Reoperations on the aortic root represent a distinctive

Transcription

1 Results of Reoperation on the Aortic Root and the Ascending Aorta Nicola Luciani, MD, Raphael De Geest, MD, Amedeo Anselmi, MD, Franco Glieca, MD, Stefano De Paulis, MD, and Gianfederico Possati, MD Divisions of Cardiac Surgery, and Cardiac Anesthesia, Catholic University, Rome, Italy; and Division of Cardiovascular and Thoracic Surgery, OLV Clinic, Aalst, Belgium Background. Reoperations on the aortic root and the ascending aorta after previous aortic valve and proximal aortic surgery are increasingly frequent and highly demanding. The scarce comparability of the published series and the heterogeneity of clinical pictures contribute to the challenges of this subgroup. Methods. Forty-one patients (2004 to 2010) who were reoperated on the aortic root and the ascending aorta for aneurysmal, pseudoaneurysmal, or infectious disease were retrospectively analyzed from a prospectively filled-in database. Results. Mean logistic European system for cardiac operative risk evaluation was 29.8%. At index reoperation, procedures were classic Bentall (51%), prosthesissparing operation (17%), supracoronary ascending aortic replacement plus aortic valve replacement-repair (22%), and root replacement using valved homografts (9.7%). Distally, the operation involved the arch in 51% of cases (17 hemiarch replacement, 4 total transverse arch, 3 elephant trunk). Operative mortality was 12% and rate of major operative morbidity was 17%. At a mean 26- months follow-up, the patients surviving the operation had a good survival and functional class. The rate of adverse events during the follow-up was acceptable. Conclusions. Reoperations on the aortic root-ascending aorta in the elective patients have respectable operative mortality-morbidity despite the high-risk profile, and are justified by the excellent follow-up survival. The mortality can be diminished by integrated surgical strategies and optimal myocardial protection. Our findings encourage complete resection of borderline dilated ascending aortic-root tissue at primary and redo operation. (Ann Thorac Surg 2011;92: ) 2011 by The Society of Thoracic Surgeons Accepted for publication April 29, Address correspondence to Dr Luciani, Division of Cardiac Surgery, Catholic University, Largo A. Gemelli, 8, Rome, Italy; nicola.luciani@tiscali.it. Reoperations on the aortic root represent a distinctive and challenging subgroup among cardiac reinterventions. In such settings the operative risk is reportedly higher than in primary aortic root procedures [1, 2]. In experienced hands, surgery on the aortic root and the ascending aorta done after previous cardiac surgical procedure other than aortic can be performed with respectable mortality [3, 4]. Conversely, the reoperations on the aortic root and the ascending aorta after previous surgery on the aortic root, the aortic valve, or the ascending aorta constitute a further subgroup of patients posing specific challenges. We are confronted not only with the risk of reentry injuries, but also with a wide spectrum of anatomic and pathologic scenarios. One problem in the analysis of this subgroup is its heterogeneity in terms of type of previous operation, indications, and concomitant procedures performed at both previous surgeries and at index reoperation. Prosthetic endocarditis is the prevailing indication in some previous studies [5, 6], while degenerative aortic aneurysm and pseudoaneurysm [1, 7], or bioprosthetic valvular failure after previous root replacement prevail in others [8]. Many series included the cases which received a nonaortic operation at primary surgery [9]. The operative mortality varies considerably with the above-mentioned factors. We aimed at describing a homogeneous subset of current (2004 to 2010) reoperations on the aortic root and the ascending aorta after previous proximal aortic and aortic valve surgery. Patients and Methods Patient Selection Our database was retrospectively reviewed to identify the elective operations on the aortic root and the ascending aorta performed at our institution (January 2004 to August 2010). Data are entered prospectively in the database. A total of 216 operations were identified and screened in order to select those complying with the following criteria: Operation on the aortic valve, aortic root, or ascending aorta has been previously performed; patients underwent reoperation on the aortic root and the ascending aorta (index reoperation, the indication to index redo surgery could be either true aneurysm, or false aneurysm, or infectious disease); patients who were reoperated for structural or nonstructural prosthetic valve deterioration are excluded. Patients receiving surgery of the aortic root and the ascending aorta at index reoperation after any other surgical procedure on the heart are excluded by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg LUCIANI ET AL 2011;92: REOPERATIONS ON THE AORTIC ROOT The local Review Board approved the study. Because the data were drawn from hospital records and treated retrospectively and anonymously, and because the present investigation did not entail any additional diagnostic or therapeutic procedure in either way, the need to obtain informed consent was waived. 