The diagnosis and management of disorders of the ascending aorta and aortic arch remain as formidable challenges in cardiovascular medicine and surgery. Until the advent of the modern era of cardiac surgery with cardiopulmonary bypass, this region of the aorta was all but inaccessible to any surgical intervention and even now definitive medical therapies remain elusive. Accurate diagnosis requires the sophisticated, yet now readily available, imaging techniques developed over the past few decades and rarely can be made on clinical grounds alone. While not as prevalent as atherosclerotic peripheral and coronary arterial occlusive disease, diseases of the proximal aorta are common enough that cardiovascular specialists are inevitably confronted with them in many different clinical settings. The sometimes enigmatic but dramatic and lethal presentation of ascending aortic dissection (AD) has attracted considerable public attention in recent years (See, for example, the 2004 Pulitzer Prize winning series by Kevin Helliker and Thomas Burton of The Wall Street Journal). Dissection and rupture, often occurring in otherwise vigorous individuals, are catastrophic biomechanical failures, and the end result of many different aortic pathologies. Aneurysmal degeneration is a common manifestation of proximal aortic disease and precedes many, but not all, instances of dissection. Congenital disease often involves obstructive lesions while atherosclerosis in this region does not. Yet even in the absence of luminal narrowing or aneurysm, the changes in aortic compliance or stiffness that accompany disease and aging have important pathophysiologic effects not only downstream in other organ systems but upstream in the heart.

In this chapter the focus is on the anatomy, microstructure, functional mechanics, and pathophysiology of the proximal aorta. This field encompasses a broad range of disciplines from multiple areas of clinical medicine, genetics, cell biology, and biomedical engineering and an attempt has been made to integrate information derived from all these areas to provide a basic level of understanding of the function, structure, and pathologies and their systemic ramifications of this unique segment of the arterial system. The preceding chapter is designed to provide a more immediately clinical discussion.

Macrostructure

The ascending aorta includes the aorta from the level of the aortic valve to the origin of the brachiocephalic artery. It is further divided into the sinus portion containing the three sinuses of valsalva and the tubular portion. The aortic valve, aortic sinuses, and coronary ostia are integral elements of the sinus portion of the aorta. The proximal aorta from the ventricular outlet to the tubular part of the ascending aorta including the sinotubular junction (STJ) is called the aortic root. The precise anatomy of the aortic root¹ is of considerable clinical importance in the surgical treatment of both aneurysmal disease and aortic valvular disease. The aortic arch, which gives rise to the brachiocephalic, left common carotid, and left subclavian arteries, extends transversely from the ascending aorta to the descending thoracic aorta distal to the subclavian origin. The region where the aortic arch joins the descending thoracic aorta is termed the aortic isthmus and is an area of important aortic pathology including coarctation and the intimal tear initiating dissection and a likely site of traumatic rupture associated with deceleration injuries (Figure 27-1).

With the exception of the trilobed sinus portion, the proximal aorta is generally round in cross-section when distended at physiologic pressures. Therefore, its size and related mechanical properties can be associated with its diameter. While in childhood the dimensions of the normal aorta bear a relationship to body size,² in adults, aortic luminal diameters are independent of weight, height, and body surface area. In a study involving 70 adults aged 17 to 89 and free of signs of cardiovascular disease, helical CT scanning was used to measure luminal diameters of the thoracic aorta at defined points from the sinuses to the diaphragm. The diameters tapered from a maximum of 3.2 (±0.42 SD) cm in men and 2.9 ± 0.38 cm in women in the ascending aorta to 2.51 ± 0.34 cm in men and 2.27 ± 0.31 cm in women at the diaphragm. Other than the small (within 1 SD) association with sex, the only important correlate was age, confirming a roughly one millimeter per decade increase in aortic diameters found in an earlier study.³

