Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Heart Valve Dis. Author manuscript; available in PMC 2010 May 4.
Published in final edited form as:
J Heart Valve Dis. 2009 September; 18(5): 488–495.
PMCID: PMC2863300

Characterization of Mitral Valve Anterior Leaflet Perfusion Patterns


Background and aim of the study

Although previous histologic studies have demonstrated the presence of blood vessels in the anterior mitral leaflet (AML) and second-order chordae (SC), little is known of the pattern of leaflet perfusion. Henhe, the pattern and source of AML perfusion was investigated in an ovine model.


Fluorescein angiograms were obtained in 17 ovine hearts immediately after heparinization and cardioplegic arrest, using non-selective left coronary artery (LCA) and selective left anterior descending (LAD), proximal, mid- and distal left circumflex (LCx) perfusion. Serial photographs using a flash/filter system to optimize fluorescence were obtained through a left atriotomy.


The proximal half of the AML was seen to be richly vascularized. A loop of vessels was consistently observed in the mitral annulus and AML; these vessels ran along the annulus, extended to the sites of SC insertion, and created anastomoses between these insertions. The SC contributed to the AML perfusion and the anastomotic loop. Selective perfusion of the LAD or proximal LCx artery (ligated before the first obtuse marginal artery) did not perfuse the AML (n = 6). Perfusion of the mid- and distal LCx (n = 7) consistently supplied the AML via SC insertion sites and annular branches.


The ovine AML is perfused by vessels that run through the SC and annulus simultaneously, and then create a communicating arcade in the leaflet. These vessels originate from the mid- and distal portions of the LCx. A loss of perfusion as a result of microvascular disease could have adverse implications. Derangements in the extensive vascular component of the mitral valve could be an important contributing factor to valve disease.

Mitral valves are traditionally thought to be avascular (1,2), and recent publications have proposed that the presence of blood vessels in the valves is pathologic (3-7). In 1852, Luschka first reported findings that demonstrated the presence of blood vessels within the leaflets of normal human heart valves (8). Luschka identified a network of vessels in the broad portion of the human mitral valve, with small vessels running to the line of coaptation and a few vessels in the chordae tendineae (8,9). This incited a century of debate regarding the presence of vessels in mature, normal human heart valves (10). The presence of leaflet vessels has now been confirmed across multiple species, including humans, most recently in side-notes to histologic studies in which immunohistochemical stains confirmed their presence within the mitral valve and chordae tendineae (8-15).

Although the presence of heart valve vessels has been demonstrated, the pattern of perfusion and sources of inflow remain unknown. Hence, the study aim was to characterize the perfusion patterns of the anterior mitral leaflet in an ovine model.

Materials and methods

To investigate the pattern of mitral valve blood vessels and their origin of inflow, fluorescein coronary angiograms were obtained in 17 adult male ovine hearts immediately after heparinization (300 IU/kg) and cardioplegic arrest (500 ml lactated Ringer’s solution with 20 mEq KC1, 25 mEq NaCO3, 100 mg lidocaine, 2.5 g mannitol). The hearts were carefully excised from the chest immediately after cardioplegic arrest. A left atriotomy was performed to expose the mitral valve using electrocautery, thereby preventing bleeding atrial edges that could leak fluorescein during subsequent infusion.

All animals received humane care in compliance with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research, and the Guide for Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institutes of Health (DHEW [NIH] publication No. 85-23, revised 1985).

The study was approved by the Stanford Medical Center Laboratory Research Animal Review Committee and conducted according to Stanford University policy.


Non-selective left coronary artery (LCA, n = 4) and selective left anterior descending (LAD, n = 2), proximal circumflex (Cx) and LAD (n = 4), mid and distal circumflex (LCx, n = 7) perfusions were used to obtain the angiograms (Fig. 1). In contrast to the human coronary circulation, ovine hearts have a left-dominant system with very early LAD/LCx bifurcation. The ovine right coronary artery is poorly developed and was not infused in the present study. Once the hearts were prepared for infusion they were positioned with retractors exposing the mitral valve and elastic vessel loops gently retracting the first-order chordae to spread out the leaflet.

