Reoperations on the thoracic aorta pose a challenge

Ascending Aorta and Aortic Root Reoperations: Are Outcomes Worse Than First Time Surgery? Jacobo Silva, MD, Luis C. Maroto, MD, Manuel Carnero, MD, Isidre Vilacosta, MD, Javier Cobiella, MD, Enrique Villagrán, MD, and José E. Rodríguez, MD Units of Cardiac Surgery and Cardiology, Hospital Clínico San Carlos, Madrid, Spain Background. The aim of this study was to analyze surgery and survival data in the midterm after aortic root and (or) ascending aorta reoperations and compare these results with those obtained after first time surgery. Methods. Over a 6-year period, 365 patients underwent an aortic root and (or) ascending aorta surgery procedure at our center. Mean patient age was 63.1 25.5 years; 27.1% were women. Fifty-eight patients had had prior ascending aorta and (or) aortic valve surgery (group I) and the remaining 307 patients were assigned to an initial surgery group (II). The reoperative procedures were Bentall in 45 (77.6%), ascending aorta and valve replacement in 8 (13.8%), and ascending aorta replacement in 5 (8.6%). Results. The reoperation group showed a worse preoperative risk profile indicated by a higher logistic European system for cardiac operative risk evaluation: group I (26.9) versus group II (9.9) (p < 0.0001). Hospital mortality was 7 of 58 (12.1%) in group I and 21 of 207 (6.8%) in group II (p 0.18; relative risk 1.9 [0.8 to 4.6]). After adjusting for the different variables, reoperation could not be identified as an independent predictor of postoperative morbidity. Survival rates (including inhospital mortality) were lower in group I at one year (77.9 1.11% vs 91.9 0.3%) and at 3 years (75.3 0.11% vs 88.9 0.03% [log-rank p 0.005]). In the multivariate analysis, reoperation (p 0.01; hazard ratio 2.6 [1.2 to 5.3]) was a determining factor for survival once corrected for variables predicting mortality during follow-up. Conclusions. Reoperations on the ascending aorta and aortic root showed acceptable morbidity and mortality. Their midterm survival was lower than for patients not requiring a repeat operation. (Ann Thorac Surg 2010;90:555 60) 2010 by The Society of Thoracic Surgeons Reoperations on the thoracic aorta pose a challenge for the surgeon. Prior studies have revealed a hospital mortality as high as 17% [1, 2]. Also, the ever increasing frequency of operations performed on the ascending aorta and aortic arch means that reoperations are expected to rise and it has been estimated that these may eventually account for 10% of all ascending aorta surgery procedures [1, 2]. Another factor to consider is that the expanding use of biologic conduits (pulmonary autografts, homografts, and xenografts) susceptible to deterioration, and ever-increasing aortic valve sparing procedures susceptible to failure during follow-up, will add to this foreseeable increase in the number of ascending aorta reinterventions [3 5]. Some studies have shown that some reinterventions (aortic and mitral procedures) are not independent predictors of hospital mortality and survival when compared with the primary surgery [6, 7]. Published reports of redo aorta procedures only provide descriptive data or attempt to identify factors capable of predicting surgeryrelated mortality or survival [1, 2]. We have found no prior study that has compared reoperation-adjusted outcomes with those observed after initial surgery. This Accepted for publication March 29, 