- Email: [email protected]
Is Total Arch Replacement Associated With Worse Outcomes During Repair of Acute Type A Aortic Dissection?
Is Total Arch Replacement Associated With Worse Outcomes During Repair of Acute Type A Aortic Dissection? Robert D. Rice, MD, Harleen K. Sandhu, MD, MPH, Samuel S. Leake, BS, Rana O. Afifi, MD, Ali Azizzadeh, MD, Kristofer M. Charlton-Ouw, MD, Tom C. Nguyen, MD, Charles C. Miller, III, PhD, Hazim J. Safi, MD, and Anthony L. Estrera, MD University of Texas Medical School at Houston, Memorial Hermann Heart and Vascular Institute, Houston, Texas
Background. As acute type A aortic dissection (ATAAD) remains a challenge, the extent of resection of the transverse arch remains debated during operative repair. The purpose of this study was to compare the outcomes of total arch repair versus ascending/proximal arch repair for ATAAD. Methods. We retrospectively reviewed our aortic database of ATAAD between October 1999 and December 2014. Patients were divided into two groups: total arch repair versus proximal arch repair (hemiarch). Indications for arch replacement during ATAAD include aneurysm greater than 5 cm, complex arch tear, and arch rupture. Inhospital and long-term outcomes were compared between the two groups using univariate analysis and multiple logistic regression analysis. Survival was analyzed using Kaplan-Meier and log rank statistics, and assessment of risk factors for survival was conducted by Cox proportional hazards regression analysis. Results. During the study period, we performed 489 repairs of ATAAD, 49 patients (10%) with total arch replacement and 440 patients (90%) with proximal arch
replacement. Patients with total arch repair were older (62.4 ± 13.4 years versus 57.9 ± 14.8 years, p [ 0.046) and had significantly increased retrograde aortic dissection, circulatory arrest, and retrograde cerebral perfusion times. The incidences of early mortality, stroke, and need for renal dialysis between the total arch and proximal arch group were not significantly different: 20.4% (10 of 49) versus 12.9% (57 of 440), 8.2% (4 of 49) versus 10.5% (46 of 440), and 27% (13 of 49) versus 17.6% (76 of 432), respectively. Late survival did not demonstrate a difference between groups. Conclusions. Acute type A aortic dissection remains a challenge associated with significant mortality and morbidity. When compared with a less aggressive resection, total arch replacement performed in an individualized fashion can be associated with acceptable early and late outcomes for ATAAD and was not associated with worse outcomes.
A
arch is ruptured [1–3]. Albeit that proximal arch replacement results in truncated circulatory arrest and cardiopulmonary bypass times, currently no clear evidence has surfaced for improvement in major outcomes with selection of one approach over another [4–11]. Additionally, total arch replacement for treatment of ATAAD may result in increased false lumen thrombosis and decreased subsequent aortic procedures [5–15]. Exploiting the concept of false lumen thrombosis has recently been applied through the deployment of stent grafts into the descending thoracic aorta during concomitant proximal arch or total arch replacement (frozen elephant trunk repair) [10, 13, 16, 17]. Early reviews of frozen elephant trunk repair have shown promise in freedom from aortic reintervention. Secondary to the
cute type A aortic dissection (ATAAD) remains a grave aortic lesion with an operative mortality reported in the range of 9% to 24% related to cardiac tamponade, acute aortic valvular insufficiency, branch vessel ischemia, or frank aortic rupture [1–6]. Repair of ATAAD can involve replacement of the proximal arch sparing the aortic arch branch vessels or replacement of the entire ascending and transverse aorta (total arch replacement). Total arch replacement has been traditionally selected in the following cases: complex tears where obliteration of the false lumen channel is not technically feasible with a proximal arch approach, the arch is aneurysmal, or the
Accepted for publication June 1, 2015.
(Ann Thorac Surg 2015;-:-–-) Ó 2015 by The Society of Thoracic Surgeons
Presented at the Fifty-first Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 24–28, 2015. Address correspondence to Dr Estrera, Department of Cardiothoracic and Vascular Surgery, University of Texas Medical School, 6400 Fannin St, Ste 2850, Houston, TX 77030; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
Dr Azizzadeh discloses a financial relationship with Gore and Medtronic.
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.06.007
2
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
uncertainties within the literature regarding the necessity of total arch replacement versus proximal arch replacement in the mitigation of late aortic complications and the safety thereof, we explored our institutional experience with arch replacement in the setting of ATAAD.
Patients and Methods The Committee for the Protection of Human Subjects at the University of Texas Medical School at Houston approved this study with a waiver of consent.
Patients Patients with ATAAD were analyzed from a departmental aortic database and were divided into the two subcategories described. Patients with complicated ATAAD, including hypotensive episode, cardiac tamponade, acute congestive heart failure from severe aortic insufficiency, and malperfusion syndromes, including stroke, were repaired in an emergent fashion. Patients with ATAAD with neurologic devastation (awaiting recovery) and receiving anticoagulants, if stable, were repaired in an urgent fashion with initial primary antiimpulse therapy.
Operative Approach Before repair, patients were either admitted directly to the operating room or to the cardiovascular intensive care unit and begun on antiimpulse therapy with maintenance of systolic blood pressure less than 120 mm Hg. Patients were repaired through a median sternotomy, using full heparinization therapy, cardiopulmonary bypass, and profound hypothermic circulatory arrest with retrograde cerebral perfusion. Arterial cannulation was accessed through the femoral or axillary artery or ascending aorta. Venous cannulation was through a bicaval approach. A left ventricular sump catheter was routinely placed through the right superior pulmonary vein. Cardiopulmonary bypass was initiated, and the patient was systemically cooled. The patients were systemically cold until the distal repair was completed, with myocardial protection strategy using a combination with both antegrade and retrograde cold blood cardioplegia. Electroencephalograms were previously used to monitor cerebral function and to determine time to initiate circulatory arrest. Currently, however, nearinfrared spectroscopy is used to monitor cerebral oxygenation. Open distal anastomosis was performed under circulatory arrest, and total arch replacement was selected for patients with aneurysms greater than 5 cm, a complex arch tear, or arch rupture. For patients undergoing total arch replacement, the branch vessels were preserved as a single aortic cuff off the transverse arch. Patients with known connective tissue disorders had separate bypasses performed to the great vessels. Once the distal and transverse anastomoses were completed, the graft was deaired and antegrade perfusion was reinitiated through a side-arm branch of the Dacron (C.R. Bard, Haverhill, PA) graft. Systemic warming was initiated, and the proximal reconstruction was then completed through standard cardiopulmonary bypass.
Ann Thorac Surg 2015;-:-–-
The aortic root was replaced when extensively involved in the dissection, the root diameter was greater than 5 cm or there was significant valve damage.
Definitions Acute dissection was defined as dissection occurring within 2 weeks of presentation as diagnosed by computed tomography angiogram, magnetic resonance imaging, or echocardiography. Stroke was defined as a neurologic deficit confirmed by radiograph or clinical determination by neurology consultation. Postoperative bleeding was defined as unanticipated return to the operating room for bleeding. Time from admission to operative repair was extracted from hospital records. Measurement of the aorta was determined by center flow estimation from radiographs or echocardiography of the greatest diameter of the ascending aorta, regardless of the location of the dissecting flap or hematoma. Pericardial effusion was defined as the presence of a pericardial effusion, tamponade as operative or echocardiographic evidence of right atrial or ventricular compression. Hypotension was defined as persistent systolic blood pressure less than 90 mm Hg. Aneurysm in the arch was defined as an aortic diameter greater than 5.0 cm. Smoking included both current smoking and history of smoking. Retrograde type A dissection referred to dissection without an identifiable tear in the ascending or transverse arch. Intramural hematoma was designated as circular hematoma in the ascending aorta without identifiable tear. Aortic insufficiency was defined as equal to or more than moderate grade. Rupture was designated contained or free rupture of the aorta at time of exploration. Respiratory complications included prolonged intubation longer than 24 hours, reintubation, or need for tracheostomy. Early mortality was defined as death occurring within 30 days postoperatively. Proximal arch, also referred to as hemiarch, was defined as any resection less than the total arch and was associated with open distal anastomosis.
