Epidemiology

Published figures on the incidence of aortic dissection likely underestimate the actual occurrence rate, since misdiagnosis of this condition is common and the percentage of acute aortic dissection patients who expire before hospital presentation cannot be accurately estimated.² Nevertheless, acute aortic dissection is a rare event. Analysis of the Swedish National Cause of Death Register between 1987 and 2002 estimated the incidence of thoracic aortic aneurysm or dissection to be 16.3 per 100,000 men and 9.1 per 100,000 women; others have reported acute aortic dissection to affect 30 in 1 million individuals per year.³ By comparison, acute myocardial infarction (AMI) is 140-fold more common.¹ Data from the International Registry of Aortic Dissection (IRAD) indicate that the mean age of patients at presentation with aortic dissection is 63 years, with men accounting for 63% of cases.¹

Classification

Classifying aortic dissection according to anatomical location and time from onset of symptoms helps stratify risk and guide selection of initial treatment strategy (Fig. 34-1). The Stanford classification system designates dissections that involve the aorta proximal to the brachiocephalic artery (i.e., root and ascending aorta) as type A, and those that do not as type B.⁴ This distinction is clinically important because dissection involving the ascending aorta is a key determinate of early death and major morbidity. Location of the intimal tear does not influence Stanford dissection type. In the older DeBakey classification scheme, a type I dissection originates within the ascending aorta and extends for a variable distance beyond the take-off the innominate artery. A DeBakey type II dissection is confined to the ascending aorta, and a type III dissection originates in the descending thoracic aorta beyond the origin of the left subclavian artery and terminates above (type IIIA) or extends below (type IIIB) the level of the diaphragm.⁵ Although there is no single universally accepted classification system, the Stanford classification scheme is most often used in practice today and will be used throughout this chapter. The terms communicating and noncommunicating refer to the presence or absence, respectively,  of blood flow between the true and false lumens of the aorta.

Figure 34-1 Aortic dissection type according to De Bakey and Stanford classification systems.

(From Nienaber C, Eagle KA: Aortic dissection: new frontiers in diagnosis and management. Part I: from etiology to diagnostic strategies. Circulation 108:628–635, 2003.)⁷

Aortic dissection is acute if presentation occurs within 14 days of the onset of symptoms and chronic if more than 2 weeks have elapsed. Morbidity and mortality are highest within the acute phase; patients who have survived without treatment for 2 weeks are self-selected for better short- and intermediate-term outcomes.

In practice, diagnosis of aortic dissection depends on demonstration with imaging of an intimal flap with separation of true and false lumens. In type A dissection, the true lumen is usually displaced along the inner curvature of the aortic arch and continues caudally along the medial aspect of the descending thoracic aorta. Aortic branch vessel blood flow may derive from either the true or false lumen; alternatively, flow may be sluggish or absent within the false lumen, or branch vessels may be completely occluded at or near their origins.

Pathogenesis

Forces that weaken the medial layer of the aorta increase the probability of dilation, aneurysm formation, and dissection (Box 34-1). Acquired and genetic diseases that mediate this process are discussed later. In classic acute aortic dissection, the initiating event is an intimal tear through which blood rapidly surges distally into the media under systolic pressure, splitting the layers of the aortic wall and creating an intimal flap that separates the true from the false lumen.

Intimal Tear

Contemporary imaging modalities or autopsy findings identify the primary entry tear in approximately 90% of cases. It is most frequently located a few centimeters above the level of the aortic valve along the greater curvature of the aorta in cases of type A dissection and accounts for nearly 60% of all cases. Compared with other locations in the ascending aorta, the proximal few centimeters of the greater curvature are exposed to relatively greater hemodynamic, shear, and torsional force. A pivot region located in the descending thoracic aorta just beyond the insertion of the ligamentum arteriosum where the relatively mobile arch meets the fixed descending thoracic aorta is the second most common entry site for intimal tears, which will then propagate as a type B dissection (30% of cases). Arch entry occurs in 7% of cases. The abdominal aorta is the least common site for entry (3% of cases), despite the high prevalence of intima media ulcers in patients with atherosclerotic disease in this segment.^6–8

The dissecting hematoma usually propagates in an anterograde direction, although retrograde extension can occur. By this mechanism, as many as 20% of dissections that originate in the distal arch or descending thoracic aorta may involve the ascending aorta.⁹ In rare cases, a second tear may occur, resulting in a three-channel dissection¹⁰ (Fig. 34-2).