Figure 1
A) Fluorescein infusion into the mid-circumflex artery distal to the first obtuse marginal. B) Position of the heart during infusion and image acquisition. The annulus subtending the anterior mitral leaflet (AML) and the mitral saddle horn (location on ...

A Nikon digital D70 SLR camera with a 60-mm macro lens (Nikon Corp., Tokyo, Japan) was used to take serial pictures of the leaflet approximately every 5s during fluorescein infusion (1 ml 25% fluorescein; HUB Pharmaceuticals LLC, Rancho Cucamonga, CA, USA) in 500 ml cardioplegia). Blue gel filters were placed over the four flash units (Nikon Close-Up Speedlight System) and a #15 yellow lens filter (The Tiffen Company, Hauppauge, NY, USA) were used to augment angiogram fluorescence. The images were viewed and saved using Adobe Photoshop CS3 Extended (Adobe Systems Inc, San Jose, CA, USA).


Given the wide variety of anatomical definitions used to describe the mitral valve complex the terms used were defined. The morphology of the ovine and human mitral complex closely resemble one another; thus, the appropriate human anatomical terms were used. Also, although sheep have a plantegrade as opposed to orthograde posture, the standard anatomical position used for human anatomy was used. The mitral annulus was the junctional zone of the discontinuous fibrous and muscular tissue that joins the left atrium and left ventricle or separates the mitral and aortic valves and anchors the hinge portion of the anterior and posterior leaflets (16). This was determined in both visual and tactile fashion, and is labeled in selected figures. The commissure was taken as the line of separation between the anterior (aortic/septal) and posterior (mural) leaflets and its terminus near the annulus at the junction between the A1 and P1 (Carpentier nomenclature) was termed the ‘anterior commissure’ and the other (A3/P3) the ‘posterior commissure’, according to common clinical surgical usage. The anterior leaflet extends from the annulus with the saddle-horn (point in the fibrous annulus corresponding to the aortic commissure between the left and non-coronary cusps) just anterior (closer to the A1/P1 junction) to its center. The posterior leaflet has three sub-leaflets (scallops, P1, 2, 3) and subtends from the muscular annulus. The papillary muscle with attachments to the leaflets most near the anterior commissure (A1/P1/annular junction) is termed ‘left superior’, and the other with attachments most near the posterior commissure (A3/P3/annular junction) is termed ‘right inferior’ (17). The chordae tendineae extending to the leaflet edge are ‘primary’ (marginal) chordae, and those inserting in the body of the leaflet are termed ‘secondary’ (strut) chordae. The location of secondary chord insertion was determined by visual inspection of the underside of the leaflet.


An abundance of blood vessels was observed supplying the normal ovine anterior mitral leaflet. Figures Figures22 and and33 illustrate typical angiograms. The annular half of the AML had a rich vascular supply via a loop of vessels that was consistently observed in the mitral annulus and AML: the vessels ran along the annulus, particularly that subtending the AML, extended to the sites of secondary chordal insertion, and created anastomoses between these insertion sites (Fig. 4A-C). Although not every infusion produced the entire anastomotic loop, portions were seen after each infusion, and no infusion demonstrated an entirely different pattern. In some cases the annular branches extending into the leaflet covered the secondary chordal insertion sites before angiograms appeared to extend from these sites, thus making it impossible to determine with certainty in these cases that the chordal insertion sites were perfusing the leaflet.

Figure 2
Examples of fluorescein angiograms. The schematic in the lower left corner of each image illustrates the vessel of interest in the angiogram and location of the image (red box) relative to the mitral annulus and valve. A) Fluorescein angiogram after circumflex ...
Figure 3
Examples of fluorescein angiograms. The schematic in the lower left corner of each image illustrates the vessel of interest in the angiogram and location of the image (red box) relative to the mitral annulus and valve, A) Series of images after circumflex ...
Figure 4
Schematic illustration of the general perfusion pattern of vessels in the anterior mitral leaflet. A) Annular branches. B) Commissure branches extending out to second-order chordal insertion sites (marked). C) Vessels emanating from the chordal insertion ...