2010. Address correspondence to Dr Silva, Hospital Clínico San Carlos, C/Profesor Martín Lagos s/n, Madrid 28040, Spain; e-mail: jsilva8252@yahoo.es. study was designed to compare surgery results and midterm survival between ascending aorta reinterventions and first time ascending aorta procedures. Material and Methods Study Design In this retrospective cohort study, we compared inhospital and midterm follow-up data in patients undergoing an ascending aorta and (or) aortic root surgical procedure who had had previous surgery (group I) or were undergoing this type of surgery for the first time (group II). The data examined were prospectively obtained from all patients at our unit undergoing an ascending aorta and (or) aortic root procedure, isolated or combined with surgery on the aortic arch, over a six-year period (January 2004 to December 2009). Posthospital discharge data were obtained by contacting the patients by phone. This retrospective study was approved by the Ethics Committee at our center. Variables Assessed: Definitions of Events The factors included in the analysis as independent variables were those comprising the European system for cardiac operative risk evaluation (EuroSCORE) [8], and clinical, anthropometric, and intraoperative variables known to affect morbidity and mortality according to the results of prior studies [1, 2]. 2010 by The Society of Thoracic Surgeons 0003-4975/$36.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2010.03.092

556 SILVA ET AL Ann Thorac Surg ROOT AND ASCENDING AORTA REOPERATIONS 2010;90:555 60 Abbreviations and Acronyms AAR ascending aorta replacement ARF acute renal failure AVR aortic valve replacement CABG coronary artery bypass grafting CKD chronic kidney disease COPD chronic obstructive pulmonary disease CPB cardiopulmonary bypass Cr creatinine HR hazard ratio IQR interquartile range OR odds ratio RR relative risk PS previous stroke PHT pulmonary hypertension SD standard deviation Two events were recorded: hospital mortality, defined as death within 30 days of surgery or before hospital discharge, and complicated postoperative course, defined as the presence of at least one of the following complications: revision surgery due to bleeding, postoperative stroke, a need for prolonged mechanical ventilation greater than 48 hours, mediastinitis, or acute renal failure requiring hemofiltration or hemodialysis. During follow-up, patient status (alive, expired, or lost) and New York Heart Association (NYHA) functional class were recorded. Study Population From January 2004 to December 2009, 365 patients underwent an ascending aorta and (or) aortic root procedure at our center. The mean age of this study population was 63.2 years: standard deviation (SD) 13.4; 99 of the patients were women (27.1%). Of these patients, 58 (15.9%) had undergone a prior aortic valve and (or) ascending aorta procedure; 50 having had one previous operation and the remaining 8 having been operated on twice. The mean time between the most recent procedure and reoperation was 131.7 months (2 to 377 months). Surgical Technique Surgery was performed under general anesthesia and cardiopulmonary bypass (CPB) during moderate hypothermia (32 C) with proximal cannulation of the aortic arch. When an arch operation was contemplated or possible accident during resternotomy, cannulation was achieved through the right axillary artery and the patient was cooled to 28 C, with selective cerebral perfusion