Statistics Univariate comparisons of preoperative risk factors, intraoperative variables, and postoperative outcomes between the total arch and ascending or proximal arch groups were computed by contingency table analysis for discrete variables and unpaired t test or Wilcoxon rank sum analysis for continuous variables, depending on the distribution of the data. First-order assessments for relationships between treatment group, potentially related covariates, and outcomes, such as short-term mortality, were made by stratified contingency table analysis, and covariate effects were evaluated for confounding and effect modification. Covariates that demonstrated important effects in stratified analysis were considered for further multivariable evaluation in logistic regression models for assessment of adjusted effects on short-term outcomes. Care was taken to avoid adjusted analyses based on covariates that represent intermediary elements of the causal pathway between risk factors or treatment-related
Ann Thorac Surg 2015;-:-–-
3
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
variables and outcome, although separate models were evaluated extensively for each phase of care. Long-term outcome was assessed by stratified KaplanMeier analysis and Cox proportional hazards regression. Computations were performed using SAS software version 9.4 (SAS Institute, Cary, NC). Because of the low prevalence of total arch replacement in the population and the resultant unbalanced group sample sizes, we computed post-hoc power for the main effect of extent of arch repair (total versus proximal arch) and short-term mortality, against which many other comparisons were evaluated. Post-hoc power estimates were computed
using PASS software, version 12 (NCSS, Kaysville, UT). Matched analysis, which would have reduced the sample size of the control population by more than half, was considered but was not conducted.
Results In all, 489 ATAAD repairs were completed at our institution from October 1999 to December 2014. Of these repairs, the entire arch was replaced in 49 patients and the proximal arch was replaced in 440 patients (Table 1). The patients undergoing total arch repairs were older
Table 1. Patient Demographics and Univariate Analysis of Presentation and Comorbid Conditions Variable Patient characteristics Female Age, years Comorbid conditions Hypertension Coronary artery disease Previous CABG/PCI Previous sternotomy Prior myocardial infarction Atrial fibrillation Peripheral vascular disease Prior stroke COPD Smoking Substance abuse Connective/genetica Marfan syndrome Bicuspid aortic valve Presentation Chest pain Abdominal pain Back pain Nausea/vomiting Shortness of breath Syncope Hypotension Tamponade Cardiac arrest Malperfusion syndromeb Stroke on admission eGFR,c n ¼ 465; 48;417d GFR Q1 (56.9) GFR Q2 (>56.9 to 82.1) GFR Q3 (>82.1 to 106.3) GFR Q4 (>106.3)
Total Arch n ¼ 49 (10.02%)
Hemiarch/Ascending n ¼ 440 (89.98%)
p Value
11/49 (22.5) 62.4 13.4
127/440 (28.9) 57.9 14.8
0.344 0.046
44/49 6/49 2/49 3/49 0/49 7/49 4/49 6/49 16/49 29/49 2/49 3/49 1/49 2/49
370/440 67/440 48/440 43/440 24/440 24/440 40/440 30/440 113/440 252/440 35/440 16/440 9/440 31/440
(84.1) (15.2) (10.9) (9.8) (5.5) (5.5) (9.1) (6.8) (25.7) (57.3) (7.9) (3.6) (2.1) (7.1)
0.293 0.578 0.135 0.605 0.155 0.026 1.000 0.158 0.294 0.787 0.566 0.424 1.000 0.762
376/440 (85.5) 57/440 (12.9) 206/440 (46.8) 41/440 (9.3) 86/440 (19.6) 42/440 (9.6) 89/440 (20.2) 69/440 (15.7) 15/440 (3.4) 138/440 (31.4) 32/440 (7.3) 88.6 45.3 100/417 (23.9) 104/417 (24.9) 108/417 (25.9) 105/417 (25.2)
0.739 0.010 0.284 0.451 0.412 0.409 0.303 0.626 0.019 0.635 0.254 0.266 0.329 0.329 0.329 0.329
(89.8) (12.2) (4.1) (6.1) (0.0) (14.3) (8.2) (12.2) (32.7) (59.2) (4.1) (6.1) (2.0) (4.1)
41/49 (83.7) 13/49 (26.5) 19/49 (38.8) 6/49 (12.2) 12/49 (24.5) 10/49 (20.4) 13/49 (26.5) 9/49 (18.4) 3/49 (6.1) 17/49 (34.5) 6/49 (12.2) 81.06 39.6 17/48 (35.4) 12/48 (25.0) 9/48 (18.8) 10/48 (20.8)
a b Known connective tissue disorder or hereditary/genetic disease or syndrome. Malperfusion syndrome in any vascular territory. d Gault estimated glomerular filtration rate (eGFR), mL $ min1 $ 1.73 m2. Data missing.
c
Cockcroft-
Wherever applicable, data presented as (n ¼ xxx; xx; xxx) represent number of patients in total cohort; total arch and hemiarch/ascending analyzed with available data. Categoric data are n (%) and continuous data are mean SD. CABG ¼ coronary artery bypass graft; coronary intervention.
COPD ¼ chronic obstructive pulmonary disease;
GFR ¼ glomerular filtration rate;
PCI ¼ percutaneous
4
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
Ann Thorac Surg 2015;-:-–-
(62.4 13.4 versus 57.9 14.8 years, p ¼ 0.046) and with a male predominance (77.5% versus 71.1%, p ¼ 0.34) in both groups (Table 1). Symptoms at presentation were similar between groups, except that patients with total arch replacement presented more frequently with abdominal pain (27% versus 13%, p ¼ 0.01) and with a greater incidence of retrograde type A dissection (19% versus 5%, p ¼ 0.001; Tables 1 and 2).
used for all 49 patients undergoing total arch repair, and for the vast majority of patients undergoing proximal arch replacement. Stabilization of the aortic valve was addressed with aortic valve resuspension in the majority of patients undergoing both procedures. Concomitant exploratory laparotomy was completed in a minority of patients undergoing both repair strategies. No significant intercohort differences were observed with regard to blood product usage (Table 2).
Intraoperative Characteristics Cannulation strategy was more frequently used through the femoral approach in both patient cohorts. An insignificant increased use of direct ascending cannulation was observed in the total arch repair group. Bicuspid aortic valve was encountered in 6.7%, with no preponderance for either surgical group. Previous sternotomy was required in 9.8% of patients undergoing proximal arch repair versus 6.1% of patients undergoing total arch repair, but the difference did not reach statistical significance (p ¼ 0.61). Total arch replacement was associated with significantly longer circulatory arrest time (43.1 14.3 versus 27.2 9.3 minutes, p < 0.0001) and retrograde cerebral perfusion time (38.7 14.5 versus 26.9 9.0 minutes, p < 0.0001). Retrograde cerebral perfusion was
Postoperative Outcomes Early mortality was similar between groups, with total arch at 20% (10 of 49) versus proximal arch at 13% (57 of 440; p ¼ 0.15. Patients undergoing total arch repair did not have significantly prolonged intensive care unit stays when compared with proximal arch patients (8.5 7.4 versus 9.1 8.9 days, p ¼ 0.70). Reoperations for bleeding were not significantly different between groups: total arch, 4% (2 of 49), versus proximal arch, 7% (29 of 440, p ¼ 0.76). In comparing postoperative respiratory failure, cardiac complications, and mesenteric ischemia, no significant differences were observed (Table 3). Postoperative stroke was not significantly different between groups, observed in 10.5% (46 of 440) of patients
Table 2. Intraoperative Variables Total Arch n ¼ 49 (10.02%)
Hemiarch/Ascending n ¼ 440 (89.98%)
p Value
18 (IQR 4.5–56.5) 5.10 0.9 9/49 (18.8) 7/49 (14.3) 28/49 (57.1) 9/49 (18.4) 4/49 (8.2)
10.8 (IQR 3.0–25.3) 5.14 0.9 21/440 (4.8) 73/440 (16.6) 286/440 (65.0) 66/440 (15.0) 25/440 (5.7)
0.616 0.795 0.001 0.679 0.277 0.535 0.518
35/45 (77.8) 17/45 (15.6) 3/45 (6.7) 47/49 (95.9) 43/49 (87.8) 2/49 (4.1) 106 29.1 172.6 50.1 43.1 14.3 49/49 (100) 38.7 14.5
370/429 (86.3) 44/429 (10.1) 15/429 (3.5) 370/440 (84.1) 335/440 (76.1) 38/440 (8.6) 100.6 33.8 160.3 51.4 27.2 9.3 433/440 (98.4) 26.9 9.0
0.125 0.308 0.239 0.027 0.066 0.409 0.289 0.111 <0.0001 1.000 <0.0001
0 ¼ 1, 3 (IQR 3–4.5) 0 ¼ 10, 4 (IQR 3–8) b 0 ¼ 8, 4 (IQR 2–8) b 0 ¼ 12, 8 (IQR 6–12) b 0 ¼ 33, 10 (IQR 10–20)
0 ¼ 9, 4 (IQR 3–5) 0 ¼ 87, 4 (IQR 3–7) b 0 ¼ 121, 5 (IQR 3–8) b 0 ¼ 125, 12 (IQR 6–16) b 0 ¼ 302, 10 (IQR 10–20)
0.932 0.401 0.546 0.065 0.737
Variable Admission to repair, hours (n ¼ 450; 44; 406) Aortic size, cm (n ¼ 456; 47; 409)a Retrograde type A Intramural hematoma Aortic insufficiency Rupture Coronary dissection Cannulation (n ¼ 474; 45; 429)a Femoral Ascending aorta Axillary Aortic valve resuspension Aortic root reconstruction Coronary bypass grafting Aortic cross-clamp, minutes Bypass, minutes (n ¼ 488; 49; 439)a Circulatory arrest, minutes (n ¼ 486; 49; 437)a Retrograde cerebral perfusion Retrograde cerebral perfusion time, minutes Transfusion, units Cells saved Packed red blood cells Fresh frozen plasma (n ¼ 488; 49; 439)a Platelets (n ¼ 488; 49; 439 )a Cryoprecipitates (n ¼ 488; 49; 439)a
a
b
b
b
b
a
b Missing data. Indicates number of patients who received 0 units of blood product. The median and interquartile range (IQR) here reflect data after adjusting for these numbers.