899 Surgery and Postoperative Management Computed tomography scan of the thorax was preoperatively performed in all patients in order to understand the relationships between the sternum and the aortic structures, the right ventricle, the innominate vein, and bypass grafts. Jugular vein cannulation was performed in all cases and the femoral vessels prepared before resternotomy. In cases considered to be at high risk of reentry lesion, resternotomy was performed after cannulation of the femoral vessels and institution of cardiopulmonary bypass (CPB), in order to decompress the right ventricle and the aortic aneurysm-pseudoaneurysm. This strategy may be used to reduce the risk of reentry lesions and facilitate myocardial protection. The indications, technique, and advantage of this approach are described in detail elsewhere [10]. When on complete CPB, and with the heart beating, the sternum was opened with an oscillating saw and the surrounding adhesions were dissected. The dissection was kept as limited as possible. In the presence of patent left internal thoracic artery-toleft anterior descending artery graft, particular care was paid to the identification of the graft. This was temporarily occluded in all cases at the time of aortic crossclamping. Myocardial protection was accomplished in all cases by of cold crystalloid cardioplegia (Custodiol; Köhler Chemie, Alsbach-Hänlein, Germany), usually administered retrogradely. If required, additional cardioplegia was delivered selectively in the coronary ostia. In some cases only antegrade delivery was performed. For the performance of Bentall operation, the coronary ostia were isolated subtotally in all instances and sutured directly to the tube graft (Fig 1). The Cabrol technique was never used, even in case of markedly dilated and calcified aortic root. The anastomoses were occasionally reinforced with Teflon felts in the presence of weakened aortic tissue. With the aim to minimize the risk of recurrent infection, we used a cryopreserved valved homograft to replace the root and the aortic valve whenever an infectious state existed. Whenever hypothermic circulatory arrest was foreseen to perform open distal anastomosis or arch replacement, the core temperature was cooled to 26 C (rectal temperature) and retrograde cerebral perfusion was used when the expected arrest time did not exceed 30 minutes; the core temperature was cooled to 22 C to 26 C and antegrade selective cerebral perfusion (Kazui technique [11]) was instituted in the case of longer arrest time. Preoperative risk evaluation was performed using the logistic EuroSCORE (European system for cardiac operative risk evaluation) [12]. The occurrence of major-minor postoperative morbidity during the stay in the intensive care unit was recorded. Major complications were defined as stroke, shock, sepsis, myocardial infarction, severe myocardial Fig 1. Schematic drawing: subtotal isolation of the coronary ostia in root replacement. The red brackets indicate the portion of spared aortic tissue around the ostia. dysfunction requiring mechanical circulatory support, and visceral ischemia. Operative mortality was defined as death within the 30th postoperative day. Study endpoints were the following: Operative results (mortality, major, and minor morbidity); midterm follow-up survival; and reoperation-free survival. Follow-Up The follow-up of patients surviving the operation was conducted by outpatient visits 30 days after hospital discharge and every 6 months thereafter. The visit included physical examination, New York Heart Association class, surface electrocardiogram, and transthoracic echocardiography. Telephone interviews were also performed to inquire on the patients status. Statistical Analysis Analysis was conducted with SPSS software, version 11.0 for Windows (SPSS, Chicago, IL). Descriptive statistics were mean standard deviation for continuous data and percentages for categoric data. Survival analysis was performed according to the Kaplan-Meier method, and corresponding curves were built. Results A total of 41 patients fulfilled the inclusion criteria (12.5% of all aortic operations and 3.7% of all cardiac reoperations performed in the study time span). Baseline characteristics and surgical procedures performed before index reoperation are outlined in Table 1. Previous operation was aortic valve replacement or repair in 46% of cases, supracoronary ascending aortic replacement in 39% and Bentall procedure in 15%. Additional procedures at previous operation were the following: CABG in

