased risk of spinal cord injury. J Thorac Cardiovasc Surg 2006;131: Safi HJ, Miller CC III. Spinal cord protection in descending thoracic and thoracoabdominal aortic repair. Ann Thorac Surg 1999;67: Adams JD, Angle JF, Matsumoto AH, et al. Endovascular repair of the thoracic aorta in the post-fda approval era. J Thorac Cardiovasc Surg 2009; 137: Shimono T, Kato N, Yasuda F, et al. Transluminal stent-graft placements for the treatments of acute onset and chronic aortic dissections. Circulation 2002;106:I Herold U, Piotrowski J, Baumgart D, et al. Endoluminal stent graft repair for acute and chronic type B aortic dissection and atherosclerotic aneurysm of the thoracic aorta: an interdisciplinary task. Eur J Cardiothorac Surg 2002;22: Sun L, Zhao X, Chang Q, et al. Repair of chronic type B dissection with aortic arch involvement using a stented elephant trunk procedure. Ann Thorac Surg 2010;90: Tsagakis K, Kamler M, Kuehl H, et al. Avoidance of proximal endoleak using a hybrid stent graft in arch replacement and descending aorta stenting. Ann Thorac Surg 2009;88: Gorlitzer M, Weiss G, Meinhart J, et al. Fate of the false lumen after combined surgical and endovascular repair treating Stanford type A aortic dissections. Ann Thorac Surg 2010;89: DISCUSSION DR MEHMET ATES (Izmir, Turkey): Thank you for your nice presentation. What is your criterion for choosing either an island technique or a branch technique in aortic repair? DR PACINI: This is an international registry and there is no standard protocol for surgery. Therefore, every center has its own indication for using the technique. I can speak about our indication in Bologna. We try to reimplant the epiaortic vessels together, as an island, when they are closed. However, sometimes they are very far from each other and a separate graft technique is preferred. On the other hand, I have to say that at the beginning of our experience we used the first generation of the E-vita prosthesis, which was not made with precoated dacron and had a high porosity. At that time we preferred to take out the proximal part of the E-vita prosthesis to avoid bleeding during the operation, and to use a precoated graft for the arch. In these cases the epiaortic vessels were more frequently reimplanted as separate grafts. But since 2 years ago, with the availability of the new E-vita graft, which has zero porosity, we usually reimplant the epiaortic vessels with the en bloc technique to make the operation faster. DR ERIC ROSELLI (Cleveland, OH): I think that your conclusion is reasonable that the FET technique may be comparable to the open two-stage technique for repairing chronic aortic dissection, but there is also the option of endovascular two-stage completion, which may allow you to shorten the interval between the two stages. I worry that in patients receiving the frozen elephant trunk the high rate of spinal cord injury may be related to the long operative time, because the patients get very edematous. You can imagine that their spinal cord may look like the rest of them does postoperatively, and perhaps they would benefit more from staging the repair over a one- to two-week period. Did you use spinal cord drainage in those patients, and did you look at your data to see if it made any difference in outcome? DR PACINI: In the present series, cerebrospinal fluid (CSF) drainage was not used in all patients. We looked at our data but couldn t find any protective effect of CSF drainage on spinal cord injury. DR ROSELLI: Probably the numbers are too small. DR PACINI: Maybe. However, in every case in which we use the frozen elephant trunk technique we now use CSF drainage, and we try to perfuse the left subclavian artery as much as possible to provide greater blood flow to the spinal cord during systemic circulatory arrest.