at a flow rate of 10 ml kg 1 min 1 if total circulatory arrest was needed. As substitute for the supracoronary ascending aorta, a Dacron prosthesis was used (Hemashield; Boston Scientific, Boston, MA) and when the aortic valve, aortic root, and ascending aorta needed replacing, we used a mechanical valved tube-graft (Carboart; Sorin Biomédica, Saluggia, Italy) or a biologic valved conduit constructed during surgery using a biologic prosthesis and a Dacron tube. Statistical Analysis All statistical tests were performed using the SPSS package (version 15.0; SPSS Inc, Chicago, IL). Categoric variables are expressed as n (%) and continuous variables as their means and standard deviation, or medians and interquartile ranges (IQR) if they failed to show a normal distribution. A statistic analysis was conducted to compare the distributions of the different variables in groups I and II, and a univariate analysis to determine variables linked to postoperative morbidity. The 2 test (or Fisher s exact test if any expected frequency was less than 5) was used to compare categoric variables, and the Student s t test for independent samples or the Mann-Whitney U test, if the variable did not follow a normal distribution, was used for continuous variables. Risk ratios (RR) or mean differences are provided with their 95% confidence intervals. A model was constructed for morbidity by multivariate analysis with binary logistic regression to determine independent factors able to predict postoperative hospital morbidity. The variables included in this analysis were those found to be associated with morbidity (p 0.2) in the univariate analysis and those showing a different distribution in groups I and II (p 0.1). This last step was designed to adjust the reoperation effect on the event variables for possible confounding factors. Odds ratios are provided with their 95% confidence interval. Due to the limited number of hospital deaths, adding a risk of type II statistical error, survival was analyzed with the Cox proportional hazard method including hospital mortality. Survival curves were constructed using the Kaplan-Meier method, comparing outcome between groups I and II by the log-rank test. Survival data were assessed by the Cox proportional-hazard method to identify independent predictors of mortality including on it those variables with p less than 0.2 in the univariate analysis. The level of significance was set at p less than 0.05 (two-tailed). Results Distributions of Variables in Groups I and II Table 1 shows the different variables (EuroSCORE, clinical variables, anthropometric variables, need for concomitant coronary artery surgery, and CPB and aortic cross-clamp times) recorded for groups I (reoperation) and II (primary surgery). Group I showed a higher incidence of previous stroke (PS) (p 0.0001 RR 8.6 [3.5 to 20.7]), active endocarditis (p 0.0001 RR 26.4 [7.2 to 93.3]), severe pulmonary hypertension (pulmonary systolic pressure 60 mm Hg) (p 0.0001 RR 3.8 [1.8 to 7.7]), and longer times of CPB (p 0.0001 mean difference [MD] 35.2 [19.9 to 50.5]) and aortic cross-clamping (p 0.001 MD 20.5 [10.2 to 30.9]). These differences translated to a significant difference in the logistic EuroSCORE [p 0.0001 MD 16.6 [13.4 to 19.9]).