3 900 LUCIANI ET AL Ann Thorac Surg REOPERATIONS ON THE AORTIC ROOT 2011;92: Table 1. Baseline Characteristics of the Patients Characteristics Values Age (years) (range) (39 80) Sex, male 30 (73%) Diabetes mellitus (n, %) 12 (29.3%) COPD (n, %) 9 (22%) Poor systolic function (LVEF 0.30) 12 (29.3%) (n,%) Extracardiac arteriopathy (n, %) 18 (43.9%) Renal failure (n, %) 6 (14.6%) Pulmonary hypertension ( 60 mm Hg) 11 (26.8%) (n, %) Number of previous operations (mean, 1.4 (1 3) range) Coronary artery disease (n, %) 4 (9.7%) Logistic EuroSCORE 29.8% Surgical procedure performed before index reoperation: Ascending aortic replacement 16 (39%) Bentall 6 (15%) Aortic valve replacement/repair 19 (46%) Indication to index reoperation: Aneurysm 29 (71%) Pseudoaneurysm 9 (22%) Infection 3 (7%) COPD chronic obstructive pulmonary disease; EuroSCORE European system for cardiac operative risk evaluation; LVEF left ventricular ejection fraction. two cases, mitral valve repair/replacement in three, hemiarch replacement in one and aortic valve replacement in 4 patients subjected to supracoronary ascending aortic replacement. Four patients underwent the index reoperation after more than one cardiac surgery before. The indication to the index reoperation was true aneurysmal disease in 71% of cases, pseudoaneurysm in 22% and infection in 7% (Table 1). Chronic type A aortic dissection coexisted in 14 cases. Operative Results Classic Bentall operation was most frequently performed at index reoperation (Table 2). A valved conduit with a mechanical prosthesis was employed in 17 cases, a biological valved heterograft (Shelhigh Inc, Union, NJ) was used in 3 cases, and a biological valve mounted on a tube graft in 1 case. In 4 cases, a cryopreserved valved homograft was used for root replacement. In 7 cases severe dilatation of the aortic root coexisted with a wellfunctioning mechanical aortic prosthesis implanted previously, and a prosthesis-sparing operation could be performed [13]. A tube graft was sutured to the sewing ring of the valvular prosthesis, which was left in place; reimplantation of the coronary ostia was then performed as usual. In 9 cases we performed supracoronary aortic replacement; among these, concomitant aortic valve replacement was performed in 7 cases while the aortic valve was repaired in 2 cases. Concomitant procedures performed at index reoperation were the following: CABG in 7 cases, mitral or tricuspid valve repair in 4, Morrow s myectomy in 1, total arch replacement without (4), or with, elephant trunk technique (3), closure of patent foramen ovale in 1, reconstruction of the mitroaortic junction in one. Cardiopulmonary bypass and myocardial ischemia times are considerable (Table 2). Hypothermic circulatory arrest was required in 21 cases (51%), including 9 cases in which antegrade selective cerebral perfusion was instituted (Kazui technique). Deep hypothermic circulatory arrest was necessary to perform surgery on the aortic arch in all but 2 cases, in which circulatory arrest was unavoidable to control bleeding after pseudoaneurysm rupture at resternotomy. Circulatory arrest was 6 minutes and 9 minutes long in these cases. During the study period, we observed 1.3% mortality after primary isolated aortic valve replacement and 1.9% after elective primary Bentall operation. The operative morbidity events are summarized in Table 3. Operative mortality was 12% (5 cases). Death was cardiac in origin in 4 cases (postoperative pump failure unresponsive to mechanical circulatory support). In 2 of these cases, ventricular fibrillation occurred before we could control and cross-clamp the ascending aorta, and in presence of a regurgitant aortic valve; left ventricular distension ensued. In the other 2 cases, inadequate myocardial protection was most likely at the origin of post-pump failure. In 1 of these, a patent left internal mammary artery-to-left anterior descending artery graft was injured during mediastinal reentry. Myocardial ischemia ensued until administration of cardioplegia; unplanned CABG was done at the end of the operation. The second patient died because of right heart failure due to inadequate perfusion with cardioplegia of the right coronary artery. In this case, a pseudoaneurysm which was not initially noticed, determined compression of the proximal right coronary artery and inadequate antegrade cardioplegia flow. The last case of operative mortality was due to the late rupture of a descending aortic aneurysm in a patient subjected to prosthesis-sparing operation and arch replacement. Among the deceased Table 2. Operative Results (Index Reoperations) Characteristics Values CPB time (minutes) Myocardial ischemia time (minutes) Hypothermic circulatory arrest (n, %) 21 (51%) Arrest time (minutes) Antegrade selective cerebral perfusion (n, %) 9 (22%) Operative mortality (n, %) 5 (12%) Procedure performed at index reoperation Root replacement (Bentall) 21 (51.2%) Prosthesis-sparing 7 (17.1%) Root replacement (homograft) 4 (9.7%) Ascending aortic replacement AVR/AVP 9 (21.9%) AVP aortic valve plasty; AVR aortic valve replacement; CPB cardiopulmonary bypass.

