Total Arch Replacement Versus More Conservative Management in Type A Acute Aortic Dissection Marco Di Eusanio, MD, PhD, Paolo Berretta, MD, Mariano Cefarelli, MD, Alfonsi Jacopo, MD, Giacomo Murana, MD, Sebastiano Castrovinci, MD, and Roberto Di Bartolomeo, MD Department of Cardiac Surgery, Sant’Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy
Background. Surgical management of aortic arch in type A acute dissection (TAAD) is controversial. This study compared short-term and long-term outcomes of total arch replacement (TAR) interventions versus more conservative arch management (CAM). Methods. Between 1997 and 2012, 240 patients underwent TAAD surgery in our institution; 53 (22.1%) received TAR and 187 (77.9%) received CAM. Compared with CAM patients, those undergoing TAR were younger (59.1 vs 64.4 years, p [ 0.004) and were less likely to present with cardiogenic shock (3.8 vs 14.4, p [ 0.02). Distal site of intimal tear (arch or descending aorta) was predictive of TAR management (odds ratio [OR], 9.1; p < 0.001). Results. Hospital mortality was similar in the groups (24.1% vs 22.6%; p [ 0.45), and no other significant differences were observed in terms of major postoperative complications. Age (OR, 1.047; p [ 0.007) and cardiopulmonary bypass time (OR, 1.005 per minute; p [ 0.05) emerged as independent predictors of hospital death. The
TAR management did not affect hospital mortality (propensity score [PS] adjusted OR: 1.51, p [ 0.36). On Kaplan-Meier analysis, 7-year survival (TAR, 52.1% ± 0.9% vs CAM, 57.2% ± 4.2%, log-rank p [ 0.9) and freedom from aortic re-intervention (TAR, 71.6% ± 1.3% vs CAM, 85.4% ± 3.9%, log-rank p [ 0.3) were similar. The PS-adjusted Cox regression showed no relationship between type of arch management and follow-up survival (hazard ratio [HR], 1.001; p [ 0.8) or need for re-intervention (HR, 1.507; p [ 0.4). Conclusions. In our experience TAR and CAM were associated with similar hospital mortality and morbidity rates. Nevertheless, the more extensive arch interventions were not protective for long-term survival and freedom from aortic re-intervention. Thus, in TAAD patients TAR remains indicated by site of intimal tear and patientspecific factors.
T
the patients, and thus perceived to carry an unreasonably higher operative risk. To address this controversy we assessed our institutional data on acute dissection surgery and compared short-term and long-term outcomes after different aortic arch interventions; total arch replacement (TAR) versus a more conservative arch management (CAM).
ype A acute aortic dissection (TAAD) is a catastrophic condition, for which emergent surgery represents mainstay of therapy. However, several important aspects of management urge to be clarified in order to refine technique selection criteria and improve patients’ outcomes. Among different open issues in TAAD surgery, how to manage the dissected aortic arch likely is one of the most impelling. While some surgeons advocate a conservative tear-oriented approach (most commonly involving hemiarch replacement) to minimize postoperative mortality and morbidity [1], others more aggressively propose a systematic total arch replacement (TAR) with liberal use of elephant trunk techniques (classic and frozen) to improve long-term prognosis by contrasting late aneurysm formation at the distal aorta that may require hazardous aortic reoperations or cause death by rupture [2–8]. Extensive TAR interventions, however, are not widespread because they are technically more demanding for the surgeon and more traumatic for
Accepted for publication Feb 12, 2015. Address correspondence to Dr Di Eusanio, Cardiac Surgery Department, Sant’Orsola-Malpighi Hospital, University of Bologna, Via Massarenti 9, 40128 Bologna, Italy; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2015;-:-–-) Ó 2015 by The Society of Thoracic Surgeons
Patients and Methods Study Population and Analysis Plan Between 1997 and 2012, 240 patients with DeBakey type I acute aortic dissection underwent aortic surgery in our institution. Of these, 53 (22.1%) underwent total arch replacement (TAR group) and 187 (77.9%) underwent more conservative arch management involving ascending aorta, hemiarch, or partial arch replacement (CAM group) (Fig 1). Patients were considered to have an acute dissection when the process was less than 14 days old. Preoperative, intraoperative, and postoperative data were stratified by type of arch management and results were presented using statistical methods controlling for treatment-selection bias (propensity score analysis). Data were obtained using an institutional prospective database 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.02.041
2
DI EUSANIO ET AL SURGICAL MANAGEMENT OF TAAD
supplemented by a chart review. The patients were followed by outpatients’ clinic, computed tomography or magnetic resonance (CT or MRI) imaging review, telephone calls, and civil registry. The study was approved by an Institutional Review Board and did not require individual patient consent.
