T annulus of the arterial valves and with accounts of. The Myth of the Aortic Annulus: The Anatomy of the Subaortic Outflow Tract

Transcription

1 11. Management of the Small Aortic Root The Myth of the Aortic Annulus: The Anatomy of the Subaortic Outflow Tract Robert H. Anderson, MD, FRCPath, William A. Devine, BS, Siew Yen Ho, PhD, Audrey Smith, PhD, and Roxane McKay, MD, FRCS Department of Paediatrics, National Heart and Lung Institute, London, United Kingdom; Department of Pathology, Children s Hospital of Pittsburgh, Pittsburgh, Pennsylvania; and Department of Surgery, Institute of Child Health, University of Liverpool, Liverpool, United Kingdom Surgical repair of the small aortic root is limited in part by the very structure of the outflow tract from the left ventricle. The root is not constructed on the basis of a ringlike annulus supporting the leaflets of the aortic valve. The only truly circular structure within the outflow tract is the junction of the aortic wall with the underlying ventricular structures, themselves partly muscular and partly fibrous. This circular ventriculoarterial junction is crossed by the semilunar attachments of the leaflets of the aortic valve, producing an interlinking arrangement between the expanded aortic sinuses and three triangles of fibrous tissue placed beneath the apexes of the commissures between the valve leaflets. The triangles form extensions of the left ventricle that are related, in part, to the pericardial cavity surrounding the heart. The arrangements of the attachment of the leaflets in malformed valves with two (or only one) effective leaflets are highly abnormal, although these valves are usually produced on the template of three aortic sinuses. The valve with two leaflets rarely gives problems during childhood. In valves producing critical stenosis, there is usually only one effective leaflet, a condition due to incomplete liberation of two of the anticipated three commissures. Detailed study shows that, in these malformed hearts, the attachment of the leaflets is much more annular than in normal valves, with inadequate formation of the fibrous triangles. (Ann Thoruc Surg 1991;52:640-6) he surgical literature abounds with descriptions of the T annulus of the arterial valves and with accounts of the optimal way of enlarging this purported structure to treat patients with stenotic and obstructive subarterial outflow tracts. All those who operate on hearts with malformed arterial valves are well aware that, in reality, the attachments of the leaflets are not arranged in ringlike fashion around the circumference of the ventricular outlets. Indeed, if the bases of the leaflets were attached in such annular formation, it would be impossible for the leaflets to open properly during ventricular systole. Instead, they are attached in semilunar fashion. Is it necessary that these semilunar attachments be described in terms of an annulus? Those who pay scant attention to matters of style and description are likely to argue that it matters little whether there is or is not a true annulus as long as those who operate on the malformed hearts understand the anatomy of the structures on which they work. We take a different stance because we believe it important to use words that reflect accurately the features of the structures they describe. This is particularly true for those who are beginning in the field and are Presented in part at the Current Controversies and Techniques in Congenital Heart Surgery Meeting, Baltimore, MD, Sep b9, Address reprint requests to Prof Anderson, Department of Paediatrics, National Heart and Lung Institute, Dovehouse St, London SW3 6LY, United Kingdom. attempting to understand the complex morphology of normal and abnormal arterial valves. In this review, therefore, we will present our understanding of the arrangement of the arterial valves, which is founded on the concept that the leaflets are attached in semilunar fashion rather than attached to a ringlike annulus [l]. We are well aware that the anatomy to be emphasized is already well recognized. The semilunar arrangement of the leaflets is described in current textbooks of anatomy [2] and is depicted and described as a tricorn (a hat with three peaks) in a popular textbook of surgery [3], where appropriate precedence is given to the important work of Zimmerman [4]. Accurate descriptions of the histology of the semilunar attachments of the leaflets of the arterial valves were also provided long ago by Gross and Kugel [5] and by Lewis and Grant [6]. Nevertheless, the cardiological world at large continues to describe an annulus within the outflow tracts of the ventricles. Our review is an attempt to put such terminology into the context of the observed anatomy of the leaflets of the normal valves and to show how their normal semilunar arrangement is distorted when the valves are congenitally malformed. Material and Methods The material on which our review is based comprises the large number of normal hearts that we have seen over the by The Society of Thoracic Surgeons /91/$3.50