Categoric data are n (%) and continuous data are mean SD. Wherever applicable, data presented as (n ¼ xxx; xx; xxx) represent number of patients in total cohort; total arch, hemiarch/ascending analyzed with available data.
Ann Thorac Surg 2015;-:-–-
5
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
Table 3. Postoperative Outcomes Variable Reoperation for bleeding Hypotension Stroke Myocardial infarction Cardiac arrest Gastrointestinal complication Mesenteric ischemia Exploratory laparotomy Respiratory complications Dialysisa Sepsis 24-hour mortalityb 30-day mortalityc ICU LOS, days (n ¼ 471; 46; 425)d
Total Arch n ¼ 49 (10.02)
Hemiarch/Ascending n ¼ 440 (89.98)
p Value
2/49 (4.1) 14/49 (28.6) 4/49 (8.2) 0/49 (0.0) 8/49 (16.3) 6/49 (12.2) 2/49 (4.1) 5/49 (10.2) 27/49 (55.1) 13/49 (27.1) 8/49 (16.3) 2/49 (4.1) 10/49 (20.4) 8.5 7.4
29/440 (6.6) 133/440 (30.2) 46/440 (10.5) 5/440 (1.1) 49/440 (11.1) 55/440 (12.5) 16/440 (3.6) 29/440 (6.6) 249/440 (56.6) 76/432 (17.6)d 41/440 (9.3) 7/440 (1.6) 57/440 (12.9) 9.1 8.9
0.757 0.811 0.616 1.000 0.283 0.959 0.699 0.369 0.842 0.109 0.132 0.225 0.150 0.696
a
Patients who were admitted on dialysis were not evaluated for postoperative dialysis. c d dissection. Death within 30 days of operation for type A dissection. Missing data.
b
Death within 24 hours of operation for type A
Values are n (%) unless indicated otherwise. Wherever applicable, data presented as (n ¼ xxx; xx; xxx) represent number of patients in total cohort; total arch, hemiarch/ascending analyzed with available data. ICU ¼ intensive care unit;
LOS ¼ length of stay.
who underwent proximal arch and in 8% (4 of 49) who underwent total arch replacement (p ¼ 0.62). The need for temporary postoperative dialysis was higher in the total arch cohort (13 [27.1%]) versus the proximal arch cohort (76 [17.6%], p ¼ 0.11), but did not reach statistical significance (Table 3). Median follow-up was longer for the proximal arch cohort (49 months [interquartile range: 11.5 to 102.6]) as compared with the total arch cohort (15.3 months [interquartile range: 2.5 to 71.0]). Late survival was similar between groups: 1-, 5-, and 10-year survival was 72.8%, 69.9%, and 61.2%, respectively, for total arch; and 80.2%, 75.6%, and 61.3%, respectively, for proximal arch (log rank p ¼ 0.209; Fig 1). Furthermore, after adjusting for the independent effects of age and preoperative renal dysfunction using Cox regression analysis for survival data, adjusted long-term survival was similar between total arch and proximal arch repair (p ¼ 0.603; Fig 2). Pairwise contrasts, from Cox proportional hazards model, between total arch and proximal arch replacement while controlling for the covariates of advancing age (70 years) and low estimated glomerular filtration rate (preoperative glomerular filtration rate < 60) showed that both were independently associated with an increased mortality (hazard ratio 3.0, p < 0.001; and hazard ratio 1.48, p ¼ 0.027, respectively). The status of the arch was not identified to be a significant risk factor for mortality (Table 4).
replacement, often employing the frozen elephant trunk procedure with good results [2, 10, 11, 13, 17, 18]. Although the Crawford group reported higher mortality in the total arch group (30% versus 16%), they recognized the importance of eliminating the tear and recommended individualizing the extent of repair based on the situation encountered. The less aggressive approach, namely, ascending repair only, was associated with lower early mortality at the expense of a potentially higher late aortic reintervention rate, which was what Crawford observed at 30% in follow-up, and has been observed by others more recently [2, 11]. Our study, like other recent reports, demonstrated little difference in early or late outcomes between total arch
Comment Since the account by Crawford and associates [2] cautioning the use of total arch replacement for ATAAD, many have adopted an aggressive approach of total arch
Fig 1. Kaplan-Meier estimate for late survival for total arch replacement (blue line) compared with proximal arch with or without ascending aortic replacement (green line) for type A aortic dissection.
6
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
Fig 2. Kaplan-Meier estimate for late survival adjusted for age and preoperative renal dysfunction of patients operated on for acute type a dissection, comparing total arch replacement (blue line) with proximal arch (ascending aortic repair with or without proximal arch) replacement (green line).
and proximal arch (hemiarch) replacement. At presentation, only abdominal pain and retrograde type A aortic dissection differed between groups. As expected, operative variables in the total arch group differed, with longer cardiopulmonary bypass and circulatory arrest times. What was not characterized, however, was the specific size of the transverse arch, as only the largest diameter of the ascending aorta was recorded. Moreover, the location and orientation of the tears in the transverse arch were not uniformly noted. For these reasons, it remained uncertain whether performing the total arch replacement was a true reflection of the actual disease pathology. Therefore, some degree of competing risks by indication cannot be ruled out. In patients with retrograde type A aortic dissection, where no tear was identified, total arch replacement was performed more frequently, again, possibly reflecting the complexity of the arch pathology. Ultimately, the decision to replace the transverse arch was left to the discretion of the operating surgeon and, as such, remains a significant limitation of most surgical series, because accounting for surgeon skill and experience was not objectified. However, it has been suggested that repairs of ATAAD by dedicated aortic teams may yield improved results [19]. Many differing approaches to cannulation strategy and cerebral perfusion during circulatory arrest have been reported. We continue to use retrograde cerebral perfusion, as we have reported acceptable results. With most cases in this study undergoing retrograde cerebral perfusion, the
Ann Thorac Surg 2015;-:-–-
incidence of stroke did not differ between groups, 8.2% in the total arch group versus 10% in the proximal arch group, an incidence similar to other current reports. With little difference observed in our experience between perfusion strategies, the use of retrograde cerebral perfusion has recently allowed more frequent central cannulation for ATAAD. More important than the location of cannulation or the cerebral adjunct utilized for repairs of the transverse arch, however, remains the use of neuromonitoring. Recent reports have suggested little difference between retrograde cerebral perfusion and antegrade cerebral perfusion for arch replacement [20, 21]. It is neuromonitoring, whether near infrared spectroscopy, electroencephalography, or transcranial Doppler, that allows for early identification of cerebral malperfusion, providing the opportunity for correction [22]. Currently, it is assumed that most surgeons adopt a conservative, patient-specific approach to the surgical management of ATAAD involving the transverse arch accepting the potential for late complications. Assuming adherence to the general principles of ATAAD repair, which include replacement of the enlarged ascending aorta with an open distal anastomosis, resection of the all tears, obliteration of the false lumen, promoting the true lumen flow, and reestablishing aortic valve competency, the primary goal is not reducing late complications but rather achieving early survival [6]. Considering this, however, it is becoming apparent that in centers with specialization in aortic disease management, the extent of arch resection may have less bearing than previously thought. That has led some groups to adopt the frozen elephant trunk for most cases of ATAAD [10, 13, 18]. Some have reported early mortality rates of 5% using the frozen elephant trunk in the setting of ATAAD, recognizing that the patient cohort was younger and delayed in presentation [17]. Initial outcomes were promising, with 30-day mortality rates ranging from 2% to 16% and permanent neurologic dysfunction rates of 2% to 5%. These are comparable to conservative techniques of ascending repair only [8, 11, 22]. Long-term radiographic follow-up remains limited, however, and some have reported false lumen thrombosis rates of greater than 90%, with aortic reintervention rates of 2% to 6% [11]. Hence, a more aggressive approach, assuming early results are comparable, may now be warranted to diminish the late complications of aortic reintervention.
Study Limitations This retrospective study should be viewed with inherent limitations. This series consisted of a tertiary referral, single-center review with selection bias, as some patients
Table 4. Adjusted Effects of Risk Factors on Long-Term Survival, Cox Proportional Hazards Regression Variable Hemiarch ascending Age >70 years Glomerular filtration rate <60 CL ¼ confidence limits;
HR ¼ hazard ratio.
