Pathophysiology and Surgical Treatment of Type A Acute Aortic Dissection

Abstract

Objectives: We report the pathophysiology and treatment results of type A acute aortic dissection from our 20-year experience. Methods: We studied 673 patients with type A acute aortic dissection who underwent initial treatment from 1994 through July 2014. We divided these patients into two groups. The former group comprised 448 patients from 1994 through 2008, and the latter group comprised 225 patients from 2009 onward, when the current strategy of initial treatment and surgical technique including the early organ reperfusion therapies were established.

Results: Women were significantly often presented than men in patients over 60 years of age. Thrombosed-type dissection accounted for more than half in patients over 70 years, and significantly often complicated pericardial effusion and cardiac tamponade than patent type. Malperfusion occurred in 26% of patients. Central repair operations were performed in 579 patients. In-hospital mortality for all patients was 15%, and for the patients who underwent central repair operations was 10%. Former period of operation, malperfusion, and preoperative cardiopulmonary arrest were significant risk factor of in-hospital death. Preoperative left main trunk (LMT) stents were placed in eight patients and superior mesenteric artery (SMA) intervention was performed in five, they were effective to improve the outcome. From 2009 onward, in-hospital mortality was 5.0% and there was no significant risk factor.

Conclusion: Surgical results of type A acute aortic dissection were dramatically improved in the past 20 years. Early reperfusion strategy for the patients with malperfusion improved the outcomes. (This article is a translation of Jpn J Vasc Surg 2015; 24: 127–134.)

Introduction

According to the annual report by the Japanese Association for Thoracic Surgery,¹⁾ the number of patients undergoing surgery for acute Stanford type A aortic dissection has annually increased, and the results of surgery have improved. However, the mortality rate is still approximately 10%, and further improvement is necessary. In our hospital, various techniques have been developed and modified to improve the results of surgery. In 1990, the Emergency and Critical Care Center was established. In 2000, the Cardiovascular Center was established; the arrangement of an emergency surgery system has been promoted. We analyze the pathophysiology of type A dissection, and summarize/report changes in surgical treatment and its results.

Subjects and Methods

Subjects

The subjects were 673 patients with acute Stanford type A aortic dissection who underwent initial treatment in our center between 1994 and July 2014. For diagnosis, computed tomography (CT) was primarily performed, and transesophageal ultrasonography was performed in all patients who underwent surgery. We excluded patients with cardiopulmonary arrest on arrival in whom a diagnosis was made based on autopsy imaging (AI) findings. The annual changes in the number of patients are shown in Fig. 1. Surgery was basically indicated for all type A patients, including acute-phase thrombosed-type patients. Central repair (CR) was performed for 86% of 579 patients, but not for 94 patients (14.0%) due to various reasons. Conservative treatment was selected for 39 patients with thrombosed-type aortic dissection based on the general condition or course. Fifteen patients did not wish to undergo surgery due to an advanced age. The presence of irreversible organ damage related to malperfusion syndrome was suggested in 17 patients. Ten patients were considered to be unable to resist surgery due to concomitant systemic diseases. In five patients, cardiac tamponade after arrival led to cardiac arrest, making surgery impossible.

Fig. 1. Change in the numbers of patients with type A acute aortic dissection who had been admitted to our center. CR indicates central repair.

As the study period was 20 years, 225 patients (CR: 181 patients) treated after 2009, when initial treatment and surgical methods were established as current strategies, were regarded as latter-period patients, and 448 (CR: 398) treated before 2008 as former-period patients to compare the results of surgery.

Surgery

As arterial lines for cardiopulmonary bypass, the unilateral femoral artery was used in the initial phase, axillary/femoral arteries from 2002, and side branch anastomosed to the axillary/femoral arteries after 2005. The left or right side was selected based on CT findings, and the absence of true-lumen collapse was confirmed using transesophageal or surgical-field ultrasonography at the start of extracorporeal circulation. For brain protection in patients treated by ascending or partial arch replacement, circulatory arrest at a rectal temperature of 25°C was performed in the initial phase, retrograde cerebral perfusion from 1998, and anterograde selective cerebral perfusion from 2006. In those treated by total arch replacement, anterograde selective cerebral perfusion was performed from the initial phase.