4 Ann Thorac Surg LUCIANI ET AL 2011;92: REOPERATIONS ON THE AORTIC ROOT Table 3. Perioperative Complications Characteristics Values Major complications Stroke/TIA 2 (20%) Sepsis 1 (10%) Myocardial ischemia 1 (10%) Pacemaker implantation 2 (20%) Post-pump myocardial failure 4 (40%) Minor complications Renal insufficiency 5 (45%) Respiratory failure 2 (18%) Inotropic support 24 hours 4 (37%) TIA transient ischemic attack. patients, previous operation was aortic valve replacement in 3, ascending aortic replacement in 1, and Bentall in 1. Among the same patients, 3 received index reoperation for aneurysmatic disease and 2 for pseudoaneurysm. The mortality in these subgroups was 10.3% and 22%, respectively. The mean intensive care unit stay was days. In 2 cases of severe post-pump failure, mechanical cardiopulmonary support was instituted by extracorporeal membrane oxygenation. We recorded 1 case of stroke (monoplegia), and 1 case of transient ischemic attack. Minor complications occurred in a total of 11 cases. Two patients (4.9%) required the implantation of a pacemaker for complete atrioventricular block. Follow-Up Results Mean follow-up time was months (range, 2 79 months); follow-up was 100% complete. One patient died suddenly during the second postoperative month. At the end of the follow-up all the remaining patients were alive and in good general conditions. The average New York Heart Association class among the follow-up survivors was ; three patients were in class III and the remainders were in class II or I. One patient who received redo-bentall with homograft at index reoperation was reoperated again for development of severe aortic regurgitation. During the follow-up, 2 patients required implantation of pacemaker for complete atrioventricular block and 1 patient had pneumonia, which was resolved completely. There were no further instances of adverse events. Kaplan- Meier survival curves for overall survival, major adverse event-free survival, and minor adverse eventfree survival are reported in Figure 2. Comment 901 Reoperations on the aortic root and the ascending aorta are a peculiar subgroup among cardiac reinterventions, and are believed to carry higher risk due to surgical complexity and variability in clinical and anatomic features [14]. Previous series indicate that these operations are clinically justified by respectable mortality ranging between 6.5% and 18% [6, 7]. Indications for redo procedures on the aortic root can be classified into 3 major categories: aortic, infective, and valvular indications (namely dysfunction of prosthetic valve or repaired native valve). The experiences published so far are difficult to compare because the patients groups are often inhomogeneous with respect to the type of indication to index reoperation. Subgroup analysis can be helpful to characterize better the surgical results. We analyzed a series of patients undergoing elective reoperation on the aortic root and the ascending aorta for either true aneurysmal, pseudoaneurysmal aortic disease or infectious state. We restricted the collection of study cases to recent years. In contrast with previous reports, we enrolled only patients with a history of surgery on the aortic valve and the proximal aorta rather than of any other cardiac procedure. Operative mortality was 12%, which is similar to the previously published rates. Such finding should, in any case, be appraised in the light of the complex profile of these patients, which is underlined by the 51% rate of hypothermic circulatory arrest and 22% rate of antegrade selective cerebral perfusion. These rates are considerably Fig 2. (A) Kaplan-Meier survival overall curve for the entire study population. (B) Kaplan-Meier survival curve for the entire study population: major adverse event-free survival. (C) Kaplan-Meier survival curve for the entire study population: any adverse event-free survival (major and minor adverse events) among patients who survived the operation.