2 Ann Thorac Surg 1991;52:64M CONGENITAL HEART ANDERSON ET AL 641 Circlet supporting Fig 1. Long-axis section through the outflow tract of the left ventricle showing the cylindrical section that must be removed so as to take out the entire aortic valve. Within this circlet, which could be slipped on the finger like a ring, the leaflets of the valve are attached in semilunar fashion (see Fig 7). Semilunar attachment of leaflets within "ring" years and the abnormal hearts with valvar aortic stenosis in the cardiopathological collections of our institutes. None of the normal hearts have been formally analyzed, but some have been sectioned histologically to determine the mode of attachments of the valve leaflets and to ascertain their connections to the fibrous skeleton of the heart. We will make reference to two more detailed and formalized studies of hearts with malformed valves. One [7] has been published, and one is in preparation (McKay R, Smith A, Leung MI', Arnold R, Anderson RH. Morphology of critical aortic stenosis in infancy with reference to haemodynamic function and clinical management. Unpublished observations). The published account was devoted to the deformed aortic valve having two leaflets. This material was analyzed by Angelini primarily, we being part of her collaborative team. The other series involved 21 hearts with "critical aortic stenosis" seen in infancy. They have been analyzed in depth by ourselves and colleagues at the Institute of Child Health, University of Liverpool. This material has, as yet, not been published and, therefore, will be considered only in the broadest terms as they apply to the concepts to be discussed. Glossary and Etymology It should be clear from our introduction that part of the problem of providing an adequate description of the anatomy of the subarterial outflow tracts is an inappropriate use of words. It is important, therefore, to establish the precise meaning of the words used in our descriptions, so those who disagree will have clear definitions on which to base their dissent. ANNULUS. According to the The Shorter Oxford English Dictionary [8], annulus is the diminutive form of the Latin word anus, the latter meaning a ring. Although rings as worn nowadays can take various forms, the same dictio- nary defines this word as a circle or circlet as carried on the finger. This is how we will use the word. SEMILUNAR ATTACHMENTS. If the area of outflow tract supporting the leaflets of an arterial valve is removed in its entirety, it takes the form of a ring that could be slipped on the finger (Fig 1). Within this circlet, however, the valve leaflets themselves are attached in semilunar fashion, with three such half-moons being present in the normal valve. The attachment of each leaflet rises to its zeniths at the peripheral attachments of the commissures and is at its nadir at the base of the leaflet. CUSP. Although used by most pathologists and surgeons as synonymous with the word leaflet, the word cusp means a projection or point [9] and is used appropriately by dentists when describing elevations on the surfaces of the teeth. COMMISSURE. This word means a seam or the place where two bodies touch or unite [ 101. When applied to an arterial valve, therefore, the commissure should, logically, describe the line from the periphery to the center of the valve orifice over which the surfaces of adjacent leaflets coapt (Fig 2). Often, however, pathologists describe only the peripheral part of this junction in terms of the commissure. VENTRICULOARTERIAL JUNCTION. Perhaps the most important point requiring emphasis for the understanding of the anatomy of arterial valves is that the semilunar attachments of the leaflets (see Fig 1) produce a discrepancy between the anatomical and physiological ventriculoarterial junctions. The anatomical junction, made between the ventricle and the arterial trunk it supports, takes the form of a circle over which the fibrous wall of the trunk is

3 642 CONGENITAL HEART ANDERSON ET AL Ann Thorac Surg 1991;52:640-6 attached to the supporting ventricular structures. In the left ventricle, these supporting structures are in part muscular and, because of the fibrous continuity between the leaflets of the aortic and mitral valves, in part fibrous (Fig 3). The physiological junction, in contrast, is marked by the semilunar locus of attachment of the leaflets. The entire area beneath this locus is, in terms of hemodynamic pressures, part of the left ventricle. The area distal to the locus is, in the same hemodynamic terms, part of the aortic sinus. The arrangement of the attachments, therefore, results in three segments of ventricular wall becoming part of the aortic sinus and three triangular areas of arterial wall being incorporated into the outflow tract of the ventricle (Fig 4). TRIANGLES VERSUS TRIGONES. To understand the complex arrangement of the aortic root, it is also necessary to emphasize the relationship of the semilunar attachment of the leaflets to the fibrous skeleton of the heart. The skeleton is often displayed as an extensive structure enclosing the attachments of all four cardiac valves [ll]. Dissections as well as histological studies [6, 121 show this not to be the case. In the aortic root, the fibrous skeleton forms the posterior wall of the outflow tract in the area Fig 3. View of the opened outflow tract of the left ventricle showing how the leaflets of the aortic valve are supported in part by the musculature of the left ventricle and in part by fibrous continuity with the aortic leaflet of the mitral valve. where the leaflets of the aortic valve are in fibrous continuity with those of the mitral valve (Fig 5; see Fig 3). The two ends of this area of fibrous continuity are thickened to form the left and right fibrous trigones, respectively. These trigones are the best-formed parts of the skeleton. The right fibrous trigone itself is continuous with the membranous part of the septum, and these two structures together form the central fibrous body through which the axis of atrioventricular conduction tissue penetrates to pass from the muscular atrioventricular septum to the crest of the ventricular septum (see Fig 5). Sections can be taken from the base of the leaflets of the aortic valve where they are attached to the muscular walls of the left ventricle to give the spurious impression that a cordlike collagenous ring does provide support for them. Serial sections, however, show that this cordlike structure is also arranged in semilunar fashion. It becomes incorporated in the substance of the arterial wall as the attachments of the leaflets rise to the peripheral extent of the commissures (Fig 6). Indeed, beneath the apex of the b) TI commissures, only a relatively thin layer of the arterial wall separates the inside of the left ventricle from the Fig 2. (a) Photograph and (b) line interpretation of aortic value as extracardiac space. These thinner areas of fibrous wall can seen from above with the leaflets in closed position. Note how the lines of coaptation, or commissures, extend from the periphery to the center accurately be described as the interleaflet (see Fig of the valve orifice (open and closed stars). 4) [71.