For plasty of a dissected aortic central stump, Teflon felt strips were fixed/reinforced to the intimal and adventitial sides by interrupted horizontal mattress suture with a 12–16 needle in the initial phase. From 1995, GRF glue was introduced, and Teflon felt reinforcement was discontinued. As a result, accurate hemostasis was achieved, and the operation time was shortened, improving acute-phase results. However, in the late phase, false aneurysms at the proximal anastomotic site frequently occurred.²⁾ From November 2003, a method to adhere the dissected lumen with fibrin glue and reinforce it with Teflon felt was adopted. For ascending replacement, glue was not used at the peripheral stump, and intimal-/adventitial-side Teflon felt reinforcement was conducted. However, GRF glue, BioGlue, was used in some patients.

The extent of aortic replacement was determined based on the entry site. When the extent involved the ascending aorta or lesser curvature side of the arch, ascending or partial arch replacement was performed. When the extent involved the greater curvature side of the arch and proximal descending part, total arch replacement was conducted. When there was no entry, from ascending aorta to proximal descending aorta replacement alone was performed. To date, this strategy has not been changed since the initial phase.

The management of organ malperfusion has annually changed. In the initial phase, CR was promptly performed, but, currently, a strategy to simultaneously achieve organ reperfusion while promoting preparations for CR is adopted. To treat left coronary ischemia, coronary angiography and stent insertion into the left coronary trunk are initially performed (Fig. 2).³⁾ To treat abdominal ischemia, an atom tube is directly inserted into the superior mesenteric artery (SMA) under laparotomy (SMA intervention), and, after reperfusion, CR is performed (Fig. 3).⁴⁾ To treat serious lower limb ischemia, a brachial artery-femoral artery external shunt is prepared in the emergency room (ER). Dissection is often complicated by blood flow disturbance of the renal artery. However, conditions vary: poor contrast enhancement of the entire unilateral kidney on contrast-enhanced CT, that of a portion, and contrast enhancement in the delayed phase. Even when considering changes in the blood creatinine level, it is difficult to define acute-phase renal blood flow disturbance; in this study, patients with such disturbance were not included.

Fig. 2. Typical findings of the patients with left coronary malperfusion. (A) Electrocardiography of 52-year-old man revealed severe ischemic change. (B) Left coronary angiography showed stenosis of left main trunk. (C) After stent placement to left main trunk, his cardiac function recovered and he underwent subsequent central repair operation.

Fig. 3. Superior mesenteric artery intervention. We expose the proximal portion of superior mesenteric artery and find the branch vessel which is not affected by dissection (black arrow). 8 Fr atom tube is inserted (dotted arrow) to this branch vessel and arterial perfusion from femoral arterial sheath was manually performed using a syringe.

Results

The patient background with respect to age is shown in Table 1. Overall, the mean age was 65 ± 12 years, and the male-to-female ratio was 359:314. With respect to age, the proportion of males was higher in young subjects, and that of females was higher in elderly subjects. There were significant differences in sex, with a border of 60 years (P <0.01). On arrival, only 13 patients were diagnosed with Marfan’s syndrome based on typical physical findings. Thrombosed-type dissection accounted for 34% of all subjects, but its incidence was high in elderly subjects: 1/3 of those aged 60 years to 69 years and ≥50% of those aged 70 years or older were classified as thrombosed type. In females, the incidence of thrombosed-type dissection was significantly higher (P <0.01), and that of organ blood flow disturbance was significantly lower (P <0.01). Of 579 patients who underwent CR, bloody pericardial effusion was observed at the time of surgery in 306 (52.8%), and cardiac tamponade in 186 (32.1%). When comparing thrombosed-type patients with false lumen patent-type patients, the incidences of pericardial hemorrhage and cardiac tamponade in the former (n = 124, 69.3%; and n = 83, 46.4%, respectively) were significantly higher than in the latter (n = 182, 45.5%; and n = 103, 25.8%, respectively) (P <0.01). On the echographic assessment of aortic insufficiency, data could be obtained from 437 patients. The severity was evaluated as severe in 37 patients (8%) and moderate in 82 (19%).

Table 1.

Patients’ age and backgrounds

Age	No. of patients	Male (%)	Thrombosed type	Malperfusion (%)
27–39	21	15 (71)	0	6 (29)
40–49	69	55 (80)	7 (10)	27 (39)
50–59	130	85 (65)	21 (16)	41 (32)
60–69	200	105 (53)	68 (34)	51 (26)
70–79	181	73 (40)	95 (52)	36 (20)
80–	72	26 (36)	38 (53)	11 (15)

Patients with both imaging findings, such as CT findings, and ischemic symptoms were regarded as having organ malperfusion. Overall, the incidence was 26% (n = 172). With respect to age, the incidence in subjects aged 40 years to 49 years was the highest (39%). It gradually decreased with age. The malperfusion sites consisted of the coronary artery in 46 patients (6.8%), carotid artery in 53 (7.9%), spinal artery in 8 (1.2%), abdominal artery in 21 (3.4%), and lower limb artery in 76 (11%). In the thrombosed-type patients, the incidence of organ blood flow disturbance was 6.1% (n = 11), being lower than in the false lumen patency-type patients (36%, n = 161).