The aortic wall is composed of three layers: The intima, the media, and the adventitia. The normal intima consists of a single layer of highly biochemically active endothelial cells, a subendothelial connective tissue layer, and the internal elastic lamina. The media is composed of vascular smooth muscle cells (VSMCs), which are also biochemically active, and an extracellular matrix. The extracellular matrix contains elastin, collagen, and glycoproteins including fibrillin microfibrils and imparts unique elastomeric properties to the aortic wall.⁴ The outer limit of the media is bounded by the external elastic lamina. The adventitia is composed of collagen containing connective tissue and includes nerve fibers and the vaso vasora that supply the outer layers of the media. The inner layers of the media receive nutrient supply from the aortic lumen. Anatomic variation in aortic wall thickness in the absence of disease, is determined by the number of concentric fibromuscular layers in the media, and is linearly related to aortic diameter, not only in man but also remarkably across multiple mammalian species of markedly different size.⁵

Microstructure

The unique, mechanical distensibility and elastic recoil of the aorta and its pathophysiologic features are dependent upon the structural architecture and complicated developmental and homeostatic processes of the media. The basic structure involves the lamellar unit which consists of two fenestrated elastic lamellae encasing smooth muscle cells and extracellular matrix components.⁶ A simplified schematic representation has been presented by Dingemans et al.⁷ (Figure 27-2) who found 53 to 78 concentric, lamellar layers in their study of human ascending aortas.

Elastic lamellae and their interconnections are made up of elastic fibers that demonstrate “a bewildering jigsaw of molecular interactions involving fibrillin, elastin, and various microfibril-associated molecules…”.⁸ Microfibrils are structural elements found in most tissues and most abundantly associated with basement membranes and elastic fibers. An important milestone in the understanding of microfibril biology has been the discovery of fibrillin as a major glycoprotein in microfibrils.⁹ Fibrillins exist in at least three isoforms whose primary structures are dominated by calcium binding epidermal growth factor (cbEGF) domains that adopt a rodlike structure in the presence of calcium.⁸ Fibrillin molecules also have compact, globular regions that appear to be related to their capacity for extensibility.¹⁰ Elastic fiber development involves the deposition of tropoelastin, the soluble precursor of elastin, on a preformed matrix of fibrillin-rich microfibrils and, under appropriate conditions, elastin undergoes a process of ordered self-aggregation called coacervation and becomes cross-linked by lysyl oxidase.⁸ Elastic fibers are thus a composite material composed of cross-linked elastin and fibrillin-rich microfibrils and arranged in fenestrated elastic lamellae with irregular protrusions that reach toward smooth muscle cells. Fibulin-5 is an extracellular matrix protein abundantly expressed in great vessels, cardiac valves, and other elastic tissues during embryogenesis as well as in adult life and appears to have a critical function as a scaffold protein organizing links between elastic fibers and cells.¹¹ Based on studies of comparative anatomy, the incorporation of elastin with the system of microfibrils appears to be necessary to support the high pressure closed circulatory systems of higher vertebrates.¹² Recognizing the critical roles of tropoelastin, fibrillin-1, and fibulin-5 in the formation and maintenance of elastic fibers and their attachments and interactions with vascular cells, brings the promise of bioengineered vascular grafts with the mechanical characteristics of normal arteries.¹³

Elastic fiber biology is associated with more than 30 reported molecules (Table 27-1).⁴

Immunoelectron microscopy of specimens of adult human ascending aorta have revealed important observations on the relationships of collagen, fibrillin, fibronectin, elastin, and VSMCs in the aortic media.⁷ Thick collagen fibers containing types I, III, and V collagen are closely associated with elastic lamellae. Collagen fibers are most numerous close to elastic lamellae and are oriented circumferentially, roughly parallel to the main axis of the VSMCs, and to the irregular protrusions of the elastic lamellae extending into the interlamellar space. Collagen fibers more centrally in the interlamelar space are thin and more randomly oriented. Elastic lamellae are nearly straight in distended aortas but wavy in those prepared in a collapsed state. Collagen fibrils in the adventitia are on an average very thick compared with most collagen fibrils in the media, but those in the intima are thinner. Within the media, fibril diameters are strikingly variable. The shapes of the collagen fibrils on cross-section are also highly variable with most being rounded but others irregular and angular especially near the adventitia. Some of these collagen fibrils, irregularly shaped on cross section, appeared to be loose and spiraled when seen longitudinally and may be degenerated forms. Type VI collagen as well as fibrillin was found in the oxytalen fibers that connect smooth muscle cells to the elastic elements with the actual cell attachments involving fibronectin. Type IV collagen, which generally seems to be a major basement membrane component, was found in aggregates located further from cell surfaces than fibronectin.⁷

