Editorial Comment: What really matters in aortic dissection?

Aortic dissection affects between 5 and 30 people per million population per year, while the incidence of acute myocardial infarction exceeds 4400 per million per year [1]. More than 70% of dissection patients are hypertensive before the event and 30% have pre-existing aneurysmal dilatation of the ascending aorta [2]. Many are misdiagnosed on initial evaluation, and around 30% are discovered at autopsy [3]. It can be difficult to differentiate between dissection and myocardial infarction on clinical history or initial findings and this delays surgical repair. Seventy percent of dissection patients have electrocardiographic changes and around 25% manifest troponin elevation [4]. For those with troponin elevation, 14% have ST segment elevation, 14% ST depression and 36% T-wave inversion. Though dissection commonly involves the coronary ostia, an additional mechanism for troponin release is haemodynamic stress through acute aortic regurgitation. Those with troponin elevation have a four times elevated risk of early death [3]. Information from the International Registry of Acute Aortic Dissection (IRAD) shows that 15% of patients suffer acute myocardial ischaemia through coronary malperfusion and 3% present with myocardial infarction [1]. As for primary myocardial infarction, these patients are likely to experience myocardial stunning with cardiogenic shock or prehospital cardiac arrest. The overall number with combined dissection plus acute myocardial infarction is probably underestimated, because they die before reaching the hospital.

While no one disputes the value of protocol-driven antiplatelet therapy or thrombolysis for myocardial infarction, this has become a frequent and unnecessary risk factor in aortic dissection surgery. Inhibition of coagulation promotes bleeding during surgical repair and, intuitively, it facilities the propagation of dissection by preventing thrombosis in the false lumen [5]. It provides the surgeon with a further taxing problem in an already difficult condition. Some prefer to postpone surgery until the coagulation profile is improved. So far, there is little evidence to support or refute this approach, but an interim death may raise concerns.

During aortic root repair, it is common to find periostial dissection around the right coronary ostium, because the primary tear begins in the convexity with propagation of the false lumen proximally and distally. Periostial dissection is found less frequently in the left coronary sinus. Autopsy findings show the dissection to extend down the coronary artery through the outer one-third of the media in a circumferential or spiral passage. Haematoma within the vessel wall alters the tolerance of the coronary endothelium to shear stress, which may result in a secondary intimal tear. Narrowing of the true lumen by the false lumen or flap may interrupt flow. The process is similar to that of spontaneous coronary dissection, so lessons may be learnt from the treatment of that disorder [6]. Selective coronary angiography may worsen ischaemia if contrast injection increases pressure in the false lumen or the guide wire tears the delicate dissected membranes [7]. If extensive mural thrombus is present with no luminal flap, stenting may simply displace haematoma proximal or distal to the stent. Long segments of stent may be needed to completely exclude the false lumen. In spontaneous coronary dissection, intravascular ultrasound is recommended to guide stent deployment, but if the lumen remains adequate, conservative management is advisable [7]. Coronary bypass surgery is an option, but because of branch occlusion by dissection, this may not revascularize the ischaemic territory. Surgical revascularization is also rarely achieved soon enough to abort the infarction process.

In this issue, Imoto et al. [8] retrospectively review 516 acute Type A dissection patients, 15% of whom had coronary artery involvement documented variously by direct intraoperative inspection, coronary angiography or echocardiographic detection of wall motion abnormality. Of those with coronary dissection, 64% had myocardial ischaemia. The right coronary was most frequently affected, with the left coronary involved in only 32 (6%) of the 516 patients. Left coronary artery involvement was associated with widespread anterolateral ischaemia, ST segment elevation, extensive hypokinesia and low cardiac output state. While the overall mortality for coronary dissection patients was 24%, it was 7.4% in those without ischaemia vs 33% for those with. Relative mortalities were 15% for right coronary involvement, 47% for left main dissection and 100% when both coronaries were involved. Specific risk factors for mortality were myocardial ischaemia, left main coronary involvement and preoperative cardiac arrest. Patients with widespread anterolateral ischaemia, low cardiac output state and preoperative cardiopulmonary resuscitation had universally poor outcome. This led the authors to suggest that preoperative stenting of the left main stem is advisable before surgical repair when left coronary involvement is suspected. This occurred in only 7 (1.35% of all) patients, but for those subject to this strategy, the incidence of fatal low cardiac output state was said to be lower [8].