Leaflet vessel inflow was seen to arise from the mid- and distal circumflex artery via annular and chordal branches. Second-order chordae contributed perfusion to the AML and the anastomotic loop (Figs. 2A,D and 3A,B). Selective perfusion of the LAD and proximal LCx artery (ligated before the first obtuse marginal) did not perfuse the AML (n = 6). Perfusion of the mid- and distal LCx (n = 7) consistently supplied the AML via the second-order chordal insertion sites and annular branches. Non-selective LCA perfusion (n = 4) yielded similar perfusion patterns to those obtained with mid- and distal LCx perfusion.


The results of the present study have highlighted the presence of a complex web of blood vessels within the mitral leaflet tissue. The finding that mitral leaflet vessels are fed by inflow through the chordae tendineae supports previous histologic evidence of vessels in the chordae spanning the distance between the papillary muscles and the leaflet (15). The redundant inflow from the mid- and distal LCx via both annular and papillary muscle branches, and the anastomotic loop that is formed, may indicate the significance of this vasculature, as may the historical evidence that anterior mitral leaflet vasculature persists across multiple species (11,13,14). The predominance of the vessels in the leaflet near the annulus was consistent with the presence of leaflet cardiac muscle cells in this region (11) and the greater thickness of the annular edge of the leaflet (>1 mm) (18) which substantially exceeds the ability of oxygen to diffuse from surrounding blood (diffusion limit ~0.2 mm) (19).

As the possibility of mitral valve tissue engineering nears reality, there is renewed interest in the ultrastructure of the valve. A re-examination has revealed that this is not a tissue predominated by passive connective tissue elements, but instead a tissue with contractile potential (cardiac myocytes, smooth muscle and valvular interstitial cells), nervous innervation (afferent and efferent), and metabolic activity with oxygen demands (11,13,14,18,20-26). The blood vessels in the valve are not vestigial holdovers, but are likely a necessary component. This represents a significant challenge to the development of a tissue-engineered valve that resembles a native valve.

There is evidence that vessels in the mitral leaflets may be implicated in valve disease. Montiel has shown that human mitral leaflet vessels are subject to the same pathologic changes as other vascular beds in the body (11). In patients with hypertension, the walls of leaflet vessels were hypertrophied compared to those of normotensive patients. Patients with chronic rheumatic valvular heart disease had pathologic changes in their leaflet vasculature. The loss of mitral valve vessels, for example as a result of microvascular disease, may cause leaflet ischemia which would have adverse repercussions, although this remains to be proven. As the present understanding of the contractile elements of the valve advances, so too may an understanding be required a need for a blood supply and the pathophysiologic changes associated with the loss of such a supply. This could lead to scarring or changes in the ultrastructure of the leaflet, or contribute to the changes observed in mitral valve fibroelastic deficiency seen in the elderly.

Historically, interest in the mitral valve vessels has stemmed from interest in their potential role in native valve endocarditis and the idea that, in some cases, the bacteria could originate within the valve rather than on the surface. In 1912, Rosenow demonstrated that the intravenous injection of bacteria (streptococci, pneumococci and Staphylococcus albus) into the blood-stream of rabbits resulted in subendothelial clumps of bacteria in the heart valves, and hemorrhages within the structure of the valve (27). This led Rosenow to conclude that the localization of bacteria within the valve depended on the presence of small blood vessels that were embolized by clumps of bacteria, leading to endocarditis. Currently, endocarditis is understood to be a valve-surface phenomenon at sites of abnormal flow or valve injury, where the valvular endocardium is disrupted; however, endocarditis does occur in the rare patient without factors, which suggests a predisposing endocardial disruption (1).

The discovery of anti-angiogenic proteins within the valves and chordae, in conjunction with the assertion that normal heart valves and chordae are avascular, has led some to conclude that the presence of vessels may trigger a cascade of pathologic processes that is ultimately diagnosed as heart valve disease, and that therapies that inhibit angiogenesis may prevent this pathologic cascade (4-7). If, however, it is accepted that blood vessels are a normal leaflet and chordal constituent, this complicates such a line of reasoning. Nevertheless, the loss of anti-angiogenic signals and neovascularization may still be a novel therapeutic target. Molecular targets, such as endothelial alpha-v/beta-3, that differentiate between activated versus resting endothelial cells or otherwise identify neovascularization, may be particularly useful in this regard (28).