Parameter Estimate
HR
95% CL
p Value
0.13828 1.09875 0.39199
0.871 3.000 1.480
0.52–1.47 2.12–4.25 1.05–2.09
0.6041 <0.0001 0.0273
Ann Thorac Surg 2015;-:-–-
either did not undergo intervention for reasons of death in transit or refused surgery. Additionally, bias related to the operating surgeon’s preference remains difficult to account for, because each surgeon had an approach to ATAAD based on experience and comfort level. Statistical power for the main effect of total arch replacement was low, given the modest number of cases managed with this technique. Post-hoc power for the main comparison between total arch replacement and 30day mortality was 34%. A sample size of 2,000 (200 total arch and 1,800 proximal arch/ascending) would have been required to raise power to the traditional threshold of 80% if event rates and the ratio between types of repair were held constant. Consequently, the small size of the total arch group makes statistical inference about the comparative mortality in the two treatment groups difficult. However, it also indicates that the effect size between the groups is small. Until more data are available, the results should be viewed as preliminary.
Conclusion Acute type A aortic dissection remains a challenge associated with significant mortality and morbidity. When compared with a less aggressive resection, total arch replacement performed on an individualized basis can be associated with acceptable early and late outcomes in ATAAD and was not associated with worse outcomes. Finally, further studies regarding approaches to reduce late complications are required. We wish to thank Troy Brown for editorial assistance and Chris Akers for illustrations.
References 1. Trimarchi S, Nienaber CA, Rampoldi V, Myrmel T, Suzuki T, Mehta R. Contemporary results of surgery in acute type A aortic dissection: the International Registry of Acute Aortic Dissection experience. J Thorac Cardiovasc Surg 2005;129:112–22. 2. Crawford ES, Kirklin JW, Naftel DC, Svensson LG, Coselli JS, Safi HJ. Surgery for acute dissection of ascending aorta. Should the arch be included? J Thorac Cardiovasc Surg 1992;104:46–59. 3. Moon MR, Sundt TM, Pasque MK, et al. Does the extent of proximal or distal resection influence outcome for type A dissections? Ann Thorac Surg 2001;71:1244–9. 4. Kazui T, Washiyama N, Bashar AH, et al. Surgical outcome of acute type A aortic dissection: analysis of risk factors. Ann Thorac Surg 2002;74:75–81. 5. Mody PS, Wang Y, Geirsson A, Kim N, et al. Trends in aortic dissection hospitalizations, interventions, and outcomes among Medicare beneficiaries in the United States, 20002011. Circ Cardiovasc Qual Outcomes 2014;7:920–8. 6. Elefteriades JA. What operation for acute type A dissection? J Thorac Cardiovasc Surg 2002;123:201–3.
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
7
7. Kim JB, Chung CH, Moon DH, et al. Total arch repair versus hemiarch repair in the management of acute DeBakey type I aortic dissection. Eur J Cardiothorac Surg 2011;40: 881–7. 8. Rylski B, Beyersdorf F, Kari FA, Schlosser J, Blanke P, Siepe M. Acute type A aortic dissection extending beyond ascending aorta: limited or extensive distal repair. J Thorac Cardiovasc Surg 2014;148:949–54. 9. Watanuki H, Ogino H, Minatoya K, et al. Is emergency total arch replacement with a modified elephant trunk technique justified for acute type A aortic dissection? Ann Thorac Surg 2007;84:1585–91. 10. Shrestha M, Fleissner F, Ius F, et al. Total aortic arch replacement with frozen elephant trunk in acute type A aortic dissections: are we pushing the limits too far? Eur J Cardiothorac Surg 2015;47:361–6. 11. Li B, Ma WG, Liu YM, Sun LZ. Is extended arch replacement justified for acute type A dissection? Interact Cardiovasc Thorac Surg 2015;20:120–6. 12. Uchida N, Shibamura H, Katayama A, et al. Operative strategy for acute type A aortic dissection: ascending aortic or hemiarch versus total arch replacement with frozen elephant trunk. Ann Thorac Surg 2009;87:773–7. 13. Sun L, Qi R, Zhu J, et al. Total arch replacement combined with stented elephant trunk implantation: a new “standard” therapy for type a dissection involving repair of the aortic arch? Circulation 2011;123:971–8. 14. Zhang H, Lang X, Lu F, et al. Acute type A dissection without intimal tear in arch: proximal or extensive repair? J Thorac Cardiovasc Surg 2014;147:1251–5. 15. Dake MD, Miller DC, Semba CP, et al. Transluminal placement of endovascular stent-grafts for the treatment of descending thoracic aortic aneurysms. N Engl J Med 1994;331:1729–34. 16. 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;125:1550–3. 17. Ma WG, Zheng J, Zhang W, et al. Frozen elephant trunk with total arch replacement for type A aortic dissections: does acuity affect operative mortality? J Thorac Cardiovasc Surg 2014;148:963–72. 18. Di Bartolomeo R, Pantaleo A, Berretta P, et al. Frozen elephant trunk surgery in acute aortic dissections. J Thorac Cardiovasc Surg 2014 Dec 2; [E-Pub ahead of print]. 19. Lenos A, Bougioukakis P, Irmie V, et al. Impact of surgical experience on outcome in surgery of acute type A aortic dissection. Eur J Cardiothorac Surg 2014 Dec 11; [E-Pub ahead of print]. 20. Ganapathi AM, Hanna JM, Schechter MA, et al. Antegrade versus retrograde cerebral perfusion for hemiarch replacement with deep hypothermic circulatory arrest: does it matter? A propensity-matched analysis. Thorac Cardiovasc Surg 2014;148:2896–902. 21. Usui A, Miyata H, Ueda Y, et al. Risk-adjusted and case-matched comparative study between antegrade and retrograde cerebral perfusion during aortic arch surgery: based on the Japan Adult Cardiovascular Surgery Database: the Japan Cardiovascular Surgery Database Organization. Gen Thorac Cardiovasc Surg 2012;60:132–9. 22. Estrera AL, Garami Z, Miller CC, et al. Cerebral monitoring with transcranial Doppler ultrasonography improves neurologic outcomes during repairs of acute type A aortic dissection. J Thorac Cardiovasc Surg 2005;129: 277–85.
DISCUSSION DR RUGGERO DE PAULIS (Rome, Italy): You say that 10% of your patients underwent total arch replacement. This was probably driven by the anatomical situation. And then you
correctly pointed out we may extend our indication among the other 90%. So how are you going to select which one should be selected to undergo total arch replacement?
8
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
DR RICE: So intraoperatively when we open the aorta after we have put the patient on retrograde cerebral perfusion, we will look at the tear characteristics within the transverse arch. If it is unrepairable and it extends into the branch vessels, then we typically would convert to a total arch strategy at that point. DR BAVARIA (Philadelphia, PA): When he did the multivariate analysis, though, there was no difference when it was adjusted for renal failure and acquired immunodeficiency syndrome, so it is interesting. DR ESTRERA: Dr De Paulis, I think that is one of the real big issues with our study but also may be one of the strengths, because there were seven surgeons who did the operations throughout that whole period, and I have to comment that there is a selection bias because it depended on who the surgeon was. And there are some surgeons, like myself and Dr Safi, who are going to be a little more aggressive at doing a more extensive operation if indicated and other surgeons may be less. But I never fault that, because in the end, as Rob pointed out early, as John Elefteriades had mentioned a decade ago, we want to have a live patient. So if we have to do a redo operation later on for a root or arch, I am okay with that. I am not going to be disappointed at doing a reoperation, it is more business for us, but in the end we have a live patient. And so I bring that point up also to comment about Dr Shresta’s paper (on use of the frozen elephant trunk for acute type A aortic dissection), because he is an outstanding surgeon so
Ann Thorac Surg 2015;-:-–-
he can have those good results in this setting. In the real world, we must remember that 90% of these patients were transferred from outside hospitals and sometimes those patients cannot be done at Centers of Excellence like all the people who spoke here today. For these cases that cannot transfer, I am happy if they just do a tube graft and get out of the operating room, and then we can deal with all the other issues at a later date. DR DE PAULIS: But that was actually the question. The selection of that 10% of patients was already driven by the anatomical condition. Now we want to know in the remaining 90% when you can be more aggressive. Okay, we can be more aggressive in all cases like and we can probably treat all dissection with a frozen elephant trunk, but that is the answer we are looking for. DR ESTRERA: Yes, but we are still going to be very selective. In my practice, honestly, if the tear is extended into the arch, I am going to replace the arch. If the arch is large or completely disrupted, I am going to replace it. Or if the proximal descending is large, I am going to replace and do some kind of elephant trunk. We have been limited here because we do not have the Evita and the Thoraflex. But granted, seeing these presentations, and very nice job by The Society of Thoracic Surgeons has done for putting this together, it educates me as well as the rest of the audience as to all the different options we do have as we become more aggressive in the future. And in my opinion, we probably should be more aggressive, but it is going to take steps.