Figure 34-2 Three-channeled aortic dissection.

A, Computed tomographic image demonstrates three-channel descending aortic dissection in Marfan syndrome patient. B, Schematic representation of dissection is provided: region 1 represents thrombosed false lumen, region 2 is true lumen significantly diminished in size, and region 3 designates contrast-enhanced false lumen.

(From Yoshida S, Shidoh M: Three channeled aortic dissection. Postgrad Med J 79:532, 2003.)

Blood within the false lumen may reenter the true lumen anywhere along the length of the dissection. Reentry may be protective because of spontaneous decompression of the false lumen that may reduce the risks of rupture and/or development of malperfusion syndromes.

Aortic Rupture and End-Organ Malperfusion

Aortic rupture, defined as tearing in the vessel wall that results in extravascular hemorrhage, most commonly occurs with trauma (e.g., motor vehicle collision) but may occur as a complication of the primary dissection.¹¹ Rupture into the pericardial space resulting in cardiac tamponade occurs in type A dissection, whereas rupture into the left pleural space is usually encountered with type B dissection. Dissection-mediated end-organ ischemia or infarction occurs from (1) mechanical compression of aortic branch vessels by false lumen hematoma, (2) extension of the dissection plane across the ostium of the branch vessel, or (3) dynamic vessel inlet obstruction caused by an oscillating intimal flap. Compromise of coronary, brachiocephalic, mesenteric, renal, spinal, and iliac circulations can occur and result in a myriad of clinical presentations. Occlusion of the left ventricular (LV) outflow tract by an intimal flap has been reported (Fig. 34-3).

Figure 34-3 Aortic dissection intimal flap prolapsing into left ventricle.

Horizontal plane transesophageal echocardiographic image captures prolapse of intimal flap (arrows) from a proximal Stanford type A aortic (AO) dissection into left ventricular (LV) cavity.

(From Rosenszweig BP, Goldstein S, Sherid M, et al: Aortic dissection with flap prolapse into the left ventricle. Am J Cardiol 77:214–216, 1996.)

Predisposing Genetic Factors

As is true for aneurysms, any process that leads to destruction or degeneration of the major supporting elements of the aortic media (elastin, collagen, smooth muscle cells [SMCs]) can predispose to development of dissection. The histopathological term medial degeneration refers to noninflammatory destruction or fragmentation of elastic lamellar units, dropout of SMCs, and accumulation of mucopolysaccharide ground substances (not always in distinct cystic spaces), which characterize the final common pathway for a variety of processes that affect the integrity of the aortic media (Fig. 34-4).

Figure 34-4 Cystic medial degeneration.

A, Hematoxylin and eosin microscopic section of aorta reveals fragmentation and loss of elastin fibers with cyst-like structures present within media. B, Movat pentachrome stain emphasis medial interlamellar cystic “drop out.”

(From Maleszewski JJ, Miller DV, Lu J, et al: Histopathologic findings in ascending aortas from individuals with Loeys-Dietz syndrome [LDS]. Am J Surg Pathol 33:194–201, 2009.)

Marfan’s syndrome (MFS) is the most common inherited connective tissue disorder, with an estimated prevalence of 1 case per 3000 to 5000 individuals irrespective of racial, ethnic, or geographic considerations.^18–21 If untreated, aortic disease in MFS is progressive and incompatible with normal longevity. Half of patients in whom aortic dissection occurs at younger than 40 years of age have a history of MFS.²² Mutations in the gene encoding fibrillin-1 (FBN1), a critical component of the microfibrils that help form cellular adhesions in the extracellular matrix (ECM), are largely responsible for disease expression.²⁰ Medial degeneration is not pathognomic of MFS and may be present in numerous other conditions.²³

Marfan’s syndrome disease expression is heterogenous, but the presence of an aortic root aneurysm (aortic diameter z score ≥2) and ectopia lentis are sufficient to make the diagnosis even in the absence of a family history²⁴ (Box 34-2A). Conversely, in the absence of either of these two clinical features, the revised Ghent nosology for MFS outlines a scoring system based on the presence of FBN1 mutation and/or other key systemic features of MFS to make the diagnosis (Box 34-2B). Systemic features suggestive of MFS include positive wrist sign (entire distal phalanx of adducted thumb extends beyond ulnar border of the palm), positive thumb sign (tip of thumb covers entire fingernail of fifth finger when wrapped around contralateral wrist), pectus excavatum, pneumothorax, dural ectasia, hindfoot deformity, and protrusio acetabuli. Other clinical features less strongly associated with MFS include mitral valve prolapse, various abnormal facial features, and thoracolumbar kyphosis.