The proteoglycans dermatan sulfate, heparan sulfate, and chondroitin sulfate were also demonstrated in the work cited above in the matrix material of aortic elastic lamellae. Of the extracellular structures described only the cell-associated fibronectin positive layers, the elastin-associated material, and the oxytalan fibers were devoid of proteoglycans. Dermatan (decorin) was associated with collagen fibers and is thought to be related to the control of collagen fibrillogenesis, heparan (perlecan) was found within basal laminalike material, where it is thought to be a binding site for growth factors and cytokines, and the large interstitial chondroitin proteoglycans (versican) seem to function in maintaining shape and sustaining the compression generated by pulsatile forces.⁷

The embryologic development and homeostatic processes of the aortic media are clearly more complicated than a static view of its microstructure. VSMCs possess receptor mechanisms that allow the detection and a corresponding response to shear stress and tension as well as other cell-signaling events.¹⁴ This response involves matrix modifications such as the deposition of collagen fibers, tropoelastin, or microfibrils resulting from the activation of transcription factors and subsequent gene expression. The aortic wall is best viewed as an intricate biochemical mechanism having its origin in the developing embryo and subject to a continuum of development, growth, maintenance, and senescence. While the aorta develops as a component of the embryonic circulatory system, most of its functional life is spent under very different conditions of vessel geometry, blood pressure and, therefore, wall tension since the aorta grows and blood pressure increases steadily from the embryonic state to adulthood with further changes occurring with senescence. The corresponding microstructure changes as well since aortic mesodermal cells respond to these factors in a more or less continuous way as can be seen from bioengineering studies designed to produce substitute autologous arteries.^15,16,17 There is an active and on going cell-mediated maintenance program for structural proteins, but as with most tissues, repair functions are not unlimited especially with the loss of cellular elements.

One class of molecules known to be important in matrix development and homeostasis are the matrix metalloproteinases (MMPs) and their antagonist tissue inhibitors of MMPs (TIMPs). Remodeling is an important element of tissue growth and morphogenesis and this requires matrix degradation, which seems to be one component of MMP function. However, MMPs have a more expansive regulatory role in several extracellular matrix processes including cell migration, proliferation, and apoptosis as well as angiogenesis and wound healing.¹⁸

Mechanical Properties of the Proximal Aorta

Blood is delivered to the proximal aorta in boluses depending upon the stroke volume and pressure generated by left ventricular systole, by the intrinsic impedence of the ventricular outflow tract and aortic root, peripheral vascular resistance, and by the viscosity of the blood. This event causes a pulse wave of flow and pressure that is repeated with every heartbeat. However, when blood reaches the arterioles and capillaries; pressure is nearly constant and flow continuous. Much of the attenuation of the amplitude of flow and pressure waves occurs in the aorta. The ratio of flow pulse amplitude to mean flow decreases by more than threefold from the aortic arch to the femoral artery.¹⁹ This phenomenon is a result of the unique elastic properties of the aortic wall which allow a calibrated level of distension and elastic recoil with each ventricular systole and thereby produce a diastolic counterpulsation qualitatively similar to that produced by the intra-aortic balloon pump used to augment the hemodynamics of a failing heart (Figure 27-3). Shadwick²⁰ points out that this pulse amplitude attenuation may serve to minimize the hydraulic power requirements of the heart by lowering the energy cost of accelerating and decelerating the blood mass in each cycle and to minimize the drag force imposed by flow on the endothelial cells lining the small vessels. There is, however, some energy lost in this process of elastic recoil. When an artery is subjected to inflation-deflation cycles, viscoelasticity, in distinction to pure elasticity, will cause the pressure-volume curve for deflation to fall below that for inflation, forming a hysteresis loop. The area enclosed in this loop represents the 15% to 20% of total strain energy input lost with each systole. Together with blood viscosity the viscoelasticity of the artery wall prevents reflected pressure waves from resonating in the arterial system.²⁰

A conceptual model of the relationships of arterial pressure and flow, termed the Windkessel (“air chamber”) was developed by Otto Frank.²¹ The model considers the circulatory system as a closed hydraulic circuit consisting of a pump connected to a chamber partially filled with incompressible water but containing a pocket of compressible air to simulate aortic compliance and elastic recoil. Thus these properties of the aorta are sometimes called Windkessel properties. The model has undergone many modifications to allow it to accurately predict measurable circulatory parameters. A four element Windkessel has been presented with elements corresponding to physiologic concepts that produces waveforms with excellent fidelity to those observed in vivo²² (Figure 27-4).

