Cardiac Valvular Pathology: Comparative Pathology and Animal Models of Acquired Cardiac Valvular Diseases

Abstract

Recent voluntary withdrawal of the ergoline-derivative Alzheimers’ drug Pergolide (Permax) resulting from demonstrated risk of cardiac valve injury illustrates the increased importance of valve injury in pharmaceutical toxicology. Following the 2001 landmark discovery of cardiac valve injury associated with the widely prescribed anti-obesity drug combination fenfluramine-phentermine, and subsequent withdrawal, the need to understand and assess cardiac valve biology and pathology both preclinically and clinically has been accentuated. Unique aspects of the developmental biology, anatomy, and physiology of cardiac valves compared to main cardiac tissue have been discovered, and key elements of the pathophysiology of various valvular injury mechanisms have been described. Although general clinical cardiac valvular disease in humans has been well characterized, animal modeling of valvular injury has proved to be difficult and undersubscribed. Additionally, both the preclinical, pharmaceutical, toxicologic assessment of valvular injury and the understanding of species-comparative valvular pathology have been limited. As discoveries and awareness grows, the purpose of this paper is to review the structure and function of cardiac valves, mechanisms, and outcomes of the common acquired human cardiac valve diseases, including those that are drug-related; to summarize comparative laboratory animal valvular pathology; and to review the literature of contemporary animal models of valvular injury.

Introduction

Diseases of the heart represent the most common cause of adult morbidity and mortality in the Western world, followed by cancer and stroke (MMWR 2006). Although atherosclerotic and myocardial injuries top the list of cardiovascular diseases, pathologic conditions of the heart valves represent a significant group of interrelated conditions that produce cardiac disease. There is a large group of congenital cardiac valvular disorders, of which some such inherited conditions may directly or indirectly produce valvular disease in later life. Acquired valvular diseases fall into categories that produce morphologic valve changes, either by primarily injuring and altering the valve tissue or by producing remodeling and alteration to valves secondary to changes in mechanics and/or hemodynamics. The broad categories include age-related degenerative processes, peri- or postinfective processes, mechanical injury, systemic disease-related processes, and drug- or toxin-related injuries. Recent reviews of acquired cardiac valvulopathies have introduced focused descriptions of human valve diseases, elements of the pathophysiology of valvular diseases, and the process of regeneration/repair (Butany et al. 2005; Schoen 2005; Veinot and Walley 2000). Much is yet unknown about the pathogenesis of the multiple causes and processes of valve injury, but it is apparent that important subtle differences exist in the presentation, morphology, and progression of the entities described so far. These differences have implications for both diagnostic characterization of valve disease and for modeling of valve disease processes. This review will provide a brief overview of the anatomy, physiology, and response to injury of the cardiac valves, focusing on the pathogenesis and morphologic features of the principal acquired human and animal cardiac valvulopathies with emphasis on the degenerative, systemic disease-related, mechanical, and drug/toxic processes. The principal human cardiac valve diseases will be compared. Animal valvular heart diseases will be described, with emphasis on known entities in laboratory animals, and, from the current literature, specific contemporary animal models of valvulopathy will be discussed.

Cardiac Valve Anatomy, Physiology, and Response to Injury

Cardiac valves are a set of complex, delicate yet resilient connective tissue structures whose function is to enable the unidirectional, hemodynamic flow of blood through the chambers of the heart. There are four heart valves, consisting of the left and right atrioventricular (AV) valves and the aortic and pulmonic semilunar valves. The right AV valve is also known as the tricuspid valve, with three leaflets. The left AV valve is also known as the bicuspid valve or the mitral valve. The valves are continuous at their base with the myocardium or the great vessel walls, where each is embedded in a concentric ring of fibrous tissues known as an annulus. The free AV valve leaf edges are attached by thin, fibrous, string-like chordae tendinae to the tips of papillary muscles, projections of myocardial tissue arising from the floor of the ventricles. The chordae prevent prolapse of the valve leaves into the atria as the valve closes during systole. The semilunar valves in the pulmonic and aortic outflow tracts have three cusps each and are so named because of their crescent semilunar shape. Based on their shape and array, these valves function as pockets, snapping open under pressure from blood in the aorta or pulmonary artery to fully occlude the outflow tracts and prevent back-flow into the ventricles during diastole. The basic anatomical structures of the heart are depicted in Figure 1.