Using an in vivo murine model of MFS, Dietz et al. demonstrated that fibrillin-1 is a key target for binding of transforming growth factor (TGF)-β, a molecule involved in activation of inflammatory, fibrotic, and metalloproteinase cell signaling pathways.²⁵ To investigate the possibility that angiotensin I (AT1)-mediated TGF-β activation could represent a pharmacological target to modify development of aortic dilation in MFS, Habashi et al. studied the effects of losartan, an AT1 receptor antagonist, on aortic aneurysm formation in transgenic mice encoding a cysteine-to-glutamine substitution at position 1039 in the fibrillin-1 gene (Fbn1^C1039G/−). Mice with this fibrillin-1 mutation, the most common class of mutation associated with MFS, demonstrate significant and progressive aortic root dilation compared with wild-type mice.²⁶ Treatment with losartan attenuated TGF-β signaling in the aortic wall and resulted in full normalization of aortic wall thickness and a marked improvement in aortic wall architecture (i.e., a decrease in elastin fiber disruption) compared with placebo or β-adrenergic receptor antagonist therapy with propranolol.

These landmark observations have advanced understanding of the molecular basis of MFS and provided new targets for therapy. In one small clinical trial in young MFS patients, initiation of losartan resulted in a significant decrease in rate of growth of the aortic root (mean change 0.46 ± 0.62 mm/yr vs. 3.5 ± 2.8 mm/yr prior to treatment; P < 0.001).²⁷ These findings were supplemented by Ahimastos et al., who conducted a randomized placebo-controlled clinical trial testing the effect of the angiotensin-converting enzyme inhibitor (ACEI) perindopril on aortic root diameter and aortic stiffness in 17 MFS patients.²⁹ At 24 weeks, circulating TGF-β levels and central pulse wave velocities were lower and aortic root diameters were smaller in perindopril-treated patients. Several larger randomized clinical trials directed at modulating TGF-β signaling are ongoing (clinicaltrials.gov; NCT00683124, NCT00782327 NCT00429364, NCT00723801).

Limited forms (i.e., forme frustes) of MFS that feature cardiovascular manifestations include mitral valve prolapse syndrome (MVPS) (mitral valve prolapse, pectus excavatum, scoliosis, and mild arachnodactyly) and the MASS phenotype (myopia, mitral valve prolapse, borderline and nonprogressive aortic root dilation, skeletal findings and striae). There is increasing awareness of familial thoracic aortic aneurysm disease (FTAAD), though candidate genes affecting both matrix and SMC components have been identified in only 20% of such patients. Vascular-type Ehlers-Danlos’ syndrome (EDS) is associated with arterial rupture and dissection, including the aorta.^24,28

Ehlers-Danlos’ syndrome (1:5000 births) comprises a heterogeneous group of disorders characterized clinically by hypermobile joints, hyperextensible skin, tissue fragility, and a predisposition to spontaneous vascular rupture. Aortic involvement occurs in EDS type IV, an autosomal dominant disorder attributed to structural defects in the pro-α1 (III) chain of type III collagen, encoded by the COL3A1 gene on chromosome 2q31.³⁰

FTAAD has been mapped to other genetic loci, including 16p13.11 (MYH11 gene), 5q13-14, and 11q23.2-q24, which are not associated with abnormalities of fibrillin or collagen.^31,32 More than five mutations in the fibrillin-1 gene have been identified in patients with familial or spontaneous thoracic aortic aneurysm and dissection, with histopathological changes characteristic of medial degeneration, yet with no demonstrable abnormalities of collagen or fibrillin in fibroblast culture.^33,34 Polymorphisms encoding the gene vitamin K epoxide reductase complex subunit 1 (VKORC1), which results in under-carboxylation of specific matrix proteins, are associated with calcification of the arterial wall. In patients carrying the C allele (CT or CC), a relative two-fold increase in the probability of developing aortic dissection has been observed.³⁵ A missense mutation in the ACTA2 gene that encodes for actin filaments in SMCs is linked to 14% of patients with FTAAD.³⁶ As a consequence of this mutation, intracellular actin filament assembly is disrupted, promoting focal areas of SMC disarray, decreased SMC contraction, and medial degeneration of the aorta. This phenotype is felt to be associated with aortic wall weakening and increased predisposition to dissection.