Ann Thorac Surg SILVA ET AL 2010;90:555 60 ROOT AND ASCENDING AORTA REOPERATIONS 557 Table 1. Preoperative and Intraoperative Variables by Study Group: Group I, Reoperation; Group II, Primary Surgery Variable Group I (n 58) Group II (n 307) p Value RR 95% CI MD a 95% CI Age 63 (SD 11.4) 63.3 (SD 13.7) 0.89 0.3 ( 4.0 3.5) a Gender (female) 19 (32.8%) 80 (26.1%) 0.29 0.7 (0.4 1.3) BSA (m 2 ) 1.76 (SD 0.4) 1.84 (SD 0.3) 0.07 0.1 ( 0.01 0.2) a HTA 30 (51.7%) 167 (54.4%) 0.76 0.9 (0.5 1.7) Diabetes 8 (13.8%) 25 (8.1%) 0.16 1.8 (0.8 4.4) Dyslipemia 17 (34.7%) 106 (40%) 0.05 0.8 (0.4 1.6) Smoker 16 (27.6%) 101 (32.9%) 0.09 1.6 (0.8 2.4) COPD 2 (3.4%) 18 (5.9%) 0.75 0.6 (0.1 2.5) PVD 2 (3.4%) 19 (6.2%) 0.55 0.5 (0.1 2.4) PS 13 (22.4%) 10 (3.3%) 0.0001 8.6 (3.6 20.7) CKD (Cr 2.26 mg/dl) 1 (1.7%) 7 (2.3%) 0.9 0.7 (0.1 6.2) egfr (ml/min/1.73 m 2 ) 69.8 (SD 3.3) 69.7 (SD 1.2) 0.97 0.1 ( 6.1 6.4) a Active endocarditis 12 (20.7%) 3 (1%) 0.0001 26.4 (7.2 93.3) Critical preoperative state 4 (86.9%) 23 (77.5%) 0.99 0.9 (0.3 2.7) Unstable angina 1 (1.7%) 20 (6.5%) 0.22 0.2 (0.03 1.9) LVEF 0.40 8 (0.138) 34 (0.111) 0.51 1.3 (0.6 2.9) MI 90 days 1 (1.7%) 6 (2%) 0.99 0.9 (0.1 7.4) Emergency surgery 4 (6.9%) 52 (16.9%) 0.052 0.4 (0.2-1,01) PHT 60 mm Hg 15 (25.9%) 26 (8.5%) 0.0001 3.8 (1.8 7.7) EuroSCORE (logistic) 26.9 (IQR 24.7) 9.9 (IQR 10.4) 0.0001 16.6 (13.4 19.9) a CPB time (minutes) 128 (IQR 76.2) 92 (IQR 62) 0.0001 35.2 (19.9 50.5) a Aortic cross-clamp time (minutes) 93 (IQR 39.5) 66 (IQR 39.5) 0.001 20.5 (10.2 30.9) a CABG 5 (8.6%) 36 (11.7%) 0.49 0.7 (0.3 1.9) NYHA p 0.90 b I 27 (46.6%) 147 (47.9%) II 15 (25.9%) 94 (30.6%) III 14 (24.1%) 53 (17.3%) IV 2 (3.4%) 16 (4.2%) a MD 95% CI. b Linear tendency test. BSA body surface area; CABG coronary artery bypass grafting; CI confidence interval; CKD chronic kidney disease (creatinine 2.26 mg/dl); COPD chronic obstructive pulmonary disease; CPB cardiopulmonary bypass; egfr estimated glomerular filtration rate; EuroSCORE European system for cardiac operative risk evaluation; HTA hypertension; IQR interquartile range; LVEF left ventricular ejection fraction; MD mean difference; MI myocardial infarction; NYHA New York Heart Association; PHT pulmonary hypertension 60 mm Hg; PS previous stroke; PVD peripheral vascular disease; RR relative risk; SD standard deviation. Surgical Results Among the types of prior surgery, isolated aortic valve replacement (AVR) was the most common recorded in 38 patients (65.5%), followed by ascending aorta replacement (AAR) in 16 (27.6%), AVR AAR in 3 (5.2%), and the Bentall technique in 1 (1.7%). Indications for repeat surgery were the following: ascending aorta and (or) aortic root aneurysm in 38 patients (65.5%); aortic dissection in 3 (5.2%); a suture line pseudoaneurysm in 5 (8.6%); and native or prosthetic endocarditis associated with an ascending aorta and (or) aortic root aneurysm in 12 (20.7%). The most common redo procedure was the Bentall technique, conducted in 45 patients (77.6%); one of these requiring the Cabrol modification [9], followed by AVR AAR in 8 (13.8%) and AAR in 5 (10.3%). In 6 (10, 3%) patients the aortic arch was partly or completely replaced. In the first time intervention group (II), the procedure most often performed was AVR AAR in 114 patients (37.1%), followed by the Bentall in 110 patients (35.8%), AAR in 64 (20.8%), aortic root sparing in 17 (5.5%) and homograft-xenograft in 2 (0.65%). In 47 (15.3%) cases, the aortic arch was partly or completely replaced. Mortality and Morbidity Overall surgery-related mortality in our patient series was 7.7% (28 of 365); 12.1% (7 of 58) in group I and 6.8% (21 of 307) in group II (p 0, 18; RR 1.9 [0.8 to 4.6]). The modes of postoperative death were neurologic (1), multiple organ dysfunction syndrome (3), intraoperative bleeding (1), and cardiogenic shock (2) in group I, and neurologic (4), multiple organ dysfunction syndrome (5), intraoperative bleeding (2), and cardiogenic shock (10) in group II. In addition, 66 patients (18.1%) suffered at least one major complication during the postoperative period; 14 of 58 (24.1%) in group I and 52 of 307 (16.9%) in group II (p 0, 19; RR 1.6 [0.8 to 3.1]). Table 2 shows the rates of mortality and major complications recorded in the two groups. Median intensive care unit stay was 2 (IQR 3)

558 SILVA ET AL Ann Thorac Surg ROOT AND ASCENDING AORTA REOPERATIONS 2010;90:555 60 Table 2. Hospital Mortality and Major Complications in Patients Undergoing Reoperation (Group I) or Primary Surgery (Group II) Group I (n 58) Group II (n 307) p Value RR 95% CI MD a 95% CI Mortality 7/58 (12.1%) 21 (6.8%) 0.18 1.9 (0.8 4.6) Major complications 14 (24.1%) 52 (16.9%) 0.19 1.6 (0.8 3.1) Bleeding 7 (12.1%) 18 (5.9%) 0.93 2.2 (0.9 5.4) PMV ( 48 h) 9 (21.4%) 33 (10.7%) 0.3 1.5 (0.7 3.4) PSp 3 (6.5%) 33 (10.7%) 0.3 1.5 (0.7 3.4) ARF-HD 1 (1.7%) 6 (2.2%) 0.99 0.8 (0.1 6.6) Mediastinitis 1 (1.7%) 6 (2.2%) 0.99 0.8 (0.01 6.6) ICU stay (days) 2 (IQR 3) 1 (IQR 2) 0.005 3.7 ( 1.3 8.7) a Hospital stay (days) 17.9 (20.9) 13.1 (SD 14.8) 0.13 4.8 ( 1.5 11) a a MD. ARF-HD acute renal failure requiring hemodialysis; CI confidence interval; ICU intensive care unit; IQR interquartile range; MD mean difference; PMV prolonged mechanical ventilation; PSp postoperative permanent stroke; RR relative risk; SD standard deviation. days in group I and 1(IQR 2) day in group II (p 0.005 MD 3.7 [ 1.3 to 8.7]). Mean hospital stay was 17.9 (SD 20.9) days in group I and 13.1 (SD 14.8) days in group II (p 0.131 MD 4.8 [ 1.5 to 11]) (see Table 2). As variables related to a complicated postoperative course, the univariate analysis identified the following factors: severe pulmonary hypertension (p 0.048), PS (p 0.003), emergency surgery (p 0.0001), unstable angina (p 0.006), need for arch surgery (p 0.001), a critical preoperative state (p 0.004), and CPB (p 0, 0001) and aortic cross-clamp (p 0.001) times. Multivariate analysis identified as independent predictors of postoperative complications: age (p 0.041, odds ratio 1.03 [1.001 to 1.05]), PS (p 0.037 RR 2.8 [1.06 to 7.6]), emergency surgery 9 (p 0.001, odds ratio 5.3 [2.6 to 10.9]), and CPB time (p 0.002, odds ratio 1.01 [1.003 to 1.014]) (see Table 3). Reoperation did not emerge as a predictor of having postoperative complications once corrected for the variables showing a different distribution in groups I and II and those determining complications (p 0.13). Follow-Up Complete follow-up data were available for 350 of the patients (95.9%), with mean and median follow-up times of 30.1 (SD 18.8) and 31.6 (IQR 34.1) months, respectively. Mean NYHA functional class was 1.4 (SD 0.6). During follow-up, 42 patients died, 13 group I and 29 group II patients, yielding overall estimated survival rates (inhospital mortality included) of 89.8 0.03% at 1 year and 86.5 0.04% at 3 years. Survival rates in group I were 77.9 1.11% at one year and 75.3 0.11% at 3 years, while rates in Fig 1. Kaplan-Meier survival analysis (log-rank) of survival in patients undergoing reoperation (group I) versus primary cardiac surgery (group II). (In-hospital deaths included.)