Ann Thorac Surg 2015;-:-–-
antegrade fashion through the graft or axillary and innominate arteries when feasible. Proximal root interventions were performed during the cooling phase when the aorta was clamped or during the rewarming period as indicated.
Statistical Analysis Surgical Technique A standard median sternotomy was used in all patients. After systemic heparinization, cardiopulmonary bypass (CPB) was instituted using a central (distal ascending aorta, aortic arch, right axillary, or innominate arteries) or a peripheral (femoral artery) site for arterial cannulation; venous drainage was obtained by means of right atrium, bicaval, or femoral cannulation, as indicated. The left side of the heart was vented through the right superior pulmonary vein or the pulmonary artery trunk. Myocardial protection was achieved with antegrade or retrograde intermittent infusion of cold crystalloid cardioplegia. The resection of the arch tear was the primary goal in all interventions. For this reason the aortic arch was always explored during a period of circulatory arrest. Brain protection was achieved with antegrade selective cerebral perfusion (ASCP) and moderate hypothermia in all cases. Our technique for ASCP has been previously described [9]. Briefly, it involves moderate hypothermia (nasopharyngeal 26 C), bilateral hemisphere perfusion, and a cerebral flow rate of 10 to 15 mL $ kg $ min adjusted to maintain a right radial arterial pressure of between 40 and 70 mm Hg. Our tools of monitoring included bilateral radial artery pressure lines, nasopharyngeal and bladder temperatures, and regional oxygen saturation in the frontal lobes by means of near-infrared spectroscopy. After distal anastomosis, CPB was re-initiated in an
Fig 1. Distribution of aortic arch in type A acute dissection procedures throughout the study period. (CAM ¼ conservative arch management; TAR ¼ total arch replacement.)
Continuous variables were expressed as mean SD or media and categoric variables as percentages. Missing data were not defaulted to negative, and denominators reflect cases reported. The Student t test and MannWhitney U test were used for continuous variables. The Pearson c2 or Fisher exact test was used for categoric variables. Univariate analyses were performed to determine relationships between measured variables and inhospital and follow-up mortality. Arch management type and variables that achieved p values less than 0.2 in the univariate analyses were introduced in multivariable analysis and Cox regression to estimate the independent effects of risk factors for hospital mortality and all-cause mortality at follow-up, respectively. From a nonparsimonious multivariable logistic regression with TAR techniques as the dependent variable, a propensity score (PS) was derived from the conditional probability that a given patient would undergo TAR management. To control for treatment selection biases, the PS for each patient was used as an adjusting variable in the binary logistic regression model and the Cox regression model. Long-term survival was investigated using Kaplan-Meier analysis and differences in survival and freedom from reintervention between groups were examined with the log-rank test. The p values less than 0.05 were considered statistically significant; SPSS software v20.0 (SPSS Inc, Chicago, IL) was used.
Ann Thorac Surg 2015;-:-–-
DI EUSANIO ET AL SURGICAL MANAGEMENT OF TAAD
Table 2. Logistic Regression Modela: Predictors of Total Aortic Arch Replacement
Results Patient Characteristics The CAM patients were older than TAR patients (64.4 years vs 59.1; p ¼ 0.004) and, although a critical status at presentation was similarly reported in the 2 groups, were more likely to be complicated by cardiogenic shock before surgery (14.4 vs 3.8; p ¼ 0.02). Furthermore, CAM subjects showed a more proximal aortic involvement by the dissection as indicated by the site of entry tears (Table 1). Logistic regression identified distal site of intimal tear (arch or descending aorta) to be the only independent predictor for TAR management (OR, 9.1; 95% CI, 4.176% to 19.911%; p < 0.001) (C statistic ¼ 0.8) (Table 2).