The results in non-surgically-treated patients differed among the reasons. None of the patients for whom conservative treatment was selected based on the state of a thrombosed false lumen or dissection died. However, the mortality rate was 87.5% in those for whom conservative treatment was selected due to an advanced age or organ malperfusion.

Of 579 patients who underwent CR, thoracic endovascular aortic repair (TEVAR) was performed for 6. To close a descending aorta entry, this procedure was selected for patients after another cardiac surgery, those in whom thoracotomy was impossible due to concomitant diseases, and those with thoracic descending aortic rupture. Four patients survived, and two died after surgery.

Of 573 patients who underwent blood vessel prosthesis implantation, ascending or partial arch replacement was selected for 453 (79.1%), and total arch replacement for 120 (20.9%). For aortic root repair, supra-coronary reconstruction after closure of the dissected lumen was performed for 531 patients, aortic root replacement for 34, and reimplantation for 8.

Of the 673 subjects, including non-CR patients, 98 (14.6%) died during admission. Of 579 patients who underwent CR, 60 (10.4%) died during admission. We investigated factors involved in death in the CR patients. Univariate analysis showed significant differences in surgery during the former period, male gender, organ malperfusion, preoperative cardiopulmonary arrest (CPA), and false lumen patent type (Table 2). However, on multivariate analysis, surgery during the former period, organ malperfusion, and preoperative CPA were significant risk factors (Table 3). We compared the results of treatment between the former and latter periods. Overall, the mortality rates during the former and latter periods were 15.4% and 12.8%, respectively, showing no significant difference. However, in 579 patients who underwent CR, they were 12.8% and 5.0%, respectively, showing a significant decrease (P <0.01). The CR-related mortality rates during the former and latter periods in patients with organ malperfusion were 30.8% and 8.6%, respectively, showing a significant decrease (P <0.01). In the latter-period patients, organ malperfusion was not a significant risk factor for death (Table 4).

Table 2.

Univariate analysis of in-hospital mortality of all patients who underwent central repair surgery

Variable	Survivors (n = 519)	Nonsurvivors (n = 60) (%)	P value
Operation in former period	347 (66.9)	51 (85.0)	0.004
Age (≥60)	329 (63.4)	39 (65.0)	0.806
Male	273 (52.6)	41 (68.3)	0.021
Cardiopulmonary arrest	25 (4.8)	11 (18.3)	0.000
Malperfusion	116 (22.4)	33 (55.0)	0.000
Thrombosed type dissection	168 (32.4)	11 (18.3)	0.026
Cardiac tamponade	167 (32.2)	19 (31.7)	0.936
Total arch replacement	106 (20.4)	12 (20.0)	0.938
Root surgery	36 (6.9)	7 (11.7)	0.186
TEVAR	4 (0.8)	2 (3.3)	0.063

TEVAR: thoracic endovascular aortic repair

Table 3.

Multivariate analysis of in-hospital mortality in all operated patients

Variable	OR	95% CI	P value
Operation in former period	3.177	1.487–6.736	0.003
Cardiopulmonary arrest	2.971	1.291–6.836	0.010
Malperfusion	4.460	2.517–7.903	0.000

CI: confidence interval; OR: odds ratio

Table 4.

Comparison of in-hospital mortality according to the period of surgery

	In-hospital mortality		P value
	1994–2008	2009–2014 (%)
All cases of CR	51/398 (12.8)	9/181 (5.0)	0.004
Malperfusion	28/91 (30.8)	5/58 (8.6)	0.002
Non-malperfusion	23/307 (7.5)	4/123 (3.3)	0.101
Cardiopulmonary arrest	11/30 (36.7)	0/6 (0)	0.075
Thrombosed type	10/134 (7.5)	1/45 (2.2)	0.205
Total arch replacement	11/65 (16.9)	1/53 (1.9)	0.007
Root surgery	4/24 (16.7)	3/19 (15.8)	0.938

CR: central repair

The results of treatment for organ malperfusion are presented below:

Coronary ischemia

Of 45 patients with coronary ischemia, percutaneous coronary intervention (PCI) was performed for 13. Of 32 non-PCI patients, CR was conducted for 28, but 14 (50%) died. Of 12 patients with left coronary ischemia, 8 died. Of 16 patients with right coronary ischemia, 6 died; those with left coronary ischemia showed unfavorable results (P = 0.127). Of 13 patients who underwent PCI (left coronary artery: 7 patients, right coronary artery: 6 patients), the heart function was improved in 12. For eight patients, CR was performed. One patient (13%) died after surgery. In one patient, there was no improvement of the heart function after PCI for the left main trunk (LMT), and CR was impossible. For four patients, CR was not indicated due to serious cerebral ischemia or liver cancer after heart-function improvement, leading to fatal outcomes. The results of CR after PCI were more favorable than those of CR in the absence of PCI (P = 0.06).

Cerebral ischemia

Of 53 patients with cerebral ischemia, coronary ischemia was present in 11. For 35 patients, CR was performed, and 10 died. In 18 patients, consciousness disorder (JCS: ≥200) was considered to be irreversible, and CR was not performed. Of these, 16 died. Of the 53 patients, 26 (49%) died. In 11 of 25 patients who tolerated surgery, recovery was achieved without sequelae; in most patients, cerebral ischemic symptoms initially appeared, but spontaneous remission was noted on arrival.

Abdominal organ ischemia

Of 23 patients with abdominal organ ischemia, CR alone or post-CR abdominal revascularization was performed for 18. Of these, nine (50%) died. For five patients, CR was performed after superior mesenteric artery blood flow was resumed (SMA intervention) through emergency laparotomy or catheter fenestration, achieving survival and significantly improving the results (P = 0.04).

Lower limb ischemia

Of 76 patients with lower limb ischemia, abdominal organ ischemia was present in 12. In seven patients with marked ischemic symptoms of the lower limbs, a brachial artery-femoral artery external shunt was prepared using a 7 Fr sheath while waiting surgery, leading to the prompt relief of the symptoms. For all patients, CR was performed. Of patients in whom lower limb ischemia persisted after CR, axillary-femoral artery bypass was additionally performed for five, and left-right femoral artery bypass for three. In 5 (6.6%), myonephropathic metabolic syndrome (MNMS) occurred, and 16 (21.1%) died. The primary cause of death was ischemia of other organs.

Discussion

According to a report from the International Registry of Acute Aortic Dissection (IRAD) in 2008, the surgery-related mortality rate in patients with type A dissection is 23.4%.⁵⁾ However, recent studies involving a single institution reported a mortality rate of 5% to 10%.^6–8) Concerning the results of surgery in our center, the in-hospital mortality rates during the past 20 years and 5 years were 10.4% and 5.0%, respectively, showing a marked improvement.

We examined patients during the past 20 years with respect to age. In young patients, the proportion of males was higher. In elderly patients, the proportion of females was higher. According to a report from the IRAD, involving comparison of males/females, the age was significantly more advanced in females, and the interval until diagnosis was prolonged, leading to unfavorable treatment results.⁹⁾ Our data showed that the results of surgery were unfavorable in males. This was possibly because the incidence of juvenile false lumen patent-type dissection, which may cause malperfusion, was higher in males. In females, the incidence of false lumen patency is significantly lower than in males, and organ blood flow disturbance is less frequent. Aortic dissection is a medial lesion, but not an arteriosclerotic disease. The loss of media-cross-linked elastic fibers may be involved in the pathogenesis of aortic dissection.¹⁰⁾ However, there are gender/age-related differences in the pathogenesis.

With respect to amputation stump plasty and anastomosis, various methods have been reported: reinforcement with Teflon felt, the use of various bioglues, and use of neomedia.⁶⁾ Although there are differences among institutions, there may have been no influence of hemorrhage or rupture at the anastomotic site on the acute-phase results of surgery. We introduced GRF glue in 1995, and experienced a large number of late false aneurysms.^2,11) Currently, reinforcement with fibrin glue and Teflon felt is conducted, and long-term results are favorable. A study reported the usefulness of BioGlue,¹²⁾ but a glue that may have a tissue toxicity should be carefully used.

For brain protection, selective anterograde cerebral perfusion is selected for all patients. In patients undergoing ascending replacement, there was no significant difference in the cerebroprotective effects in comparison with retrograde cerebral perfusion.¹³⁾ However, selective anterograde cerebral perfusion makes it possible to promptly change a strategy from ascending replacement to arch replacement. The establishment/acquisition of the procedure and development of extra-corporeal circuits have facilitated emergency surgery at night.