5 902 LUCIANI ET AL Ann Thorac Surg REOPERATIONS ON THE AORTIC ROOT 2011;92: higher than those observed in the cited references. Some authors report that hypothermic circulatory arrest and aortic arch surgery are associated with increased operative mortality [15, 16], although this is a debated issue. In our series, the frequent need for concomitant procedures has contributed to increase the CPB and myocardial ischemia times. The observed mortality is lower than predicted by the logistic EuroSCORE, which should be cautiously interpreted while evaluating the indication in these patients. Early mortality after reoperation on the aortic root and the ascending aorta is higher than mortality of primary aortic root replacement operations [2], which in the recent literature range between 2% and 3% [17, 18]. Inour series, redo surgery on the aortic root carries increased operative risk compared with reoperation on the aortic valve alone. The involvement of the sinuses of Valsalva and of the coronaries in the surgical procedure should be considered to raise the operative risk by itself [1, 19]. Myocardial failure was the major cause of death in the aortic root reoperation group, thus underlining the pivotal role of myocardial protection in these cases. To guarantee the safety of the procedure, the surgeon must pay the highest attention to the following: (1) perfect delivery of cardioplegia; and (2) perfect reattachment of the coronary ostia to the material used to replace the root, because in redo procedures the avoidance of torsion, kinking, or inadequate mobilization of the coronary arteries is technically more demanding than in primary operations. Routinely, we verify the patency and the flow in the proximal portion of both coronary arteries using the transesophageal echo probe at the end of the procedure. Unexpected CABG is associated with operative mortality due to myocardial failure in redo operations [7]. Indeed, the need for unplanned CABG for myocardial ischemia after release of cross-clamp is likely the expression of suboptimal reimplantation of the coronaries. The retrograde delivery of cardioplegia has been proposed to be superior to the antegrade one [20], as a consequence of involvement of the coronary ostia in the disease process affecting the aortic root (pseudoaneurysm, dissection, endocarditis). We have uniformly used cold crystalloid cardioplegia administered retrogradely, plus additional administration selectively, in the coronary ostia in some cases. The advantages of cold crystalloid cardioplegia are predominantly evident in these cases requiring prolonged CPB times, because 1 dose delivered in a few minutes provides stable protection without the need for repeated administrations. The occurrence of left ventricular distension during the reentry phase is an ominous herald of severe myocardial dysfunction at the end of the operation. Thus, caution should be taken in patients with an important regurgitant aortic valve; judicious cooling of core temperature to avoid premature fibrillation and expeditious control of the ascending aorta. Whenever the ascending aorta cannot be controlled and cross-clamped in a short time, transapical venting through a small left thoracotomy or positioning an endoclamp balloon within the ascending aorta should be promptly considered. The anatomic proximity of conduction tissue (His bundle) is the reason for the incidence of complete atrioventricular block after reoperations on the root, which may reach 15% [5]. We have observed a low requirement rate of permanent pacemaker implantation (4.9%); in this perspective, the prosthesis-sparing technique [13] is useful to reduce the trauma of the root at reoperation. We believe that the jugular-femoral cannulation provides an excellent working space and visibility of the surgical field. In order to minimize the risk of malperfusion in cases with chronic aortic dissection, we generally introduce a longer cannula (17 to 21 Fr) through the femoral artery (Seldinger technique) and drive its tip into the true lumen up to the level of the upper descending aorta or distal aortic arch (transesophageal echo guidance). In this way, an antegrade flow is provided using a distal cannulation site. Whenever required, we prefer to employ the Kazui technique rather than to perfuse the brain from an axillary artery cannula, as opposed to other groups [21]. Establishing CPB before resternotomy in patients at high risk is a helpful strategy to reduce the risk of reentry injuries at resternotomy and to repair the injuries whenever they occur [10]. Our observations suggest that one should be as complete as possible in the resection of diseased aortic wall, even if it is affected by borderline dilatation at primary surgery, and especially in younger patients affected by bicuspid aortic valve disease. One should limit as much as possible the extent of diseased aortic tissue that remains exposed to the blood pressure (as in coronary button reimplantation), in order to minimize the risk of recurrent aneurysm or pseudoaneurysm. We have uniformly been aggressive also in the resection of diseased aortic tissue at reintervention, as it is underlined by the high rate of use of hypothermic circulatory arrest and antegrade selective cerebral perfusion. Nonetheless, patients surviving the operation display an excellent longterm survival, not different from that of individuals who received redo isolated aortic valve surgery, and were free from any aortic disease. On this basis, reoperations of the aortic root-ascending aorta in the elective patients should be considered curative and are justified in the light of the expected survival benefit. Mortality and morbidity of these complex cases can be effectively diminished by the integrated use of different strategies, including thorough preoperative evaluation (angio-computed tomographic scan of the thorax), appropriate use of CPB sometimes started before resternotomy, optimal myocardial protection strategy, and pharmacologic adjuvant to reduce bleeding. In this series we had 5 patients aged 75 years or more, with 1 case of operative mortality. In our experience, and in contrast with other reports [3], advanced age does not represent an absolute contraindication to reoperations on the proximal aorta, although our sample size is insufficient to perform a formal risk factors analysis. In case of chronic aortic dissection, a long arterial cannula placed in the true lumen of the distal aortic arch is a helpful tool to reduce the risk of intraoperative malperfusion.