Operative Data Primary intimal tear resection was always achieved in TAR patients, but less frequently in CAM patients (100% vs 89%; p ¼ 0.008). In the 18 CAM patients not receiving primary tear resection, direct tear closure with separate stitches was performed in 9 cases while the tear was left alone in 6 because too distally located on the descending Table 1. Patient Characteristics (n ¼ 240) Variable Total Male Age, years (mean SD) Hypertension Renal failure Diabetes Marfan–BAV Smoking Cerebral vasculopathy CAD Aortic aneurysm IMH Redo Critical status: Cardiogenic shock Stroke/coma Intubated Gut ischemia Extension of aortic dissection: Ascending þ arch Ascending þ arch þ descending Intimal teara Root/ascending Arch/descending
3
CAM n
%
TAR n
%
187/240 125/187 64.4 138/173 8/173 8/173 5/187 34/173 11/173 12/157 85/187 8/187 6/187 33/187 25/187 7/187 9/187 .
77.9 66.8 11.2 80.2 4.6 4.6 2.7 19.7 6.4 7.6 45.5 4.3 3.2 17.6 13.4 3.7 4.8 .
53/240 41/53 59.2 40/53 3/53 1/53 3/53 10/53 4/53 3/49 32/53 3/53 4/53 6/53 2/53 3/53 1/53 1/53
22.1 77.4 12.3 75.5 5.7 1.9 5.7 19.6 7.6 6.1 61.5 5.7 7.5 11.3 3.8 5.7 1.9 1.9
44/174 25.3 14/53 130/174 74.7 39/53
26.4 73.6
130/165 78.8 16/47 35/165 21.2 31/47
34 66
p Value 0.14 0.004 0.29 0.5 0.33 0.25 0.52 0.48 0.5 0.13 0.45 0.16 0.14 0.02 0.7 0.28 0.21 0.01
Background. Surgical management of aortic arch in type A acute dissection (TAAD) is controversial. This study compared short-term and long-term outcomes of total arch replacement (TAR) interventions versus more conservative arch management (CAM). Methods. Between 1997 and 2012, 240 patients underwent TAAD surgery in our institution; 53 (22.1%) received TAR and 187 (77.9%) received CAM. Compared with CAM patients, those undergoing TAR were younger (59.1 vs 64.4 years, p [ 0.004) and were less likely to present with cardiogenic shock (3.8 vs 14.4, p [ 0.02). Distal site of intimal tear (arch or descending aorta) was predictive of TAR management (odds ratio [OR], 9.1; p < 0.001). Results. Hospital mortality was similar in the groups (24.1% vs 22.6%; p [ 0.45), and no other significant differences were observed in terms of major postoperative complications. Age (OR, 1.047; p [ 0.007) and cardiopulmonary bypass time (OR, 1.005 per minute; p [ 0.05) emerged as independent predictors of hospital death. The
TAR management did not affect hospital mortality (propensity score [PS] adjusted OR: 1.51, p [ 0.36). On Kaplan-Meier analysis, 7-year survival (TAR, 52.1% ± 0.9% vs CAM, 57.2% ± 4.2%, log-rank p [ 0.9) and freedom from aortic re-intervention (TAR, 71.6% ± 1.3% vs CAM, 85.4% ± 3.9%, log-rank p [ 0.3) were similar. The PS-adjusted Cox regression showed no relationship between type of arch management and follow-up survival (hazard ratio [HR], 1.001; p [ 0.8) or need for re-intervention (HR, 1.507; p [ 0.4). Conclusions. In our experience TAR and CAM were associated with similar hospital mortality and morbidity rates. Nevertheless, the more extensive arch interventions were not protective for long-term survival and freedom from aortic re-intervention. Thus, in TAAD patients TAR remains indicated by site of intimal tear and patientspecific factors.