Bicuspid aortic valve (BAV) disease is the most common congenital cardiac anomaly in adults (4:1000 live births) and is often accompanied by an aortopathy that is histologically similar but less severe than that observed in patients with MFS. Similar pathological changes have been described in patients with aortic coarctation (many of whom have BAV disease), Noonan’s syndrome, Turner’s syndrome, and polycystic kidney disease. Dilation of the root and (more commonly) the ascending aorta is present in up to 40% of patients with BAV disease and is a risk factor for dissection or rupture (Fig. 34-5).

Figure 34-5 Bicuspid aortic valve (BAV) with aortic dilation.

Electrocardiogram (ECG)-triggered breath-hold contrast-enhanced magnetic resonance angiography (MRA) using a 1.5-T imager demonstrates (A) bicuspid aortic valve (arrows) and (B) significant dilation of the root and proximal ascending aorta.

(From Arpasi PJ, Bis KG, Shetty AN, et al: MR angiography of the thoracic aorta with an electrocardiographically triggered breath-hold contrast-enhanced sequence. Radiographics 20:107–120, 2000.)

Acquired Disorders

Systemic hypertension is the most common treatable risk factor for aortic dissection and is present in approximately 75% of patients.² Hypertension accelerates the normal aging process and leads to intimal thickening, SMC apoptosis, fibrosis, loss of elasticity, and compromise of nutritive blood supply. Decreased aortic compliance and vulnerability to pulsatile forces predispose to injury and create a substrate for dissection.

Clinical Presentation

The most important element of any diagnostic algorithm for suspected acute aortic syndrome is a high clinical index of suspicion, based foremost on the presenting history and physical examination (Box 34-3). Absent an appreciation for the cardinal features of dissection, the diagnosis can be missed in a substantial number of patients. Simple clinical prediction rules have been developed to estimate probability of acute aortic dissection.⁴³ The IRAD investigators have recently confirmed the sensitivity of 12 clinical risk markers proposed in the 2010 American College of Cardiology Foundation/American Heart Association (ACCF/AHA) thoracic aortic disease guidelines² (Table 34-1). These markers, assessed at bedside, were divided into three distinct categories: predisposing factors, characteristics of the pain at time of presentation, and key physical examination findings.⁴⁴ The presence of risk factor(s) from at least one category identified 95.7% of acute aortic dissection patients in the IRAD database.

^{Table 34-1} Percentage of Patients in the International Registry of Acute Aortic Dissection (1996–2009) with Each of 12 High-Risk Clinical Markers Observed at Time of Presentation with Acute Aortic Dissection*

MFS, Marfan’s syndrome.

^* The aortic dissection detection (ADD) score aims to enhance early diagnosis of acute aortic dissection. The ADD score is calculated by determining the number of categories in which any of 12 high-risk clinical features are present in patients with symptoms suggestive of acute aortic dissection. For example, in a patient with a family history of aortic disease (category 1) and known thoracic aneurysm (also category 1), the ADD score would be 1. Likewise, the ADD score is 2 in a patient with Marfan’s syndrome (category 1) and a blood pressure differential (category 3). A retrospective analysis of the International Registry of Acute Aortic Dissection determined that among 2538 patients with acute aortic dissection, 95.7% had an ADD score ≥1. The ADD score may therefore provide the clinician with a simple and effective bedside method to inform further diagnostic testing and/or treatment in patients with suspected aortic dissection. Importantly, the negative predictive value for acute aortic dissection in patients with an ADD score of 0 has not yet been established.

From Rogers AM, Hermann LK, Booher AM, et al: Sensitivity of the aortic dissection detection risk score, a novel guideline-based tool for identification of acute aortic dissection on initial presentation. Circulation 123:2213–2218, 2011.⁴⁴