Background. As acute type A aortic dissection (ATAAD) remains a challenge, the extent of resection of the transverse arch remains debated during operative repair. The purpose of this study was to compare the outcomes of total arch repair versus ascending/proximal arch repair for ATAAD. Methods. We retrospectively reviewed our aortic database of ATAAD between October 1999 and December 2014. Patients were divided into two groups: total arch repair versus proximal arch repair (hemiarch). Indications for arch replacement during ATAAD include aneurysm greater than 5 cm, complex arch tear, and arch rupture. Inhospital and long-term outcomes were compared between the two groups using univariate analysis and multiple logistic regression analysis. Survival was analyzed using Kaplan-Meier and log rank statistics, and assessment of risk factors for survival was conducted by Cox proportional hazards regression analysis. Results. During the study period, we performed 489 repairs of ATAAD, 49 patients (10%) with total arch replacement and 440 patients (90%) with proximal arch
replacement. Patients with total arch repair were older (62.4 ± 13.4 years versus 57.9 ± 14.8 years, p [ 0.046) and had significantly increased retrograde aortic dissection, circulatory arrest, and retrograde cerebral perfusion times. The incidences of early mortality, stroke, and need for renal dialysis between the total arch and proximal arch group were not significantly different: 20.4% (10 of 49) versus 12.9% (57 of 440), 8.2% (4 of 49) versus 10.5% (46 of 440), and 27% (13 of 49) versus 17.6% (76 of 432), respectively. Late survival did not demonstrate a difference between groups. Conclusions. Acute type A aortic dissection remains a challenge associated with significant mortality and morbidity. When compared with a less aggressive resection, total arch replacement performed in an individualized fashion can be associated with acceptable early and late outcomes for ATAAD and was not associated with worse outcomes.
A
arch is ruptured [1–3]. Albeit that proximal arch replacement results in truncated circulatory arrest and cardiopulmonary bypass times, currently no clear evidence has surfaced for improvement in major outcomes with selection of one approach over another [4–11]. Additionally, total arch replacement for treatment of ATAAD may result in increased false lumen thrombosis and decreased subsequent aortic procedures [5–15]. Exploiting the concept of false lumen thrombosis has recently been applied through the deployment of stent grafts into the descending thoracic aorta during concomitant proximal arch or total arch replacement (frozen elephant trunk repair) [10, 13, 16, 17]. Early reviews of frozen elephant trunk repair have shown promise in freedom from aortic reintervention. Secondary to the
cute type A aortic dissection (ATAAD) remains a grave aortic lesion with an operative mortality reported in the range of 9% to 24% related to cardiac tamponade, acute aortic valvular insufficiency, branch vessel ischemia, or frank aortic rupture [1–6]. Repair of ATAAD can involve replacement of the proximal arch sparing the aortic arch branch vessels or replacement of the entire ascending and transverse aorta (total arch replacement). Total arch replacement has been traditionally selected in the following cases: complex tears where obliteration of the false lumen channel is not technically feasible with a proximal arch approach, the arch is aneurysmal, or the
Accepted for publication June 1, 2015.
(Ann Thorac Surg 2015;-:-–-) Ó 2015 by The Society of Thoracic Surgeons
Presented at the Fifty-first Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 24–28, 2015. Address correspondence to Dr Estrera, Department of Cardiothoracic and Vascular Surgery, University of Texas Medical School, 6400 Fannin St, Ste 2850, Houston, TX 77030; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
Dr Azizzadeh discloses a financial relationship with Gore and Medtronic.
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.06.007
2
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
uncertainties within the literature regarding the necessity of total arch replacement versus proximal arch replacement in the mitigation of late aortic complications and the safety thereof, we explored our institutional experience with arch replacement in the setting of ATAAD.
Patients and Methods The Committee for the Protection of Human Subjects at the University of Texas Medical School at Houston approved this study with a waiver of consent.
Patients Patients with ATAAD were analyzed from a departmental aortic database and were divided into the two subcategories described. Patients with complicated ATAAD, including hypotensive episode, cardiac tamponade, acute congestive heart failure from severe aortic insufficiency, and malperfusion syndromes, including stroke, were repaired in an emergent fashion. Patients with ATAAD with neurologic devastation (awaiting recovery) and receiving anticoagulants, if stable, were repaired in an urgent fashion with initial primary antiimpulse therapy.
Operative Approach Before repair, patients were either admitted directly to the operating room or to the cardiovascular intensive care unit and begun on antiimpulse therapy with maintenance of systolic blood pressure less than 120 mm Hg. Patients were repaired through a median sternotomy, using full heparinization therapy, cardiopulmonary bypass, and profound hypothermic circulatory arrest with retrograde cerebral perfusion. Arterial cannulation was accessed through the femoral or axillary artery or ascending aorta. Venous cannulation was through a bicaval approach. A left ventricular sump catheter was routinely placed through the right superior pulmonary vein. Cardiopulmonary bypass was initiated, and the patient was systemically cooled. The patients were systemically cold until the distal repair was completed, with myocardial protection strategy using a combination with both antegrade and retrograde cold blood cardioplegia. Electroencephalograms were previously used to monitor cerebral function and to determine time to initiate circulatory arrest. Currently, however, nearinfrared spectroscopy is used to monitor cerebral oxygenation. Open distal anastomosis was performed under circulatory arrest, and total arch replacement was selected for patients with aneurysms greater than 5 cm, a complex arch tear, or arch rupture. For patients undergoing total arch replacement, the branch vessels were preserved as a single aortic cuff off the transverse arch. Patients with known connective tissue disorders had separate bypasses performed to the great vessels. Once the distal and transverse anastomoses were completed, the graft was deaired and antegrade perfusion was reinitiated through a side-arm branch of the Dacron (C.R. Bard, Haverhill, PA) graft. Systemic warming was initiated, and the proximal reconstruction was then completed through standard cardiopulmonary bypass.
Ann Thorac Surg 2015;-:-–-
The aortic root was replaced when extensively involved in the dissection, the root diameter was greater than 5 cm or there was significant valve damage.
Definitions Acute dissection was defined as dissection occurring within 2 weeks of presentation as diagnosed by computed tomography angiogram, magnetic resonance imaging, or echocardiography. Stroke was defined as a neurologic deficit confirmed by radiograph or clinical determination by neurology consultation. Postoperative bleeding was defined as unanticipated return to the operating room for bleeding. Time from admission to operative repair was extracted from hospital records. Measurement of the aorta was determined by center flow estimation from radiographs or echocardiography of the greatest diameter of the ascending aorta, regardless of the location of the dissecting flap or hematoma. Pericardial effusion was defined as the presence of a pericardial effusion, tamponade as operative or echocardiographic evidence of right atrial or ventricular compression. Hypotension was defined as persistent systolic blood pressure less than 90 mm Hg. Aneurysm in the arch was defined as an aortic diameter greater than 5.0 cm. Smoking included both current smoking and history of smoking. Retrograde type A dissection referred to dissection without an identifiable tear in the ascending or transverse arch. Intramural hematoma was designated as circular hematoma in the ascending aorta without identifiable tear. Aortic insufficiency was defined as equal to or more than moderate grade. Rupture was designated contained or free rupture of the aorta at time of exploration. Respiratory complications included prolonged intubation longer than 24 hours, reintubation, or need for tracheostomy. Early mortality was defined as death occurring within 30 days postoperatively. Proximal arch, also referred to as hemiarch, was defined as any resection less than the total arch and was associated with open distal anastomosis.
Statistics Univariate comparisons of preoperative risk factors, intraoperative variables, and postoperative outcomes between the total arch and ascending or proximal arch groups were computed by contingency table analysis for discrete variables and unpaired t test or Wilcoxon rank sum analysis for continuous variables, depending on the distribution of the data. First-order assessments for relationships between treatment group, potentially related covariates, and outcomes, such as short-term mortality, were made by stratified contingency table analysis, and covariate effects were evaluated for confounding and effect modification. Covariates that demonstrated important effects in stratified analysis were considered for further multivariable evaluation in logistic regression models for assessment of adjusted effects on short-term outcomes. Care was taken to avoid adjusted analyses based on covariates that represent intermediary elements of the causal pathway between risk factors or treatment-related
Ann Thorac Surg 2015;-:-–-
3
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
variables and outcome, although separate models were evaluated extensively for each phase of care. Long-term outcome was assessed by stratified KaplanMeier analysis and Cox proportional hazards regression. Computations were performed using SAS software version 9.4 (SAS Institute, Cary, NC). Because of the low prevalence of total arch replacement in the population and the resultant unbalanced group sample sizes, we computed post-hoc power for the main effect of extent of arch repair (total versus proximal arch) and short-term mortality, against which many other comparisons were evaluated. Post-hoc power estimates were computed
using PASS software, version 12 (NCSS, Kaysville, UT). Matched analysis, which would have reduced the sample size of the control population by more than half, was considered but was not conducted.