Ann Thorac Surg SILVA ET AL 2010;90:555 60 ROOT AND ASCENDING AORTA REOPERATIONS 559 Table 3. Adjusted Multivariate Analysis for Major Complications Variable p Value OR (95% CI) Age (years) 0.041 1.03 (1.01 1.05) PS 0.037 2.8 (1.06 7.6) Emergency surgery 0.0001 5.3 (2.6 10.9) CPB time (minutes) 0.002 1.01 (1.003 1.014) CI confidence interval; CPB cardiopulmonary bypass; OR odds ratio; PS previous stroke. group II were 91.9 0.3% at one year and 88.9% 0.03% at 3 years. When the two groups were compared, a higher survival was observed for group II (log-rank p 0.005) (see Fig 1). During follow-up the causes of death in group I were neurologic (5), cardiac (4), vascular (3), and malignant tumor (1). In group II they were neurologic (7), cardiac (12), vascular (6), malignant tumor (2), and others (2). The Cox analysis identified as independent factors predicting survival: age (p 0.005; hazard ratio (HR) 1.06 [1.02 to 1.1]), chronic kidney disease (CKD) (p 0.036; HR 3.7 [1.1 to 12.3]), a high preoperative NYHA class ( 3) (p 0.012; HR 2.6 [1.2 to 5.4]), need for coronary artery bypass grafting (p 0.035; HR 2.3 [1.1 to 5.1]) and ascending aorta reoperation (p 0.01; HR 2.6 [1.2 to 5.3]) (see Table 4). Comment The number of ascending aorta and (or) or aortic root reoperations has recently increased such that many surgery teams have to confront this technically demanding procedure [1, 2, 10 19]. Today s excellent results of elective ascending aorta procedures, with a surgery-related mortality around 3% to 4% [20, 21], have encouraged these groups to undertake reoperations, accepting the theoretical increased mortality related to this type of procedure. The most common indications for a reoperation used to be an aortic root-ascending aortic aneurysm in a patient with previous aortic valve replacement. Further indications are acute dissection, the appearance of a pseudoaneurysm at the suture line, aortic insufficiency over a previously implanted homograft, autograft or xenograft and prosthetic or native endocarditis related to a dilated aortic root and (or) ascending aorta [1, 10 19].In our series, the main indication was the development of an aortic root and (or) ascending aortic aneurysm in patients who had previously had an aortic valve replacement (38 patients). It should be noted that in all but 7 patients in this subset, the diameter of the previously implanted prosthesis was 25 mm or greater, indicating considerable prior dilation of the aortic root and the need for a more radical initial surgical approach. In effect, the relation between aortic valve disease and ascending aortic dilation is well known, especially in cases of a bicuspid aortic valve [22 25], which is why we included patients with previous AVR in our study. Another significant finding was that the time period between the initial procedure and redo surgery was long (mean 132 months), stressing the need for prolonged follow-up in patients undergoing aortic valve replacement with an aortic root diameter at its limit. These long times between surgical procedures are in line with the times reported in the literature [1, 10, 16, 18]. For example, in a series of 56 patients, Kirsch and colleagues [16] recorded a mean time of 9.4 years and Szeto and colleagues [18] one of 7.9 years for 156 patients. Hospital mortality rates for reoperations on the ascending aorta reported by the different authors range from 5.4% to 17.9% [1, 10 19]. The results of these studies are nevertheless difficult to compare as the populations examined span from patients undergoing prior mitral replacement to those undergoing aortic root replacement with previous Bentall-De Bono procedure. Our mortality of 12.1% falls within this reported range, although we should stress that we did not include patients subjected to prior mitral or coronary artery surgery because we consider these form part of a patient subset with different clinical and prognostic characteristics, with no previous aortic root disease. We did, nevertheless, include patients who underwent isolated aortic valve replacement and subsequently developed an aortic root and (or) ascending aortic dilation, based on the link established in numerous studies between aortic valve disease and dilation of the aortic root and (or) ascending aorta [22 25]. By following up patients with isolated aortic valve replacement and a moderately dilated ascending aorta ( 5 cm), these studies have identified a greater incidence of acute dissection or dilation of the ascending aorta in the long term, especially in those with a bicuspid aortic valve and ascending aorta diameter greater than 4.5 cm [22 25]. The factors described as prognostic for in-hospital mortality after surgery on the ascending aorta, reoperations included [1, 10 19], are age, preoperative functional class, CPB time, CKD, emergency surgery, concomitant coronary artery bypass grafting and lung disease. Our reoperation group showed a higher, though not significant, mortality (12.1% vs 6.8%) and we also noted an uneven distribution of certain variables in the two groups. However, due to the limited number of in-hospital deaths, adding a potential risk of type II statistical error, we determined the need for a complete survival analysis with the Cox method including hospital and follow-up deaths. Clearly, advances in anesthesia and surgery techniques have improved the outcomes of these reopera- Table 4. Multiple Cox Proportional-Hazards Model for Survival (In-Hospital Mortality Included) Variable p Value HR (95% CI) Age 0.005 1.06 (1.02 1.1) CKD (Cr 2.26 mg/dl) 0.036 3.7 (1.1 12.3) Advanced NYHA ( 3) 0.012 2.6 (1.2 5.4) CABG 0.035 2.3 (1.1 5.1) Reoperation 0.01 2.6 (1.2 5.3) CABG coronary artery bypass grafting performed at the aorta intervention; CI confidence interval; CKD chronic kidney disease; Cr creatinine; HR hazard ratio; NYHA New York Heart Association.