Differences between ovine and human coronary circulation must be considered when interpreting the results of the present study. The dominance of the left coronary system in sheep and the right coronary system in humans makes it likely that the human AML will have inflow via the left and right coronary arteries (29). Similarly, there are species differences in papillary muscles’ perfusion. In sheep, both papillary muscles are perfused via obtuse marginal branches from the LCx, whereas in humans the left superior papillary muscle may receive all or part of its perfusion via diagonal branches emanating from the LAD, while the right inferior papillary muscle may receive all or part of its perfusion via branches from the right coronary system (30). Regardless of these differences, based on previous dye and histologic studies that have shown vessels in the chordae tendineae in multiple species (including humans), it is likely that the AML inflow via chordal vessels is observed in sheep and in humans, although this remains to be shown conclusively. Although the posterior mitral leaflet was not studied in the present study, the present methodology might be applied to it in the future.

In conclusion, the idea that mitral valve leaflets are commonly regarded as avascular in medical texts and publications - and thus by most physicians - suggests that the possibility of leaflet vascular pathology is rarely, if ever, considered. Derangements in the extensive vascular component of the mitral valve could well be an important contributing factor to valve disease.


The authors greatly appreciate the technical assistance of Sigurd Hartnett BA and Paul Chang BS, and the editorial assistance of Janet Yoe. The authors had full access to the data and take responsibility for its integrity. All authors have read and agree to the manuscript as written. These studies were supported in part by grants HL-29589 and HL-67025 from the National Heart, Lung, and Blood Institute. Dr. Swanson was supported by a Postdoctoral Fellowship Grant from the American Heart Association Western States affiliate. Dr. Bothe was supported by Research Grant S/06/07 from the Deutsche Herzstiftung, Frankfurt, Germany. Drs. Swanson, Bothe, and Itoh were Carl and Leah McCormell Cardiovascular Surgical Research Fellows.


Presented as an abstract at the American Heart Association Scientific Session, November 2008, New Orleans, LA, USA


Dr. Miller is a consultant for Medtronic Heart Valve Division, Inc., and previously consulted for Edwards Lifesciences, LLP and Boston Scientific Corp. He is a member of the Executive Committee of the Edwards PARTNER trial comparing percutaneous AVR methods with conventional AVR, and has spoken at the St. Jude Medical CV Surgical Residents’ Symposium.