Results In all, 489 ATAAD repairs were completed at our institution from October 1999 to December 2014. Of these repairs, the entire arch was replaced in 49 patients and the proximal arch was replaced in 440 patients (Table 1). The patients undergoing total arch repairs were older
Table 1. Patient Demographics and Univariate Analysis of Presentation and Comorbid Conditions Variable Patient characteristics Female Age, years Comorbid conditions Hypertension Coronary artery disease Previous CABG/PCI Previous sternotomy Prior myocardial infarction Atrial fibrillation Peripheral vascular disease Prior stroke COPD Smoking Substance abuse Connective/genetica Marfan syndrome Bicuspid aortic valve Presentation Chest pain Abdominal pain Back pain Nausea/vomiting Shortness of breath Syncope Hypotension Tamponade Cardiac arrest Malperfusion syndromeb Stroke on admission eGFR,c n ¼ 465; 48;417d GFR Q1 (56.9) GFR Q2 (>56.9 to 82.1) GFR Q3 (>82.1 to 106.3) GFR Q4 (>106.3)
Total Arch n ¼ 49 (10.02%)
Hemiarch/Ascending n ¼ 440 (89.98%)
p Value
11/49 (22.5) 62.4 13.4
127/440 (28.9) 57.9 14.8
0.344 0.046
44/49 6/49 2/49 3/49 0/49 7/49 4/49 6/49 16/49 29/49 2/49 3/49 1/49 2/49
370/440 67/440 48/440 43/440 24/440 24/440 40/440 30/440 113/440 252/440 35/440 16/440 9/440 31/440
(84.1) (15.2) (10.9) (9.8) (5.5) (5.5) (9.1) (6.8) (25.7) (57.3) (7.9) (3.6) (2.1) (7.1)
0.293 0.578 0.135 0.605 0.155 0.026 1.000 0.158 0.294 0.787 0.566 0.424 1.000 0.762
376/440 (85.5) 57/440 (12.9) 206/440 (46.8) 41/440 (9.3) 86/440 (19.6) 42/440 (9.6) 89/440 (20.2) 69/440 (15.7) 15/440 (3.4) 138/440 (31.4) 32/440 (7.3) 88.6 45.3 100/417 (23.9) 104/417 (24.9) 108/417 (25.9) 105/417 (25.2)
0.739 0.010 0.284 0.451 0.412 0.409 0.303 0.626 0.019 0.635 0.254 0.266 0.329 0.329 0.329 0.329
(89.8) (12.2) (4.1) (6.1) (0.0) (14.3) (8.2) (12.2) (32.7) (59.2) (4.1) (6.1) (2.0) (4.1)
41/49 (83.7) 13/49 (26.5) 19/49 (38.8) 6/49 (12.2) 12/49 (24.5) 10/49 (20.4) 13/49 (26.5) 9/49 (18.4) 3/49 (6.1) 17/49 (34.5) 6/49 (12.2) 81.06 39.6 17/48 (35.4) 12/48 (25.0) 9/48 (18.8) 10/48 (20.8)
a b Known connective tissue disorder or hereditary/genetic disease or syndrome. Malperfusion syndrome in any vascular territory. d Gault estimated glomerular filtration rate (eGFR), mL $ min1 $ 1.73 m2. Data missing.
c
Cockcroft-
Wherever applicable, data presented as (n ¼ xxx; xx; xxx) represent number of patients in total cohort; total arch and hemiarch/ascending analyzed with available data. Categoric data are n (%) and continuous data are mean SD. CABG ¼ coronary artery bypass graft; coronary intervention.
COPD ¼ chronic obstructive pulmonary disease;
GFR ¼ glomerular filtration rate;
PCI ¼ percutaneous
4
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
Ann Thorac Surg 2015;-:-–-
(62.4 13.4 versus 57.9 14.8 years, p ¼ 0.046) and with a male predominance (77.5% versus 71.1%, p ¼ 0.34) in both groups (Table 1). Symptoms at presentation were similar between groups, except that patients with total arch replacement presented more frequently with abdominal pain (27% versus 13%, p ¼ 0.01) and with a greater incidence of retrograde type A dissection (19% versus 5%, p ¼ 0.001; Tables 1 and 2).
used for all 49 patients undergoing total arch repair, and for the vast majority of patients undergoing proximal arch replacement. Stabilization of the aortic valve was addressed with aortic valve resuspension in the majority of patients undergoing both procedures. Concomitant exploratory laparotomy was completed in a minority of patients undergoing both repair strategies. No significant intercohort differences were observed with regard to blood product usage (Table 2).
Intraoperative Characteristics Cannulation strategy was more frequently used through the femoral approach in both patient cohorts. An insignificant increased use of direct ascending cannulation was observed in the total arch repair group. Bicuspid aortic valve was encountered in 6.7%, with no preponderance for either surgical group. Previous sternotomy was required in 9.8% of patients undergoing proximal arch repair versus 6.1% of patients undergoing total arch repair, but the difference did not reach statistical significance (p ¼ 0.61). Total arch replacement was associated with significantly longer circulatory arrest time (43.1 14.3 versus 27.2 9.3 minutes, p < 0.0001) and retrograde cerebral perfusion time (38.7 14.5 versus 26.9 9.0 minutes, p < 0.0001). Retrograde cerebral perfusion was
Postoperative Outcomes Early mortality was similar between groups, with total arch at 20% (10 of 49) versus proximal arch at 13% (57 of 440; p ¼ 0.15. Patients undergoing total arch repair did not have significantly prolonged intensive care unit stays when compared with proximal arch patients (8.5 7.4 versus 9.1 8.9 days, p ¼ 0.70). Reoperations for bleeding were not significantly different between groups: total arch, 4% (2 of 49), versus proximal arch, 7% (29 of 440, p ¼ 0.76). In comparing postoperative respiratory failure, cardiac complications, and mesenteric ischemia, no significant differences were observed (Table 3). Postoperative stroke was not significantly different between groups, observed in 10.5% (46 of 440) of patients
Table 2. Intraoperative Variables Total Arch n ¼ 49 (10.02%)
Hemiarch/Ascending n ¼ 440 (89.98%)
p Value
18 (IQR 4.5–56.5) 5.10 0.9 9/49 (18.8) 7/49 (14.3) 28/49 (57.1) 9/49 (18.4) 4/49 (8.2)
10.8 (IQR 3.0–25.3) 5.14 0.9 21/440 (4.8) 73/440 (16.6) 286/440 (65.0) 66/440 (15.0) 25/440 (5.7)
0.616 0.795 0.001 0.679 0.277 0.535 0.518
35/45 (77.8) 17/45 (15.6) 3/45 (6.7) 47/49 (95.9) 43/49 (87.8) 2/49 (4.1) 106 29.1 172.6 50.1 43.1 14.3 49/49 (100) 38.7 14.5
370/429 (86.3) 44/429 (10.1) 15/429 (3.5) 370/440 (84.1) 335/440 (76.1) 38/440 (8.6) 100.6 33.8 160.3 51.4 27.2 9.3 433/440 (98.4) 26.9 9.0
0.125 0.308 0.239 0.027 0.066 0.409 0.289 0.111 <0.0001 1.000 <0.0001
0 ¼ 1, 3 (IQR 3–4.5) 0 ¼ 10, 4 (IQR 3–8) b 0 ¼ 8, 4 (IQR 2–8) b 0 ¼ 12, 8 (IQR 6–12) b 0 ¼ 33, 10 (IQR 10–20)
0 ¼ 9, 4 (IQR 3–5) 0 ¼ 87, 4 (IQR 3–7) b 0 ¼ 121, 5 (IQR 3–8) b 0 ¼ 125, 12 (IQR 6–16) b 0 ¼ 302, 10 (IQR 10–20)
0.932 0.401 0.546 0.065 0.737
Variable Admission to repair, hours (n ¼ 450; 44; 406) Aortic size, cm (n ¼ 456; 47; 409)a Retrograde type A Intramural hematoma Aortic insufficiency Rupture Coronary dissection Cannulation (n ¼ 474; 45; 429)a Femoral Ascending aorta Axillary Aortic valve resuspension Aortic root reconstruction Coronary bypass grafting Aortic cross-clamp, minutes Bypass, minutes (n ¼ 488; 49; 439)a Circulatory arrest, minutes (n ¼ 486; 49; 437)a Retrograde cerebral perfusion Retrograde cerebral perfusion time, minutes Transfusion, units Cells saved Packed red blood cells Fresh frozen plasma (n ¼ 488; 49; 439)a Platelets (n ¼ 488; 49; 439 )a Cryoprecipitates (n ¼ 488; 49; 439)a
a
b
b
b
b
a
b Missing data. Indicates number of patients who received 0 units of blood product. The median and interquartile range (IQR) here reflect data after adjusting for these numbers.