4 Ann Thorac Surg 1991;52:64C-6 CONGENITAL HEART ANDERSON ET AL 643 Hernodynarnic ventriculoaortic junction lar junction Fig 4. Discrepancy between the anatomical and hemodynamic ventriculoarterial junctions. The discrepancy is due to the semilunar attachments of the leaflets. Anatomic ventriculoaortic junction Interleaflet triangle - arterial wall as ventricle Ventricle within base of sinus Results The Normal Aortic Root As described in our glossary, the area of the outflow tract of the left ventricle supporting the leaflets of the aortic valve takes the form of a cylinder (see Fig 1) in which the leaflets are attached in the fashion of three half-moons. This can best be demonstrated by spreading out the cylinder so that its surface is seen as a flat strip (Fig 7a). Removing the leaflets from their attachment demonstrates the discrepancy between the anatomical and physiological ventriculoarterial junctions (Fig 7b). When seen in the intact heart, either from below or above, there is an interlinking of the sinuses of the aorta and the supporting structures of the ventricle. The three interleaflet triangles extending beneath the apices of the commissures and formed by the relatively thin wall of the aorta are incorporated as extensions of the ventricular outflow tract. These three extensions "point" toward the transverse sinus of the pericardium, toward the roof of the membranous septum, and toward the infundibulum of the right ventricle, respectively [7]. With this arrangement of the outflow tract, the leaflets of the arterial valve, which have shorter free edges than attached margins, are free to fall back into the aortic sinuses during ventricular systole to ensure unobstructed flow of blood into the aorta. During diastole, the leaflets are forced together to coapt along a line away from their free edge. These areas of coaptation, or commissures, extend from the periphery to the center of the valve orifice (see Fig 2). As discussed, the entire cylinder forming the root and supporting the valve leaflets can be considered to represent a ring. The only circular structure within the overall ring, however, is the anatomical ventriculoarterial junction. The Aortic Valve With Two Leaflets Careful examination of these valves shows that, in most instances, there is evidence of fusion of two leaflets in a Membranous septum Central fibrous body Fig 5. Relationships of the leaflets of the aortic valve to the components of the fibrous skeleton and the axis of atrioventricular conduction tissue. Left fibrous trigone

5 644 CONGENITAL HEART ANDERSON ET AL Ann Thorac Surg 1991;52:64&6 Fig 6. Histological sections showing how a fibrous cord (a) reinforces attachment of the valve leaflet to the uentuiciilnr musculature at the base of attnchnient hit (b) becomes incorporated in the u dl of the nouta at the lezd of the peripheral attachment of the commissii yes. valve that initially had three leaflets (Fig 8a). Study of the ventricular outflow tract also supports the concept of formation of one vestigial commissure. In the majority of cases [7], there is evidence of three interleaflet triangles, although one of the triangles is rudimentary and is located between the raphe representing the fused commissure in the conjoint leaflet. In a small minority of hearts, however, the valve with two leaflets also possesses only two aortic sinuses, with no evidence of either a raphe in one of the leaflets or a vestigial third triangle within the outflow tract (Fig 8b). Irrespective of whether the valve with two effective leaflets was derived from a template with two or three leaflets initially, it is most unusual for these structures to be stenotic or incompetent in early life. As Edwards [13] pointed out, it is the onset of acquired disease, usually producing gradual fusion of the commissures from the periphery toward the center, that produces problems in valvar function. The Critically Stenotic Aortic Valve of lnfancy Analysis of a series of such valves from the collection of the Institute of Child Health at the University of Liverpool (McKay R, Smith A, Leung MI', Arnold R, Anderson RH. Morphology of critical aortic stenosis in infancy with reference to haemodynamic function and clinical management. Unpublished observations) shows that the entire pattern of formation of the leaflets is distorted compared with normal. Almost all of the valves show a relatively circular arrangement of attachment of an effectively single leaflet, with full development of only one commissure and one deep interleaflet triangle (Fig 9). Within the solitary leaflet, there is usually evidence of initial (or attempted) formation of additional commissures, a vesti- Fig 7. In these preparations of the aortic root, the circlet shown in Figure 2 has been spread to show its full circumference (a) and then the leaflets of the valve have been removed (b).