Elastic behavior is an intrinsic property of materials under conditions of loading. In the least complicated situation, that of a simple spring, the deformation (strain), or change in length of the spring relative to its original length, is directly proportional to the load placed upon it (stress). This simple relationship is linear, (E = stress/strain, where E is a constant) and the deformation is purely elastic with the material returning to its exact original state only under ideal (theoretical) conditions. Deformations that result in a permanent change in the material are termed plastic while those that result in hysteresis or the loss of energy with return toward the original configuration are termed viscoelastic. Nevertheless, materials are characterized by the modulus of elasticity (E) calculated from the stress-strain relationship, which is not a simple linear one in the case of viscoelastic, multicomponent materials. The higher this value, the stiffer the material and conversely the lower the number, the more compliant the material is. Materials demonstrating elastic behavior also have a yield point where excess strain results in plastic deformation and a point, the tensile strength, which defines the limits of structural integrity. Figure 27-5 demonstrates stress-strain curves for descending thoracic aortic strips. At low stress levels elastic behavior is dominated by the mechanical properties of elastin and at high levels near the yield point by the properties of collagen.

The aortic wall can be considered as a cylinder composed of microscopic springs that are oriented longitudinally, circumferentially, and radially within the thickness of its wall. When the aorta is pressurized, the circumferential and longitudinal springs are stretched and the radial springs are compressed. The aorta tends to dilate and/or elongate, and its wall thins. The moduli of elasticity for these different sets of “springs” are not necessarily the same. If they were, the material would be termed isotropic. If the material displays different degrees of stiffness depending on the orientation of loading it is termed anisotropic. Since the aorta has a complex composite structure when its properties are viewed as a whole it is anisotropic. To simplify analysis, the volume of tissue comprising the aortic wall is often considered to be constant, that is, the wall is incompressible and there are proportionate changes in thickness, length, and diameter. The precise form of these proportionate changes in strain, occurring in a given two dimensions, is an intrinsic material property known as Poisson’s ratio. For anisotropic materials this ratio will differ, as does the modulus of elasticity depending upon the direction of loading. Further, the stress (tension) on the innermost layers of “springs” may not necessarily be the same as on those in the outer layers and this appears to be the case in arteries.²³ It has been shown that there exists residual strain with different levels of tension in the inner and outer layers of the aortic wall when no pressure load distends the aorta, as in a beam of prestressed concrete.²⁴ This phenomenon serves to equalize wall tension across the thickness of the aortic wall during pressure loading and has been quantified by calculating the “opening angle.” That is, the angle created between the two end points and the midpoint of the circumference of a cylindrical segment of aortic tissue cut open longitudinally (which may actually be greater than 180 degrees with the aorta turning itself inside out)²⁵ (Figure 27-6).

Using a two-dimensional analysis considering average longitudinal (axial) and circumferential stress and strain,²⁶ the average circumferential Kirchhoff wall stress (τ_c) can be given by the relation:

where P is the luminal pressure, r_i the inner luminal radius, h the vessel thickness, and λ_c the circumferential stretch ratio (the ratio of the circumference at load to that at no load which is different for the outer and inner layers and this difference becomes larger the thicker the cylinder wall). Similarly, the average Kirchhoff longitudinal stress (τ_z) can be given by

where r_o is the outer luminal radius and λ_z is the longitudinal stretch ratio.

It can be seen, therefore, that the average circumferential or longitudinal wall stress is directly proportional to the pressure in the vessel and its radius and is inversely proportional to the thickness of the vessel wall. For infinitesimal deformations the stretch ratio terms become insigificant, approaching one. This analysis expands the familiar “law of Laplace” which states that wall tension is proportional to the product of pressure and radius and underscores the relentless rise in wall tension that accompanies the growth of aneurysms. The wall thickness component is also important since one can see the purely geometric component in aortic weakening that is associated with wall thinning from destructive processes. Additionally, it can be seen how small vessels can be pressurized to the same extent as the aorta and yet be thin-walled. At physiologic pressures, in normal arteries, it has been empirically determined that the ratio of internal radius to wall thickness is between 7:1 and 10:1²⁷ so that circumferential wall tension is approximately 10 times the blood pressure. If an aneurysm doubles the aortic radius, the corresponding wall tension is 20 times the blood pressure if the wall thickness is maintained and even higher if the aneurysm is thin walled.