560 SILVA ET AL Ann Thorac Surg ROOT AND ASCENDING AORTA REOPERATIONS 2010;90:555 60 tions. The use of axillary artery cannulation warrants special attention. This technique has rapidly gained popularity and in some reports has been identified as a factor capable of reducing neurologic complications and inhospital mortality [26 28]. Further, studies designed to assess the impacts of the type of primary surgery and the indication for repeat surgery on mortality have detected no differences. Thus, Szeto and colleagues [18] observed no difference in in-hospital mortality according to the type of prior surgery performed (AVR, proximal aortic reconstruction, or aortic root replacement). The different researchers report 1-year and 5-year survival rates for reoperations of around 73% to 92% and 65.7% to 81%, respectively [1, 10 19]. The survival rates recorded here were 77.9% and 75.3% (3 years), respectively. Our univariate analysis revealed a lower survival rate in our reoperation group than the primary surgery group, and once corrected for the variables correlated with mortality, reoperation emerged as an independent predictor of survival in the midterm. Further variables identified to impact survival were age, need for coronary artery bypass grafting at the aorta intervention, a high NYHA functional class, and CKD. These findings are consistent with the results of studies examining operations on the ascending aorta, including reoperations. Specifically, in the study by Szeto and colleagues [18] focusing on 146 reoperations, NYHA functional class was identified in the univariate analysis as a determinant of survival, although only an age older than 75 years was identified in the multivariate analysis. In the study by Estrera and colleagues [14] on 104 reoperations, CKD and chronic obstructive pulmonary disease appeared as determining factors for survival in their multivariate analysis using the Cox method. In conclusion, ascending aorta and (or) aortic root reoperations showed acceptable morbidity and mortality. Notwithstanding, our reoperated patients had a worse prognosis in the midterm compared with patients in the primary surgery group. References 1. Luciani GB, Casali G, Faggian G, Mazzucco A. Predicting outcome after reoperative procedures on the aortic root and ascending aorta. Eur J Cardiothorac Surg 2000;17:602 7. 2. Pugliese P, Pessotto R, Santini F, Montalbano G, Luciani GB, Mazzucco A. Risk of late reoperations in patients with acute type A aortic dissection: impact of a more radical surgical approach. Eur J Cardiothoracic Surg 1998;13:576 81. 3. Joudinaud TM, Baron F, Raffoul R, et al. Redo aortic surgery for failure of an aortic homograft is major technical challenge. Eur J Cardiothorac Surg 2008;33:989 94. 4. Bohm JO, Hemmer W, Rein JG, et al. A single-institution experience with the Ross operation over 11 years. Ann Thorac Surg 2009;87:514 20. 5. Ito M, Kazui T, Tamia Y, et al. Redo composite valve graft replacement. J Card Surg 2001;16:240 5. 6. Davierwala PM, Borger MA, David TE, Rao V, Maganti M, Yau TM. Reoperation is not an independent predictor of mortality during aortic valve surgery. J Thorac Cardiovasc Surg 2006;131:329 35. 7. Potter DD, Sundt TM III, Zehr KJ, et al. Risk of repeat mitral valve replacement for failed mitral valve prostheses. Ann Thorac Surg 2004;78:67 72. 8. Roques F, Nashef SA, Michel P, et al. Risk factors and outcome in European cardiac surgery: analysis of the Euro- SCORE multinational database of 19030 patients. Eur J Cardiothorac Surg 1999;15:816 22. 