1. Kumar V, Abbas AK, Fausto N. Robbins & Cotran Pathologic Basis of Disease. 7th ed. Elsevier Science; Amsterdam, The Netherlands: 2004.
2. Millington-Sanders C, Meir A, Lawrence L, Stolinski C. Structure of chordae tendineae in the left ventricle of the human heart. J Anat. 1998;192:573–581. [PubMed]
3. Schoen F. Evolving concepts of cardiac valve dynamics: The continuum of development, functional structure, pathobiology, and tissue engineering. Circ J Am Heart Assoc. 2008;118:1864–1880. [PubMed]
4. Kimura N, Shukunami C, Hakuno D, et al. Local tenomodulin absence, angiogenesis, and matrix metalloproteinase activation are associated with the rupture of the chordae tendineae cordis. Circ J Am Heart Assoc. 2008;118:1737–1747. [PubMed]
5. Yoshioka M, Yuasa S, Matsumura K, et al. Chondromodulin-I maintains cardiac valvular function by preventing angiogenesis. Nat Med. 2006;12(10):1151–1159. [PubMed]
6. Yoshioka J, Lee RT. Vascularization as a potential enemy in valvular heart disease. Circ J Am Heart Assoc. 2008;118:1694–1696. [PMC free article] [PubMed]
7. Kalluri R, Zeisberg E. Controlling angiogenesis in heart valves. Nat Med. 2006;12(1):1118–1119. [PubMed]
8. Luschka H. Virchow’s Archiv. 1852;4:171–192.
9. Luschka H. Die Anatomie des Menschen. 1863:385. Bd. 2, Abt. II.
10. Bayne-Jones S. Blood vessels of the heart valves. Am J Anat. 1917
11. Montiel MM. Muscular apparatus of the mitral valve in man and its involvement in left-sided cardiac hypertrophy. Am J Cardiol. 1970;26:341–344. [PubMed]
12. Nusebaum A. Archiv für Mikroskop Anat. 1912;80:450–477.
13. Cooper T, Napolitano LM, Fitzgerald MJ, et al. Structural basis of cardiac valvar function. Arch Surg. 1966;93:767–771. [PubMed]
14. De Biasi S, Vitellaro-Zuccarello L, Blum I. Histochemical and ultrastructural study on the innervation of human and porcine atrio-ventricular valves. Anat Embryol. 1984;169:159–165. [PubMed]
15. Ritchie J, Warnock JN, Yoganathan AP. Characterization of the different mitral valve chordae tendineae. Ann Thorac Surg. 2005;80:189–197. [PubMed]
16. Cohn LH, editor. Cardiac Surgery in the Adult. 3rd ed. The McGraw-Hill Companies, Inc.; USA: 2008.
17. Frater RWM. Editorial: Attitudinally correct designation of papillary muscles. J Heart Valve Dis. 2003;12:548–550. [PubMed]
18. Krishnamurthy G, Ennis DB, Itoh A, et al. Material properties of the ovine mitral valve anterior leaflet in vivo from inverse finite element analysis. Am J Physiol Heart Circ Physiol. 2008;295(3):H1141–Hl149. [PubMed]
19. Weind KL, Ellis CG, Boughner DR. Aortic valve cusp vessel density: Relationship with tissue thickness. J Thorac Cardiovasc Surg. 2002;123:333–340. [PubMed]
20. Marron K, Yacoub MH, Polak JM, et al. Innervation of human atrioventricular and arterial valves. Circ J Am Heart Assoc. 1996;94:368–375. [PubMed]
21. Taylor PM, Batten P, Brand NJ, Thomas PS, Yacoub MH. The cardiac vane interstitial cell. Int J Biochem Cell Biol. 2003;35:113–118. [PubMed]
22. Chester AH, Misfeld M, Yacoub MH. Receptor-mediated contraction of aortic valve leaflets. J Heart Valve Dis. 2000;9:250–254. discussion 254-255. [PubMed]
23. Kershaw JD, Misfeld M, Sievers HH, Yacoub MH, Chester AH. Specific regional and directional contractile responses of aortic cusp tissue. J Heart Valve Dis. 2004;13:798–803. [PubMed]
24. Weind KL, Boughner DR, Rigutto L, Ellis CG. Oxygen diffusion and consumption of aortic valve cusps. Am J Physiol Heart Circ Physiol. 2001;281:H2604–H2611. [PubMed]
25. Sonnenblick EH, Napolitano LM, Daggett WM, Cooper T. An intrinsic neuromuscular basis for mitral valve motion in the dog. Circ Res. 1967:9–15. [PubMed]
26. Itoh A, Krishnamurthy G, Swanson JC, et al. Active stiffening of mitral valve leaflets in the beating heart. Am J Physiol Heart Circ Physiol. 2009 electronic publication 10 April, 2009. [PubMed]
27. Rosenow EC. J Infect. Dis. 1912;11:210–224.
28. Waters EA, Chen J, Allen JS, Zhang H, Lanza GM, Wickline SA. Detection and quantification of angiogenesis in experimental valve disease with integrin-targeted nanoparticles and 19-fluorine MRI/MRS. J Cardiovasc Magn Reson. 2008;10:43. [PMC free article] [PubMed]
29. Frink RJ, Merrick B. The sheep heart: Coronary and conduction system anatomy, with special reference to the presence of an os cordis. Anat Rec. 1973;179:189–200. [PubMed]
30. Voci P, Bilotta F, Caretta Q, Mercanti C, Marino B. Papillary muscle perfusion. Circ J Am Heart Assoc. 1995;91:1714–1718. [PubMed]