Cardiac dynamics have been shown to increase longitudinal (axial) ascending aortic stress independent of aortic distending pressure. It is known from a careful analysis of cine aortograms that there exists axial downward displacement of the aortic root during systole and an associated twisting deformation as well.²⁸ The degree of axial displacement significantly increases local stress in the region above the STJ. Subjects with aortic insufficiency demonstrate significantly greater axial displacement and those with poor contractile function significantly less than the norm.²⁸ In addition to the effect of vigorous cardiac contraction on the strain induced in the proximal aorta, it has been shown, based on a finite element analysis, that blood flow asymmetry through the aortic root results in 21% and 10% greater wall stress in the right and noncoronary sinuses, respectively, as compared to the left.²⁹ Therefore, the convexity or greater curvature side of the proximal aorta is subject to higher regional stresses no doubt affecting the propensity for stress-induced damage.

The third dimension of aortic wall mechanics is also important. While longitudinal and axial stresses are tensile during loading and unloading of the aorta during each cardiac cycle, radial stresses may actually change from compressive during systole to tensile during diastole depending upon the residual strain in the outer layers of the media. That is, as the aorta is depressurized, the differential strain energies resident in different layers of the aorta will tend to pull the layers apart, especially if the magnitude of this differential stress exceeds the blood pressure at that point. This has important implications for the development of intramural hematoma and AD. Further, in vivo, there are additional stress and strain components related to the shear forces exerted on intimal surfaces by flowing blood and at curves and branch points. These shearing forces are of no consequence when considering the mechanical properties of isolated aortic segments in a physiology laboratory but may be of considerable importance in the local, microscopic analysis of aortic function and pathophysiology.

Creep is a materials property that is relevant to most mechanical systems. Materials with persistent or repeated loading to levels below the yield point exhibit small but cumulative plastic deformation over time. Collagen,³⁰ elastin, and other structural elements of the aortic wall are subject to this process. Therefore, the aorta dilates over time as is clearly seen by the one millimeter per decade increase in aortic size that accompanies aging.³ Fatigue is yet another relevant material process by which catastrophic failure under conditions of repeated loading can occur at loads well below the yield point. Considering that, at an average heart rate of only 60, the aorta will be subjected to 31 536 000 loading cycles per year, fatigue characteristics of its constituent elements are obviously of major importance.

Micromechanical Model of the Aortic Wall

The aorta functions to provide a predefined degree of distension when pressurized, with a limit at peak systole, and then elastic recoil as has been described. It has been difficult to purify and test elastin without contaminating microfibrils; however, recent studies reveal a linear stress-strain response with equal stiffness in circumferential and axial directions. Isolated elastin appears to have an elastic modulus of approximately 80 kPa,³¹ whereas the elastic modulus of fibrillin-rich microfibrils is approximately 80 MPa or three orders of magnitude greater.³² Collagen fibers are even stiffer with an elastic modulus between 300 and 2500 MPa and maximum strains of only 3% to 4%²⁷ and provide the outer limit of vessel wall distension.