While early reperfusion is beneficial and may facilitate antegrade delivery of cardioplegia, survival to surgery suggests adequate left coronary flow. Those with sudden left main occlusion rarely survive. The authors acknowledge that the visit to the catheter laboratory delays surgery. Catheter manipulation may precipitate bleeding and contrast media adversely affects renal function, particularly after prolonged cardiopulmonary bypass. For these reasons, Imoto et al. [8] do not advocate preoperative stenting for an occluded right coronary artery. Others caution against early dissection repair in patients with established myocardial infarction or prehospital cardiac arrest.

What really matters in aortic dissection surgery? First and foremost is survival. I wrote this editorial after successfully repairing two acute type A dissections in 24 h. Both had received the maximum antiplatelet regime. In the first patient, the dissection involved the left main and right coronary arteries. The second had circumferential periostial dissection of the right coronary artery. I consider coronary artery malperfusion to be the same as other branch malperfusion. In the antiplatelet era, there is rarely a solid thrombus in the coronary false lumen, so that aortic root repair reproducibly reperfuses the vessel. My approach is to completely transect and excise the ascending aorta just above the sinotubular junction and proximal to the innominate artery. The integrity of coronary flow is estimated by low pressure delivery of blood cardioplegia directly into the vessel irrespective of periosteal dissection. Root repair is undertaken carefully using a circumferential external Teflon ring attached with strategically placed interrupted sutures before the continuous graft suture is applied. A very small amount of Bioglue (Cryolife) is placed between the valve resuspension sutures into the dissected space to remodel the root. More cardioplegia is then delivered into the coronary ostia to reconfirm the integrity of these vessels. This process is usually complete before the systemic temperature reaches 18°C for circulatory arrest and open repair of the proximal arch. In my view, it is mandatory to excise the whole extent of the primary tear even if this involves the arch [5]. In practice, hemiarch replacement is sufficient for >90% of patients. Total arch replacement is only applied for complex arch tears. Equally, the David root repair or Bentall procedure are applied only for patients with a connective tissue disorder or aortic valve disease. Surgical bypass grafts are applied only for residual ischaemia apparent through electrocardiographic or echocardiographic findings on discontinuing cardiopulmonary bypass.

Irrespective of more extensive surgical repair involving total aortic arch and branch replacement, or distal stent graft deployment, surgical mortality in most series still exceeds 20% [9, 10]. Not every on-call surgeon is capable of extensive thoracic aortic repair with an out-of-hours team. With a simple and easily reproducible approach, the author reported 6% 30-day mortality in 100 consecutive patients in 2002 [5]. Ten years later, after 154 consecutive dissection repairs, this mortality remains at 7% despite the antiplatelet therapy issue. The GERAADA database explains why this might be. Risk factors for surgical mortality are age (P = 0.007), preoperative cardiopulmonary resuscitation (P = 0.04) and, importantly, duration of cerebral perfusion or hypothermic circulatory arrest (P = 0.04) [10]. While not wishing to over-emphasize the importance of time, delay before reaching the operating room, duration of cardiopulmonary bypass and hypothermic circulatory arrest all impact on outcome in elderly or high-risk groups. The use of hybrid operating theatres that include catheter laboratory equipment may improve outcomes, but most of the hospitals have yet to obtain these facilities. In the meantime, the maxim ‘simple is safe’ has merit in aortic dissection.

REFERENCES