Cardiac valves develop within the fetal heart from areas of endothelial-mesenchymal transdifferentiation within the lining of the atrioventricular canal that form the endocardial cushions. Cell proliferation elongates the cushions, whereas expansion of the interstitial matrix forms the leaflets and cusps of the valves (Hinton et al 2006). Fully formed valves consist of a surface layer of valvular endocardial endothelial cells (VEC), valvular interstitial cells (VIC), and the interstitial extracellular matrix (ECM). There is a complex, dynamic relationship between these three components that imbue the valves with properties that allow their remarkable function and responsiveness to the variable hemodynamics of the heart, age-related changes, and various pathophysiologic processes.

VEC primarily provide, as in all blood vessels, a nonthromobogenic surface of the valve leaves under the conditions of powerful hemodynamic forces of the blood flow. The VEC are known to have unique properties that make them phenotypically and physiologically unique compared to endothelial cells from other levels of the vasculature, including responsiveness to biochemical stimuli and physical forces, and regulation of inflammatory and immune responses (Aird 2007). Additionally, there is a specific interrelationship between the VEC and the VIC based on embryologic origin where a portion of endothelial cells overlying the cardiac valve cushions invades the primitive mesenchyme and differentiates and proliferates into the mesenchymal VIC in the process known as endothelial-mes-enchymal transdifferentiation (Arciniegas et al 1992; Armstrong and Bischoff 2004; Eisenberg and Markwald 1995). It has been postulated that VEC can revert to this transdifferentiation process in response to injury and provide the source of an incipient population of proliferating VIC producing the thickened, hypercellular, and expanded matrix of damaged valves (Aikawa et al 2006; Paranya et al 2001).

VIC are primarily fibroblast cells that originate from the endothelial-mesenchymal transdifferentiation in the cardiac cushions and are responsible for valve structure via maintenance of the ECM. VIC synthesize and degrade the ECM via secretion of ECM components such as collagen, fibronectin, chondroitin sulfate, and prolyl-4-hydroxylase (Messier et al 1994), and degrading enzymes such as matrix metalloproteinases (MMP) and cathepsin D (Rabkin et al 2001). VIC are phenotypically dynamic cells capable of transforming into myofibroblasts and smooth-muscle VIC that are present in remodeled and diseased heart valves (Walker et al 2004). In this key transformation, the VIC express smooth muscle actin in addition to mesenchymal markers such as vimentin (Taylor et al 2003).

The ECM consists of three layers of organized connective tissue that are comparable between valve types and function together to provide the shape changes and deformations that accompany the cyclic flow of blood. The innermost layer (the ventricularis in semilunar valves, and the atrialis in AV valves) is composed of elastic fiber-rich connective tissue that stretches and retracts during systole and diastole (Schoen 1997). The medial layer is known as the spongiosa and is composed of loose collagenous connective tissue rich in glycosaminoglycans (ground substance) that acts to absorb shear forces and shock between the layers of the valve during cyclic valve motion. The thick outermost or fibrous layer provides strength and stiffness to maintain structural integrity by way of dense regular collagenous connective tissue fibers (Schoen 2005). The ECM also provides a structural framework for cell movement and aggregation, diffusion of growth factors and cytokines/chemokines (Schroeder et al 2003), and passage of nerve fibers through the valve (Marron et al 1996).

Cardiac valvular response to injury and age-related degeneration follows several final common pathways, the most preeminent of which is the generation of excess myxomatous matrix within the ECM accompanied by redifferentiation and proliferation of VIC (Durbin and Gotlieb 2002). Other responses to injurious stimuli can include a predominantly fibrotic reaction in the ECM, lipid deposition, or calcification of the valve (McDonald et al 2002a; Schoen and Edwards 2001).

In response to injury or hemodynamic stresses, VIC become activated and transformed into a myofibroblastic cell type. These cells express smooth muscle actin (α-SMA) and are capable of producing and remodeling excess ECM, and of elaboration of proinflammatory mediators (Fondard et al 2005; Rabkin et al 2001; Walker et al 2004). The transformation of VIC has been widely studied, and key regulators of the process include mechanical stress; transforming growth factor-β (TGFβ), which directly induces the switch to activated myofibroblasts; platelet-derived growth factor; and endothelin-1 (Jian et al 2002; Xu et al 2002). Other regulators of VIC myofibroblasts include growth hormone, interferon-γ, and basic fibroblast growth factor; the latter two are known inhibitors of TGFβ-mediated induction of myofibroblasts (Desmoulière et al 2005; Serini and Gabbiani 1999).