Since stress-strain curves (see Figure 27-5) demonstrate a smooth transition between the elastic compliance of elastin and the stiffness of collagen, there must exist a mechanism to dampen this transition. In the classic Maxwell model of viscoelasticity, the elastic component is represented by a spring and the transition damper by a dashpot (analogous to the device attached to a spring-loaded door to keep it from slamming shut). The precise microstructural mechanism for this coupling remains to be defined but a model can be proposed from available research. A modified Maxwell model³³ has been described using the disconnecting hook element of Wiederhielm³⁴ to describe the relationship of smooth muscle cells, elastin, and collagen in human brachial arteries. This model has been criticized on the basis of additional engineering measurements made on collagenase digested specimens by Van Bavel et al. in rat mesenteric arteries and a serial element model proposed³⁵ (Figure 27-7 and Figure 27-8). This serial element model may better explain the collagen data since excessive strains exceeding the 3% to 4% observed for collagen fibers may be required for the hook model. In electron micrographs, collagen exists in substantial circumferential bundles and pericellular interlacing fibrils.³⁶ Sokolis and collegues have described a three part constitutive mathematical expression for the nonlinear stress-strain relationship of the aortic wall represented graphically in Figure 27-5,³⁷ the first part corresponding to low stresses up to a uniaxial load of 40 kPa (which would roughly correspond to an intraluminal pressure of 30 torr), the second part up to 120 kPa (intraluminal pressure roughly 90 torr), and the third part for higher stresses. Attempting to correlate the three part stress-strain observation with aortic microstructure, micrographs of rabbit descending aorta were fixed at various stress levels from each of the three parts and demonstrated progressive straightening of wavy elastic lamellae through part I, then extension and compaction in parts II and III. Collagen remained corrugated through part I but tended to reorient toward the direction of stress. During stressing into part II, the collagen fibers continued to reorient and some began to uncoil. Well into part III the collagen fibers were almost straight and densely packed (Figure 27-9). In this study of uniaxial, longitudinal loading, circumferentially oriented elastic fibers, and collagen bundles did not change from their unstressed wavy, or tangled appearance suggesting that these circumferential elements are not tensioned with orthogonal uniaxial loading and are not firmly attached to the longitudinal elements. This feature allows for differential biaxial distensibility since orthogonal elements must slide over one another, yet also facilitates separation of layers under radial tension (as in dissection or intramural hematoma). Silver et al.³⁸ have proposed that at least some of the collagen fibrils in the media are in series with smooth muscle cells and this collagen-smooth muscle cell network is in parallel with parallel networks of collagen and elastic tissue, as in the model of VanBavel et al.

Smooth muscle cells in the aortic media have been termed spannmuskeln and those in muscular distributing arteries ringmuskeln by Benninghof (Figure 27-10)²⁷ with the latter arrangement functioning to regulate arterial caliber and the former configuration, unique to the aorta, and large elastic arteries, providing for the maintenance of aortic distensibility. As described above, proximal aortic spannmuskeln are suspended in the interlamellar space (see Figure 27-2), likely under tension, given that their long axis of orientation is circumferential, parallel to the orientation of the major collagen fibers. It appears that oxytalen fibers containing fibrillin microfibrils as well as type VI collagen are responsible for suspending smooth muscle cells under tension along with elastic lamellar extensions. That these VSMC-elastic lamellar connections are likely of considerable mechanical strength has been demonstrated by Dingemans et al. in an ultrastructural study of severely pathologic areas of aorta where these connections often remained intact.³⁹ Since elastic lamellae are wavy in the unstressed state and the waviness resolves under tension, it appears that the fibrillin-rich microfibrils within elastic fiber extensions and oxytalen fibers become tensioned along with VSMCs, which shift from a radial-tilt along waves of elastic lamellae to a more nearly straight circumferential orientation. Fibrillin-rich microfibrils are much stiffer than elastin, approaching the lower limit of collagen stiffness but are, to some degree, extensible.¹⁰ Further, microfibrils are incorporated into elastic fiber structure perhaps in a parallel configuration with elastin within microelements and creating a serial microelement arrangement as proposed by van Bavel et al. and Silver et al. with the elastin-fibrillin-smooth muscle cell interaction acting as the “dashpot” at strains below those at which collagen fibers are engaged. Once elastic lamellar deformation has occurred, smooth muscle cells and microfibrils are tensioned, and, after some small additional strain, given the high elastic modulus of microfibrils, the much tougher, stiff collagen fibers are tensioned preventing further strain on all elastic elements and acting like a door stop. This arrangement would allow for smooth muscle cell signaling in response to tension and strain both for elastic and nonextensible components as well as explaining the three part wall tension-radius curves. Since fibrillin with its EGF domains is intimately involved in this mechanical cell signaling mechanism, interference with the fibrillin-smooth muscle cell relationship could have important pathophysiologic implications for the maintenance not only of microfibrils but for collagen and elastic fibers as well.