Categoric data are n (%) and continuous data are mean SD. Wherever applicable, data presented as (n ¼ xxx; xx; xxx) represent number of patients in total cohort; total arch, hemiarch/ascending analyzed with available data.
Ann Thorac Surg 2015;-:-–-
5
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
Table 3. Postoperative Outcomes Variable Reoperation for bleeding Hypotension Stroke Myocardial infarction Cardiac arrest Gastrointestinal complication Mesenteric ischemia Exploratory laparotomy Respiratory complications Dialysisa Sepsis 24-hour mortalityb 30-day mortalityc ICU LOS, days (n ¼ 471; 46; 425)d
Total Arch n ¼ 49 (10.02)
Hemiarch/Ascending n ¼ 440 (89.98)
p Value
2/49 (4.1) 14/49 (28.6) 4/49 (8.2) 0/49 (0.0) 8/49 (16.3) 6/49 (12.2) 2/49 (4.1) 5/49 (10.2) 27/49 (55.1) 13/49 (27.1) 8/49 (16.3) 2/49 (4.1) 10/49 (20.4) 8.5 7.4
29/440 (6.6) 133/440 (30.2) 46/440 (10.5) 5/440 (1.1) 49/440 (11.1) 55/440 (12.5) 16/440 (3.6) 29/440 (6.6) 249/440 (56.6) 76/432 (17.6)d 41/440 (9.3) 7/440 (1.6) 57/440 (12.9) 9.1 8.9
0.757 0.811 0.616 1.000 0.283 0.959 0.699 0.369 0.842 0.109 0.132 0.225 0.150 0.696
a
Patients who were admitted on dialysis were not evaluated for postoperative dialysis. c d dissection. Death within 30 days of operation for type A dissection. Missing data.
b
Death within 24 hours of operation for type A
Values are n (%) unless indicated otherwise. Wherever applicable, data presented as (n ¼ xxx; xx; xxx) represent number of patients in total cohort; total arch, hemiarch/ascending analyzed with available data. ICU ¼ intensive care unit;
LOS ¼ length of stay.
who underwent proximal arch and in 8% (4 of 49) who underwent total arch replacement (p ¼ 0.62). The need for temporary postoperative dialysis was higher in the total arch cohort (13 [27.1%]) versus the proximal arch cohort (76 [17.6%], p ¼ 0.11), but did not reach statistical significance (Table 3). Median follow-up was longer for the proximal arch cohort (49 months [interquartile range: 11.5 to 102.6]) as compared with the total arch cohort (15.3 months [interquartile range: 2.5 to 71.0]). Late survival was similar between groups: 1-, 5-, and 10-year survival was 72.8%, 69.9%, and 61.2%, respectively, for total arch; and 80.2%, 75.6%, and 61.3%, respectively, for proximal arch (log rank p ¼ 0.209; Fig 1). Furthermore, after adjusting for the independent effects of age and preoperative renal dysfunction using Cox regression analysis for survival data, adjusted long-term survival was similar between total arch and proximal arch repair (p ¼ 0.603; Fig 2). Pairwise contrasts, from Cox proportional hazards model, between total arch and proximal arch replacement while controlling for the covariates of advancing age (70 years) and low estimated glomerular filtration rate (preoperative glomerular filtration rate < 60) showed that both were independently associated with an increased mortality (hazard ratio 3.0, p < 0.001; and hazard ratio 1.48, p ¼ 0.027, respectively). The status of the arch was not identified to be a significant risk factor for mortality (Table 4).
replacement, often employing the frozen elephant trunk procedure with good results [2, 10, 11, 13, 17, 18]. Although the Crawford group reported higher mortality in the total arch group (30% versus 16%), they recognized the importance of eliminating the tear and recommended individualizing the extent of repair based on the situation encountered. The less aggressive approach, namely, ascending repair only, was associated with lower early mortality at the expense of a potentially higher late aortic reintervention rate, which was what Crawford observed at 30% in follow-up, and has been observed by others more recently [2, 11]. Our study, like other recent reports, demonstrated little difference in early or late outcomes between total arch
Comment Since the account by Crawford and associates [2] cautioning the use of total arch replacement for ATAAD, many have adopted an aggressive approach of total arch
Fig 1. Kaplan-Meier estimate for late survival for total arch replacement (blue line) compared with proximal arch with or without ascending aortic replacement (green line) for type A aortic dissection.
6
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
Fig 2. Kaplan-Meier estimate for late survival adjusted for age and preoperative renal dysfunction of patients operated on for acute type a dissection, comparing total arch replacement (blue line) with proximal arch (ascending aortic repair with or without proximal arch) replacement (green line).
and proximal arch (hemiarch) replacement. At presentation, only abdominal pain and retrograde type A aortic dissection differed between groups. As expected, operative variables in the total arch group differed, with longer cardiopulmonary bypass and circulatory arrest times. What was not characterized, however, was the specific size of the transverse arch, as only the largest diameter of the ascending aorta was recorded. Moreover, the location and orientation of the tears in the transverse arch were not uniformly noted. For these reasons, it remained uncertain whether performing the total arch replacement was a true reflection of the actual disease pathology. Therefore, some degree of competing risks by indication cannot be ruled out. In patients with retrograde type A aortic dissection, where no tear was identified, total arch replacement was performed more frequently, again, possibly reflecting the complexity of the arch pathology. Ultimately, the decision to replace the transverse arch was left to the discretion of the operating surgeon and, as such, remains a significant limitation of most surgical series, because accounting for surgeon skill and experience was not objectified. However, it has been suggested that repairs of ATAAD by dedicated aortic teams may yield improved results [19]. Many differing approaches to cannulation strategy and cerebral perfusion during circulatory arrest have been reported. We continue to use retrograde cerebral perfusion, as we have reported acceptable results. With most cases in this study undergoing retrograde cerebral perfusion, the
Ann Thorac Surg 2015;-:-–-
incidence of stroke did not differ between groups, 8.2% in the total arch group versus 10% in the proximal arch group, an incidence similar to other current reports. With little difference observed in our experience between perfusion strategies, the use of retrograde cerebral perfusion has recently allowed more frequent central cannulation for ATAAD. More important than the location of cannulation or the cerebral adjunct utilized for repairs of the transverse arch, however, remains the use of neuromonitoring. Recent reports have suggested little difference between retrograde cerebral perfusion and antegrade cerebral perfusion for arch replacement [20, 21]. It is neuromonitoring, whether near infrared spectroscopy, electroencephalography, or transcranial Doppler, that allows for early identification of cerebral malperfusion, providing the opportunity for correction [22]. Currently, it is assumed that most surgeons adopt a conservative, patient-specific approach to the surgical management of ATAAD involving the transverse arch accepting the potential for late complications. Assuming adherence to the general principles of ATAAD repair, which include replacement of the enlarged ascending aorta with an open distal anastomosis, resection of the all tears, obliteration of the false lumen, promoting the true lumen flow, and reestablishing aortic valve competency, the primary goal is not reducing late complications but rather achieving early survival [6]. Considering this, however, it is becoming apparent that in centers with specialization in aortic disease management, the extent of arch resection may have less bearing than previously thought. That has led some groups to adopt the frozen elephant trunk for most cases of ATAAD [10, 13, 18]. Some have reported early mortality rates of 5% using the frozen elephant trunk in the setting of ATAAD, recognizing that the patient cohort was younger and delayed in presentation [17]. Initial outcomes were promising, with 30-day mortality rates ranging from 2% to 16% and permanent neurologic dysfunction rates of 2% to 5%. These are comparable to conservative techniques of ascending repair only [8, 11, 22]. Long-term radiographic follow-up remains limited, however, and some have reported false lumen thrombosis rates of greater than 90%, with aortic reintervention rates of 2% to 6% [11]. Hence, a more aggressive approach, assuming early results are comparable, may now be warranted to diminish the late complications of aortic reintervention.
Study Limitations This retrospective study should be viewed with inherent limitations. This series consisted of a tertiary referral, single-center review with selection bias, as some patients
Table 4. Adjusted Effects of Risk Factors on Long-Term Survival, Cox Proportional Hazards Regression Variable Hemiarch ascending Age >70 years Glomerular filtration rate <60 CL ¼ confidence limits;
HR ¼ hazard ratio.