Valves are subject to hemodynamic forces in a way that accentuates and perpetuates the tissue responsiveness to injury. Thus small changes in the form or functional integrity of valve leaves and cusps tend to foster ongoing proliferation and scarring, with greater deformation and retraction producing more valve dysfunction and even more changes in hemodynamics. The functional outcomes of cardiac valve injury fall into two categories: valve regurgitation (also known as valvular insufficiency), where blood flows in a reverse direction owing to failure of a valve to close completely; or stenosis, where blood flow is obstructed by narrowing of the lumen owing to failure of a valve to open completely (Cannistra 2005). The secondary effects of these major categories of valve dysfunction may include ventricular and atrial hypertrophy and/or dilation, myocardial ischemia, aortic dilation, hypertension, pulmonary edema and fibrosis, pleural effusion, chronic passive congestion of the liver, ascites, peripheral edema, and thromboembolism (Schoen 2001).

Human Valvular Pathology

1. Age-Related Onset of Valvular Disease

Aortic Valvular Stenosis

Aortic valvular stenosis (AVS) in adults is most commonly associated with gradual calcification of the normal aortic valve trileaflet. AVS may occur secondary to damage by rheumatic fever, or in association with the occurrence of a congenital bicuspid aortic valve. The disease progresses from the base of the cusps to the leaflets, eventually causing a reduction in leaflet motion and decreased valve area. Regardless of the underlying cause, AVS is an ongoing disease process characterized by lipid accumulation, inflammation, and mineralization with many similarities to atherosclerosis (Bonow et al 2006; Hughes et al 2005). In congenital bicuspid valve, the aortic valve may be stenotic from birth or may acquire stenosis as the valve leaves enter a progressive cycle of hemodynamic stress, resulting in increasing thickening, rigidity, and retraction of the valve (Braunwald 1998). Since the obstruction develops gradually, usually over decades, the left ventricle adapts with hypertrophy that increases the thickness of the left ventricular wall but maintains the normal chamber volume. However, compared to nonhypertrophied hearts, reduced coronary blood flow per unit mass of left ventricle can lead to increased sensitivity to ischemic injury, with higher mortality rates following myocardial infarction (Gaash et al 1990).

Aortic Sclerosis

Aortic sclerosis is defined as irregular valve thickening, mostly resulting from fibrosis, without obstruction to the ventricular outflow tract. Aortic sclerosis is present in about 25% of adults over 65 years of age and although considered idiopathic and progressive, it is often associated with hypertension, smoking, disorders of serum LDL, and/or diabetes mellitus (Stewart et al 1997). Even though left ventricular outflow does not appear obstructed, these individuals have an approximately 50% increased risk for myocardial infarction and death compared to an individual with a normal aortic valve (Otto et al 1999).

Mitral Annular Calcification

The fibrous annular rings define the circular shape of the blood flow tracts between the chambers of the heart and provide a dynamic but resilient concentric anchor for the base of the heart valves. Mitral valve annular calcification occurs in the aged population, with higher occurrence in women. Mitral annular calcification inhibits the full range annular systolic contraction and may limit valve leaflet closure, resulting in mitral regurgitation. It may also involve the conduction system, causing various degrees of atrioventricular block, and may increase the risk of stroke (Benjamin et al 1992).

Myxomatous Valvular Degeneration

Mitral valve prolapse (MVP) syndrome, also called “floppy valve syndrome,” is the result of an age-related, chronic myxomatous degeneration of the atrioventricular valves and occurs in adults at a rate of 1%–2.5% (Freed et al 1999). MVP may be familial or nonfamilial, idiopathic, or associated with certain syndromes such as Marfan syndrome or other connective tissue diseases, as well as Von Willebrand’s disease (Rosenberg et al 1983). The pathogenesis of the valve-specific myxomatous degeneration is unclear, but valvular interstitial cells have been identified as responsible for mediating extracellular matrix degradation by secretion of catabolic enzymes (Rabkin et al 2001). The result is expansion and distortion of the valve leaves with accumulation of nodular to diffuse, poorly organized, proteoglycan-rich extracellular matrix and areas of fibrosis in the spongiosa of the leaf (Rippe et al 1980). This results in systolic billowing of redundant valve leaf into the left atrium with or without mitral regurgitation. Retraction of the irregular and nodular fibrotic valve leaves may lead to, and progressively exacerbate, the regurgitation. Approximately 40% of individuals affected with MVP have involvement of the tricuspid valve, and 2%–10% have morphologic changes with prolapse of the pulmonic and aortic valves (Bonow et al 2006). In addition, prominent deposition of abnormal proteoglycan in neural and conduction tissues in extravalvular areas of the heart distinguish MVP hearts from normal hearts, suggesting that MVP is a manifestation of a general cardiac connective tissue myxomatous change (Morales et al 1992).