6 Ann Thorac Surg LUCIANI ET AL 2011;92: REOPERATIONS ON THE AORTIC ROOT References 1. Etz CD, Plestis KA, Homann TM, et al. Reoperative aortic root and transverse arch procedures: a comparison with contemporaneous primary operations. J Thorac Cardiovasc Surg 2008;136: Silva J, Maroto L, Carnero M, et al. Ascending aorta and aortic root reoperations: are outcomes worse than first time surgery? Ann Thorac Surg 2010;90: Girardi LN, Krieger KH, Mack CA, Lee LY, Tortolani AJ, Isom OW. Reoperations on the ascending aorta and the aortic root in patients with previous cardiac surgery. Ann Thorac Surg 2006;82: David TE, Feindel CM, Ivanov J, Armstrong S. Aortic root replacement in patients with previous heart surgery. J Card Surg 2004;19: Szeto WY, Bavaria JE, Bowen FW, et al. Reoperative aortic root replacement in patients with previous aortic surgery. Ann Thorac Surg 2007;84: Schepens M, Dossche KM, Morshuis WJ. Reoperations on the ascending aorta and aortic root: pitfalls and results in 134 patients. Ann Thorac Surg 1999; Kirsch M, Radu NC, Mekontso-Dessap A, Hillion ML, Loisance D. Aortic root replacement after previous surgical intervention on the aortic valve, aortic root, or ascending aorta. J Thorac Cardiovasc Surg 2006;131: Raanani E, David TE, Dellgren G, Armstrong S, Ivanov J, Feindel CM. Redo aortic root replacement: experience with 31 patients. Ann Thorac Surg 2001;71: Dougenis D, Daily BB, Kouchoukos NT. Reoperations on the aortic root and ascending aorta. Ann Thorac Surg 1997;64: Luciani N, Anselmi A, De Geest R, Martinelli L, Perisano M, Possati G. Extracorporeal circulation by peripheral cannulation before redo sternotomy: indications and results. J Thorac Cardiovasc Surg 2008;136: Kazui T, Yamashita K, Washiyama N, et al. Aortic arch replacement using selective cerebral perfusion. Ann Thorac Surg 2007;83:S Roques F, Michel P, Goldstone AR, Nashef SA. The logistic EuroSCORE. Eur Heart J 2003;24: Luciani N, Anselmi A, de Geest R, Glieca F, Possati G. Facilitated aortic root substitution: technique and results of the prosthesis-sparing operation. J Thorac Cardiovasc Surg 2010;139: Crawford ES, Crawford JL, Safi HJ, Coselli JS. Redo operations for recurrent aneurysmal disease of the ascending aorta and transverse aortic arch. Ann Thorac Surg 1985;40: Hagl C, Ergin MA, Galla JD, et al. Neurologic outcome after ascending aorta-aortic arch operations: effect of brain protection technique in high-risk patients. J Thorac Cardiovasc Surg 2001;121: Apostolakis E, Akinosoglou K. The methodologies of hypothermic circulatory arrest and of antegrade and retrograde cerebral perfusion for aortic arch surgery. Ann Thorac Cardiovasc Surg 2008;14: Etz CD, Homann TM, Silovitz D, et al. Long-term survival after the Bentall procedure in 206 patients with bicuspid aortic valve. Ann Thorac Surg 2007;84: Kindo M, Billaud P, Gerelli S, Levy F, Mazzucotelli JP, Eisenmann B. Twenty-seven-year experience with composite valve graft replacement of the aortic root. J Heart Valve Dis 2007;16: Dossche KM, Tan ME, Schepens MA, Morshuis WJ, de la Rivière AB. Twenty-four year experience with reoperations after ascending aortic or aortic root reoplacement. Eur J Cardiothorac Surg 1999;16: Byrne JG, Karavas AN, Leacche M, et al. Impact of concomitant coronary artery bypass grafting on hospital survival after aortic root replacement. Ann Thorac Surg 2005;79: Strauch JT, Spielvogel D, Lauten A, et al. Axillary artery cannulation: routine use in ascending aorta and aortic arch replacement. Ann Thorac Surg 2004;78:103 8.