T
the patients, and thus perceived to carry an unreasonably higher operative risk. To address this controversy we assessed our institutional data on acute dissection surgery and compared short-term and long-term outcomes after different aortic arch interventions; total arch replacement (TAR) versus a more conservative arch management (CAM).
ype A acute aortic dissection (TAAD) is a catastrophic condition, for which emergent surgery represents mainstay of therapy. However, several important aspects of management urge to be clarified in order to refine technique selection criteria and improve patients’ outcomes. Among different open issues in TAAD surgery, how to manage the dissected aortic arch likely is one of the most impelling. While some surgeons advocate a conservative tear-oriented approach (most commonly involving hemiarch replacement) to minimize postoperative mortality and morbidity [1], others more aggressively propose a systematic total arch replacement (TAR) with liberal use of elephant trunk techniques (classic and frozen) to improve long-term prognosis by contrasting late aneurysm formation at the distal aorta that may require hazardous aortic reoperations or cause death by rupture [2–8]. Extensive TAR interventions, however, are not widespread because they are technically more demanding for the surgeon and more traumatic for
Accepted for publication Feb 12, 2015. Address correspondence to Dr Di Eusanio, Cardiac Surgery Department, Sant’Orsola-Malpighi Hospital, University of Bologna, Via Massarenti 9, 40128 Bologna, Italy; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2015;-:-–-) Ó 2015 by The Society of Thoracic Surgeons
Patients and Methods Study Population and Analysis Plan Between 1997 and 2012, 240 patients with DeBakey type I acute aortic dissection underwent aortic surgery in our institution. Of these, 53 (22.1%) underwent total arch replacement (TAR group) and 187 (77.9%) underwent more conservative arch management involving ascending aorta, hemiarch, or partial arch replacement (CAM group) (Fig 1). Patients were considered to have an acute dissection when the process was less than 14 days old. Preoperative, intraoperative, and postoperative data were stratified by type of arch management and results were presented using statistical methods controlling for treatment-selection bias (propensity score analysis). Data were obtained using an institutional prospective database 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.02.041
2
DI EUSANIO ET AL SURGICAL MANAGEMENT OF TAAD
supplemented by a chart review. The patients were followed by outpatients’ clinic, computed tomography or magnetic resonance (CT or MRI) imaging review, telephone calls, and civil registry. The study was approved by an Institutional Review Board and did not require individual patient consent.
Ann Thorac Surg 2015;-:-–-
antegrade fashion through the graft or axillary and innominate arteries when feasible. Proximal root interventions were performed during the cooling phase when the aorta was clamped or during the rewarming period as indicated.
Statistical Analysis Surgical Technique A standard median sternotomy was used in all patients. After systemic heparinization, cardiopulmonary bypass (CPB) was instituted using a central (distal ascending aorta, aortic arch, right axillary, or innominate arteries) or a peripheral (femoral artery) site for arterial cannulation; venous drainage was obtained by means of right atrium, bicaval, or femoral cannulation, as indicated. The left side of the heart was vented through the right superior pulmonary vein or the pulmonary artery trunk. Myocardial protection was achieved with antegrade or retrograde intermittent infusion of cold crystalloid cardioplegia. The resection of the arch tear was the primary goal in all interventions. For this reason the aortic arch was always explored during a period of circulatory arrest. Brain protection was achieved with antegrade selective cerebral perfusion (ASCP) and moderate hypothermia in all cases. Our technique for ASCP has been previously described [9]. Briefly, it involves moderate hypothermia (nasopharyngeal 26 C), bilateral hemisphere perfusion, and a cerebral flow rate of 10 to 15 mL $ kg $ min adjusted to maintain a right radial arterial pressure of between 40 and 70 mm Hg. Our tools of monitoring included bilateral radial artery pressure lines, nasopharyngeal and bladder temperatures, and regional oxygen saturation in the frontal lobes by means of near-infrared spectroscopy. After distal anastomosis, CPB was re-initiated in an
Fig 1. Distribution of aortic arch in type A acute dissection procedures throughout the study period. (CAM ¼ conservative arch management; TAR ¼ total arch replacement.)