Concerning arterial lines for cardiopulmonary bypass, various methods, such as cardiac apex blood in-flow and perfusion to the aortic arch using Seldinger’s method, have been reported in addition to perfusion through the axillary/femoral arteries, which we adopted.^14,15) However, there is no safe, standardized perfusion method for all patients. An adequate arterial line must be reviewed based on preoperative CT findings, and the aorta should be examined using transesophageal ultrasonography and surgical-field ultrasonography after the start of perfusion. If true-lumen collapse is observed, a change of the arterial line must be promptly considered.¹⁶⁾

The incidence of organ malperfusion was 26%, being similar to that previously reported.¹⁷⁾ It is important to establish strategies to prevent/manage this disturbance. We have performed PCI for patients with coronary ischemia, SMA intervention by emergency laparotomy for those with abdominal ischemia, and external shunting for those with marked ischemia of the lower limbs, improving the results of treatment. As a result, organ malperfusion was not a risk factor for surgery-related death in the latter-period patients.

Revascularization of ischemic organs prior to CR is useful for evaluating whether or not CR should be indicated. For patients with dissection-related occlusion of the left coronary trunk or cardiogenic shock, left coronary artery stenting is initially performed. If there is no improvement of the heart function, CR is not indicated, considering that survival is difficult. For patients with abdominal organ ischemia, emergency laparotomy is primarily performed, and, if extensive intestinal necrosis is observed, treatment is discontinued at that point. These strategies significantly improved the results of surgery in patients who underwent CR, but there has been no significant improvement in the results of treatment involving all patients who consulted our hospital. Although it is ideal to improve the results of treatment in all patients who consulted our hospital, it may also be necessary to restrict further medical expenses under a diagnosis of irreversible organ disorder. Concerning the carotid artery, the results were extremely unfavorable, because the permissible time until reperfusion is limited, and because a revascularization procedure prior to CR has not been established. However, a recent study reported the favorable results of positive surgical treatment for patients with cerebral ischemia.¹⁸⁾ In patients for whom we did not indicate surgery due to consciousness disorder, the mortality rate was 89%; we cannot rule out the presence of patients in whom survival may have been achieved by positive surgery. In the future, indication criteria for surgery should be reviewed.

We have indicated surgery for thrombosed-type dissection patients in the acute phase within 24h after onset. Of 71 patients with type A thrombosed-type dissection during the latter period, 1 (1.4%) with heparin-induced thrombocytopenia (HIT)-related multiple thrombosis died after surgery. Based on the intraoperative findings, we previously reported that the incidences of pericardial hemorrhage and cardiac tamponade in thrombosed-type dissection patients were higher than in false lumen patent-type patients, and that dissection occurred in a layer closer to the outer membrane.¹⁹⁾ In Europe and the United States, the condition is termed intramural hematoma. However, intimal tears were noted in most patients with this disorder; therefore, we emphasized that this type of dissection should be regarded as reentry-free dissection.²⁰⁾ However, whether or not surgery should be indicated has been controversial: a study indicated that conservative treatment led to a favorable prognosis; another study recommended that surgery should be indicated for a limited number of patients based on the reference values of aortic diameter or false lumen thickness; and another study suggested that the position of ulcer-like projection (ULP) should be considered. In the current guidelines prepared by the Japanese Circulation Society, it is described that surgery should be indicated for thrombosed-type aortic dissection with pericardial hemorrhage or persistent pain, and that conservative treatment should be started for asymptomatic thrombosed ascending dissection. Currently, more accurate ULP diagnosis may be possible, with advances in multidetector computed tomography (MDCT) and cardiac-gated CT indication criteria for surgery using new diagnostic imaging must be established.

Acute Stanford type A aortic dissection requires emergency treatment, but its conditions vary, with anatomical changes. To improve the results of surgery, it is important to establish a typical technique and provide individualized management for individual patients. Furthermore, unexpected events, such as changes in the false lumen, new development of malperfusion, and rupture, may occur during/after surgery; therefore, an ability to adequately manage individual patients by evaluating intra-/postoperative changes is necessary. Recently, the widespread use of MDCT and advances in intraoperative transesophageal ultrasonography have markedly improved the assessment of the general condition, facilitating adequate individualized management. This may have contributed to the improvement of the results.

Conclusion

The results of surgery for acute Stanford type A aortic dissection have markedly improved. In particular, ischemic organ reperfusion prior to CR was effective in patients with organ blood flow disturbance.

Disclosure Statement

There is no conflict of interest for the authors.