6 Ann Thorac Surg 1991;52:64&6 CONGENITAL HEART ANDERSON ET AL 645 Fig 8. Aortic valves having two leaflets but (a) with one conjoint leaflet and three sinuses and (b) with two leaflets and two sinuses. gial remnant of an interleaflet triangle being seen on the ventricular aspect. The putative commissures, however, are represented only as folds in the surface of the leaflet and the aortic aspect of the wall of the sinus, dividing the aortic root into three potential sinuses (Fig 10). Only one of these structures is related to a properly formed interleaflet triangle and thus has the potential to function as a true commissure. Usually this commissure extends to the area of fibrous continuity with the mitral valve. It should be emphasized that the folds are on the aortic aspect of the attachment of the leaflet and correspond to raphes within the leaflet. Incision of the leaflet limited to the length of the raphes, however, would extend only to its abnormal ventricular attachment. Such cuts would be unlikely to liberate the leaflet to permit it to open in unobstructed fashion. Comment Although the concept of an annulus supporting the valve leaflets is ingrained in the surgical vocabulary and un- likely to be displaced for some time, if ever, it seems to us that its acceptance provides a poor basis for the understanding of normal and abnormal valvar function. It is paradoxical that, among the valves we have examined, those with an almost annular attachment of their leaflets are the most stenotic and produce problems the earliest during life. The understanding of the function of the normal arterial valve must be based on the knowledge that its leaflets are arranged in semilunar fashion and that the commissures between the leaflets function properly only when their peripheral attachment to the arterial wall is well above the basal attachment of the leaflets they separate. Study of the arterial root in this fashion also emphasizes the important discrepancy to be found between the anatomical and hemodynamic ventriculoarterial junctions. This, in turn, highlights the relationships of the triangles found immediately beneath the apexes of the commissures. These areas are formed by the relatively thin wall of the arterial trunk between the arterial sinuses and can be the site of aneurysmal protrusions from the outflow tract [14]. In short, the understanding of normal

7 646 CONGENITAL HEART ANDERSON ET AL Ann Thorac Surg 1991;52:64M Fig 9. Formation of the so-called unicuspid and u n icomm iss u ral valve seen most frequently when aortic stenosis is present in infants. The valve has three potential commissures, but two are vestigial, being marked by raphes in the leaflet and by rudimentary triangles on their ventricular aspect. and abnormal arterial valves is enhanced by the appreciation of the semilunar attachment of the leaflets. Such knowledge would, in our opinion, be advanced by the abandonment of the myth of the valve annulus. A B // Fold in wall Apices of commissures at sinotubular junction NORMAL CRITICAL STENOSIS true commissure Fig 10. Fundamental differences in arrangement of (a) the normal valve leapets and (b) the pattern seen in most of the hearts with critical aortic stenosis seen during infancy in the Liverpool study. References 1. Anderson RH. Editorial note: the anatomy of arterial valvar stenosis. Int J Cardiol 1990;26: Williams PL, Warwick R, Dyson M, Bannister LH. Gray s anatomy. 37th ed. Edinburgh: Churchill Livingstone, 1989; Kirklin JW, Barratt-Boyes BG. Cardiac surgery. Morphology, diagnostic criteria, natural history, techniques, results, and indications. New York: John Wiley, Zimmerman J. The functional and surgical anatomy of the heart. Ann R Coll Surg Engl 1966;39: Gross L, Kugel MA. Topographic anatomy and histology of the valves in the human heart. Am J Pathol 1931;7: Lewis T, Grant RT. Observations relating to subacute infective endocarditis. Heart 1923;10: Angelini A, Ho SY, Anderson RH, et al. The morphology of the normal aortic valve as compared with the aortic valve having two leaflets. J Thorac Cardiovasc Surg 1989;98: Little W, Fowler HW, Coulson J. In: Onions CT, ed. The shorter Oxford English dictionary on historical principles. 3rd ed. Oxford: Clarendon, 1973: Little W, Fowler HW, Coulson J. In: Onions CT, ed. The shorter Oxford English dictionary on historical principles. 3rd ed. Oxford: Clarendon, 1973: Little W, Fowler HW, Coulson J. In: Onions CT, ed. The shorter Oxford English dictionary on historical principles. 3rd ed. Oxford: Clarendon, 1973: Lockhart RD, Hamilton GF, Fyfe FW. Anatomy of the human body. London: Faber and Faber, 1959: Anderson RH, Becker AE. Cardiac anatomy-an integrated text and colour atlas. London and Edinburgh: Gower- Churchill Livingstone, 1980: Edwards JE. The congenital bicuspid aortic valve. Circulation 1961;23: McAlpine WA. Heart and coronary arteries. An anatomical atlas for clinical diagnosis, radiological investigation and surgical treatment. New York: Springer-Verlag, 1975:1549.