Continuous variables were expressed as mean SD or media and categoric variables as percentages. Missing data were not defaulted to negative, and denominators reflect cases reported. The Student t test and MannWhitney U test were used for continuous variables. The Pearson c2 or Fisher exact test was used for categoric variables. Univariate analyses were performed to determine relationships between measured variables and inhospital and follow-up mortality. Arch management type and variables that achieved p values less than 0.2 in the univariate analyses were introduced in multivariable analysis and Cox regression to estimate the independent effects of risk factors for hospital mortality and all-cause mortality at follow-up, respectively. From a nonparsimonious multivariable logistic regression with TAR techniques as the dependent variable, a propensity score (PS) was derived from the conditional probability that a given patient would undergo TAR management. To control for treatment selection biases, the PS for each patient was used as an adjusting variable in the binary logistic regression model and the Cox regression model. Long-term survival was investigated using Kaplan-Meier analysis and differences in survival and freedom from reintervention between groups were examined with the log-rank test. The p values less than 0.05 were considered statistically significant; SPSS software v20.0 (SPSS Inc, Chicago, IL) was used.
Ann Thorac Surg 2015;-:-–-
DI EUSANIO ET AL SURGICAL MANAGEMENT OF TAAD
Table 2. Logistic Regression Modela: Predictors of Total Aortic Arch Replacement
Results Patient Characteristics The CAM patients were older than TAR patients (64.4 years vs 59.1; p ¼ 0.004) and, although a critical status at presentation was similarly reported in the 2 groups, were more likely to be complicated by cardiogenic shock before surgery (14.4 vs 3.8; p ¼ 0.02). Furthermore, CAM subjects showed a more proximal aortic involvement by the dissection as indicated by the site of entry tears (Table 1). Logistic regression identified distal site of intimal tear (arch or descending aorta) to be the only independent predictor for TAR management (OR, 9.1; 95% CI, 4.176% to 19.911%; p < 0.001) (C statistic ¼ 0.8) (Table 2).
Operative Data Primary intimal tear resection was always achieved in TAR patients, but less frequently in CAM patients (100% vs 89%; p ¼ 0.008). In the 18 CAM patients not receiving primary tear resection, direct tear closure with separate stitches was performed in 9 cases while the tear was left alone in 6 because too distally located on the descending Table 1. Patient Characteristics (n ¼ 240) Variable Total Male Age, years (mean SD) Hypertension Renal failure Diabetes Marfan–BAV Smoking Cerebral vasculopathy CAD Aortic aneurysm IMH Redo Critical status: Cardiogenic shock Stroke/coma Intubated Gut ischemia Extension of aortic dissection: Ascending þ arch Ascending þ arch þ descending Intimal teara Root/ascending Arch/descending
3
CAM n
%
TAR n
%
187/240 125/187 64.4 138/173 8/173 8/173 5/187 34/173 11/173 12/157 85/187 8/187 6/187 33/187 25/187 7/187 9/187 .
77.9 66.8 11.2 80.2 4.6 4.6 2.7 19.7 6.4 7.6 45.5 4.3 3.2 17.6 13.4 3.7 4.8 .
53/240 41/53 59.2 40/53 3/53 1/53 3/53 10/53 4/53 3/49 32/53 3/53 4/53 6/53 2/53 3/53 1/53 1/53
22.1 77.4 12.3 75.5 5.7 1.9 5.7 19.6 7.6 6.1 61.5 5.7 7.5 11.3 3.8 5.7 1.9 1.9
44/174 25.3 14/53 130/174 74.7 39/53
26.4 73.6
130/165 78.8 16/47 35/165 21.2 31/47
34 66
p Value 0.14 0.004 0.29 0.5 0.33 0.25 0.52 0.48 0.5 0.13 0.45 0.16 0.14 0.02 0.7 0.28 0.21 0.01