9. Cabrol C, Pavie A, Gandjbakhch I, et al. Complete replacement of the ascending aorta with reimplantation of the coronary arteries: new surgical approach. J Thorac Cardiovasc Surg 1981;81:309 15. 10. Dougenis D, Daily BB, Kouchoukos NT. Reoperations on the aortic root and ascending aorta. Ann Thorac Surg 1997;64: 986 92. 11. Schepens MA, Dossche KM, Morshuis WJ. Reoperations on the ascending aorta and aortic root: pitfalls and results in 134 patients. Ann Thorac Surg 1999;68:1676 80. 12. Vallely MP, Hughes C, Bannon PG, Hendel PN, French BG, Bayfield MS. Composite graft replacement of the aortic root after previous cardiac surgery: a 20-year experience. Ann Thorac Surg 2000;70:851 5. 13. Raanani E, David TE, Dellgren G, Armstrong S, Ivanov J, Feindel CM. Redo aortic root replacement: experience with 31 patients. Ann Thorac Surg 2001;71:1460 3. 14. Estrera Al, Miller C III, Porat E, et al. Determinants of early and late outcome for reoperations of the proximal aorta. Ann Thorac Surg 2004;78:837 45. 15. Girardi LN, Krieger KH, Mack CA, Lee LY, Tortolani AJ, Isom OW. Reoperations on the ascending aorta and aortic root in patients with previous cardiac surgery. Ann Thorac Surg 2006;82:1407 12. 16. Kirsch EW, Radu N, Mekontso-Dessap AD, 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;131:601 8. 17. David TE, Feindel CM, Ivanov J, Armstrong S. Aortic root replacement in patients with previous heart surgery. J Card Surg 2004;19:325 8. 18. Szeto WY, Bavaria JE, Bowen FW, et al. Reoperative aortic root replacement in patients with previous aortic surgery. Ann Thorac Surg 2007;84:1592 9. 19. Malvindi PG, Putte BP, Robin HH, Schepens MA, Morshuis WJ. Reoperations on the aortic root: experience in 46 patients. Ann Thorac Surg 2010;89:81 6. 20. Prifti E, Bonacchi M, Frati G, et al. Early and long-term outcome in patients undergoing aortic root replacement with composite graft according to the Bentall s technique. Eur J Cardiothorac Surg 2002;21:15 21. 21. Gott VL, Gillinov AM, Pyeritz RE, et al. Aortic root replacement. Risk factor analysis of a seventeen-year experience with 270 patients. J Thorac Cardiovasc Surg 1995;109:536 44. 22. Russo CF, Mazzetti S, Garatti A, et al. Aortic complications after bicuspid aortic valve replacement: long-term results. Ann Thorac Surg 2002;74: S1773 6. 23. Pieters FA, Widdershoven JW, Gerardy AC, Geskes G, Cheriex EC, Wellens HJ. Risk of aortic dissection after aortic valve replacement. Am J Cardiol 1993;72:1043 7. 24. Borger MA, Preston M, Ivanov J, et al. Should the ascending aorta be replaced more frequently in patients with bicuspid aortic valve disease? J Thorac Cardiovascular Surg 2004;128: 677 83. 25. Matsuyama K, Usui A, Akita T, et al. Natural history of a dilated ascending aorta after aortic valve replacement. Circ J 2005;69:392 6. 26. Halkos ME, Kerendi F, Myung R, Kilgo P, Puskas JD, Chen EP. Selective antegrade cerebral perfusion via right axillary artery cannulation reduces morbidity and mortality after proximal aortic surgery. J Thorac Cardiovasc Surg 2009;138: 1081 9. 27. Gulbins H, Pritisanac A, Ennker J. Axillary versus femoral cannulation for aortic surgery: enough evidence for a general recommendation? Ann Thorac Surg 2007;83:1219 24. 28. Gurbuz A, Emrecan B, Yilik L, et al. Aortic reoperations: experience with 23 patients using axillary artery cannulation. Int Heart J 2005;46:1099 104.