Parameter Estimate
HR
95% CL
p Value
0.13828 1.09875 0.39199
0.871 3.000 1.480
0.52–1.47 2.12–4.25 1.05–2.09
0.6041 <0.0001 0.0273
Ann Thorac Surg 2015;-:-–-
either did not undergo intervention for reasons of death in transit or refused surgery. Additionally, bias related to the operating surgeon’s preference remains difficult to account for, because each surgeon had an approach to ATAAD based on experience and comfort level. Statistical power for the main effect of total arch replacement was low, given the modest number of cases managed with this technique. Post-hoc power for the main comparison between total arch replacement and 30day mortality was 34%. A sample size of 2,000 (200 total arch and 1,800 proximal arch/ascending) would have been required to raise power to the traditional threshold of 80% if event rates and the ratio between types of repair were held constant. Consequently, the small size of the total arch group makes statistical inference about the comparative mortality in the two treatment groups difficult. However, it also indicates that the effect size between the groups is small. Until more data are available, the results should be viewed as preliminary.
Conclusion Acute type A aortic dissection remains a challenge associated with significant mortality and morbidity. When compared with a less aggressive resection, total arch replacement performed on an individualized basis can be associated with acceptable early and late outcomes in ATAAD and was not associated with worse outcomes. Finally, further studies regarding approaches to reduce late complications are required. We wish to thank Troy Brown for editorial assistance and Chris Akers for illustrations.
References 1. Trimarchi S, Nienaber CA, Rampoldi V, Myrmel T, Suzuki T, Mehta R. Contemporary results of surgery in acute type A aortic dissection: the International Registry of Acute Aortic Dissection experience. J Thorac Cardiovasc Surg 2005;129:112–22. 2. Crawford ES, Kirklin JW, Naftel DC, Svensson LG, Coselli JS, Safi HJ. Surgery for acute dissection of ascending aorta. Should the arch be included? J Thorac Cardiovasc Surg 1992;104:46–59. 3. Moon MR, Sundt TM, Pasque MK, et al. Does the extent of proximal or distal resection influence outcome for type A dissections? Ann Thorac Surg 2001;71:1244–9. 4. Kazui T, Washiyama N, Bashar AH, et al. Surgical outcome of acute type A aortic dissection: analysis of risk factors. Ann Thorac Surg 2002;74:75–81. 5. Mody PS, Wang Y, Geirsson A, Kim N, et al. Trends in aortic dissection hospitalizations, interventions, and outcomes among Medicare beneficiaries in the United States, 20002011. Circ Cardiovasc Qual Outcomes 2014;7:920–8. 6. Elefteriades JA. What operation for acute type A dissection? J Thorac Cardiovasc Surg 2002;123:201–3.
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
7
7. Kim JB, Chung CH, Moon DH, et al. Total arch repair versus hemiarch repair in the management of acute DeBakey type I aortic dissection. Eur J Cardiothorac Surg 2011;40: 881–7. 8. Rylski B, Beyersdorf F, Kari FA, Schlosser J, Blanke P, Siepe M. Acute type A aortic dissection extending beyond ascending aorta: limited or extensive distal repair. J Thorac Cardiovasc Surg 2014;148:949–54. 9. Watanuki H, Ogino H, Minatoya K, et al. Is emergency total arch replacement with a modified elephant trunk technique justified for acute type A aortic dissection? Ann Thorac Surg 2007;84:1585–91. 10. Shrestha M, Fleissner F, Ius F, et al. Total aortic arch replacement with frozen elephant trunk in acute type A aortic dissections: are we pushing the limits too far? Eur J Cardiothorac Surg 2015;47:361–6. 11. Li B, Ma WG, Liu YM, Sun LZ. Is extended arch replacement justified for acute type A dissection? Interact Cardiovasc Thorac Surg 2015;20:120–6. 12. Uchida N, Shibamura H, Katayama A, et al. Operative strategy for acute type A aortic dissection: ascending aortic or hemiarch versus total arch replacement with frozen elephant trunk. Ann Thorac Surg 2009;87:773–7. 13. Sun L, Qi R, Zhu J, et al. Total arch replacement combined with stented elephant trunk implantation: a new “standard” therapy for type a dissection involving repair of the aortic arch? Circulation 2011;123:971–8. 14. Zhang H, Lang X, Lu F, et al. Acute type A dissection without intimal tear in arch: proximal or extensive repair? J Thorac Cardiovasc Surg 2014;147:1251–5. 15. Dake MD, Miller DC, Semba CP, et al. Transluminal placement of endovascular stent-grafts for the treatment of descending thoracic aortic aneurysms. N Engl J Med 1994;331:1729–34. 16. 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;125:1550–3. 17. Ma WG, Zheng J, Zhang W, et al. Frozen elephant trunk with total arch replacement for type A aortic dissections: does acuity affect operative mortality? J Thorac Cardiovasc Surg 2014;148:963–72. 18. Di Bartolomeo R, Pantaleo A, Berretta P, et al. Frozen elephant trunk surgery in acute aortic dissections. J Thorac Cardiovasc Surg 2014 Dec 2; [E-Pub ahead of print]. 19. Lenos A, Bougioukakis P, Irmie V, et al. Impact of surgical experience on outcome in surgery of acute type A aortic dissection. Eur J Cardiothorac Surg 2014 Dec 11; [E-Pub ahead of print]. 20. Ganapathi AM, Hanna JM, Schechter MA, et al. Antegrade versus retrograde cerebral perfusion for hemiarch replacement with deep hypothermic circulatory arrest: does it matter? A propensity-matched analysis. Thorac Cardiovasc Surg 2014;148:2896–902. 21. Usui A, Miyata H, Ueda Y, et al. Risk-adjusted and case-matched comparative study between antegrade and retrograde cerebral perfusion during aortic arch surgery: based on the Japan Adult Cardiovascular Surgery Database: the Japan Cardiovascular Surgery Database Organization. Gen Thorac Cardiovasc Surg 2012;60:132–9. 22. Estrera AL, Garami Z, Miller CC, et al. Cerebral monitoring with transcranial Doppler ultrasonography improves neurologic outcomes during repairs of acute type A aortic dissection. J Thorac Cardiovasc Surg 2005;129: 277–85.
DISCUSSION DR RUGGERO DE PAULIS (Rome, Italy): You say that 10% of your patients underwent total arch replacement. This was probably driven by the anatomical situation. And then you
correctly pointed out we may extend our indication among the other 90%. So how are you going to select which one should be selected to undergo total arch replacement?
8
RICE ET AL AORTIC ARCH REPAIR IN ACUTE TYPE A DISSECTION
DR RICE: So intraoperatively when we open the aorta after we have put the patient on retrograde cerebral perfusion, we will look at the tear characteristics within the transverse arch. If it is unrepairable and it extends into the branch vessels, then we typically would convert to a total arch strategy at that point. DR BAVARIA (Philadelphia, PA): When he did the multivariate analysis, though, there was no difference when it was adjusted for renal failure and acquired immunodeficiency syndrome, so it is interesting. DR ESTRERA: Dr De Paulis, I think that is one of the real big issues with our study but also may be one of the strengths, because there were seven surgeons who did the operations throughout that whole period, and I have to comment that there is a selection bias because it depended on who the surgeon was. And there are some surgeons, like myself and Dr Safi, who are going to be a little more aggressive at doing a more extensive operation if indicated and other surgeons may be less. But I never fault that, because in the end, as Rob pointed out early, as John Elefteriades had mentioned a decade ago, we want to have a live patient. So if we have to do a redo operation later on for a root or arch, I am okay with that. I am not going to be disappointed at doing a reoperation, it is more business for us, but in the end we have a live patient. And so I bring that point up also to comment about Dr Shresta’s paper (on use of the frozen elephant trunk for acute type A aortic dissection), because he is an outstanding surgeon so
Ann Thorac Surg 2015;-:-–-
he can have those good results in this setting. In the real world, we must remember that 90% of these patients were transferred from outside hospitals and sometimes those patients cannot be done at Centers of Excellence like all the people who spoke here today. For these cases that cannot transfer, I am happy if they just do a tube graft and get out of the operating room, and then we can deal with all the other issues at a later date. DR DE PAULIS: But that was actually the question. The selection of that 10% of patients was already driven by the anatomical condition. Now we want to know in the remaining 90% when you can be more aggressive. Okay, we can be more aggressive in all cases like and we can probably treat all dissection with a frozen elephant trunk, but that is the answer we are looking for. DR ESTRERA: Yes, but we are still going to be very selective. In my practice, honestly, if the tear is extended into the arch, I am going to replace the arch. If the arch is large or completely disrupted, I am going to replace it. Or if the proximal descending is large, I am going to replace and do some kind of elephant trunk. We have been limited here because we do not have the Evita and the Thoraflex. But granted, seeing these presentations, and very nice job by The Society of Thoracic Surgeons has done for putting this together, it educates me as well as the rest of the audience as to all the different options we do have as we become more aggressive in the future. And in my opinion, we probably should be more aggressive, but it is going to take steps.