2. Mechanical Induction of Valvular Disease

Trauma

Trauma can produce valvular lesions in addition to, or independent of, myocardial injury. When significant myocardial injury exists, the most common associated valve lesions occur in the atrioventricular valves and are usually associated with fatality. The most common valve lesion associated with nonfatal chest trauma is aortic valve rupture or leaf cusp perforation resulting in aortic regurgitation (Boudoulas 2003). In addition to damage to the valve leaves or cusps, atrioventricular valve regurgitation can result from injury to the heart owing to ruptured chordae tendinae, papillary muscle rupture, or papillary muscle contusion. In these instances, once the tricuspid or mitral valve becomes partially prolapsed, the hemodynamic alterations foster the cycle of expanded myxomatous matrix and fibrosis, producing thickening, irregularity, and rigidity as the valve undergoes remodeling.

Dilated Cardiomyopathy

Mitral regurgitation is the most common outcome of dilated cardiomyopathy as changes in the left ventricular architecture and wall motion irregularities produce misalignment of valve mechanics (Soler-Soler 2000). Valve leaf remodeling resulting from altered hemodynamics ensues. Although not as likely to be affected, the changes can occur in the right atrioventricular valve secondary to right ventricular architecture change or right ventricular papillary muscle dysfunction (Raman et al 2003).

3. Other Disease-Related Valvular Disease

Rheumatic Fever

Rheumatic fever is an inflammatory disease that may develop after an untreated infection with group A β-hemolytic streptococcus bacteria and can involve the heart, joints, skin, and central nervous system. Rheumatic fever is common worldwide and is responsible for many cases of damaged heart valves as a consequence of rheumatic carditis. Since the beginning of the 20th century, rheumatic fever has been far less common in the United States, but since the 1980s there have been a few outbreaks. Rheumatic fever primarily affects children between ages 6 and 15 and occurs approximately 20 days after “strep throat” or scarlet fever. In up to a third of cases, the underlying streptococcal infection may not have caused any symptoms. The rate of development of rheumatic fever in individuals with untreated streptococcal infection is estimated to be 3% (Stollerman 1995).

Worldwide, rheumatic fever is still the underlying cause of the majority of acquired valve diseases, particularly producing mitral stenosis and/or regurgitation, aortic stenosis and/or regurgitation, and tricuspid stenosis. Valve injury is a consequence of general heart tissue inflammation, which occurs following streptococcal infection and is a result of an autoimmune reaction with both humoral and cellular-mediated components. Antibodies to the bacterial M-protein, a streptococcal virulence factor with antiphagocytic properties, and antibodies to streptococcal surface carbohydrates cross-react with endothelium and cardiac proteins including myosin and laminin (Veinot 2006). Myosin cross-reactive T-lymphocytes infiltrate the heart and create diffuse inflammation, degeneration, and remodeling. Over time the valve leaves appear fibrotic, thickened, and neovascularized, with chronic inflammation, fusion at commissures, and fibrosis (Schoen and Edwards, 2001). There may also be thickening and shortening of chordae tendinae associated with these changes in atrioventricular valves.

Infective Endocarditis

Valve injury and functional deficits may occur secondary to nonrheumatic inflammation and/or to thrombosis/vegetations (a propagating accumulation of coagulation proteins and platelets) on the valve surface. Although infective endocarditis can occur on normal valves in patients experiencing sepsis or bacteremia, it has a predilection to valves that have underlying congenital, degenerative, rheumatic, or functional/ hemodynamic alterations. The infective endocarditis may produce thrombotic valvular vegetations, valve leaf destruction and remodeling, and/or regurgitation (Boudoulas 2003). Valves most commonly affected are the AV and aortic, whereas the pulmonic valve is very rarely involved. In affected patients, tricuspid lesions occur more commonly in intravenous drug users, possibly because of direct exposure to pathogens via the venous routes of drug administration.

Sterile thrombotic endocarditis occurs in patients with chronic systemic diseases including malignancy, tuberculosis, renal failure, systemic lupus erythematosis, and HIV/AIDS. The combination of a hypercoagulable state and endothelial damage is thought to predispose to this condition, which has a clinical course and outcome similar to the thrombotic valvular vegetations in infective endocarditis.

Ischemic Cardiomyopathy

Papillary muscle damage and dysfunction as a consequence of ischemic cardiac disease can produce atrioventricular valve regurgitation (Silver and Silver 2001). More common in the mitral valve, it may be accompanied by ventricular dysfunction as a result of mural or septal infarction, or chronic cardiomyopathy. Acute injury can include papillary muscle rupture/detachment of the chordae tendinae.

Carcinoid Syndrome

Carcinoid syndrome occurs in 50% or more of patients diagnosed with late-stage and/or metastatic (to the liver) carcinoid tumors. These tumors arise from enterochromaffin cells in the intestine and secrete high levels of serotonin and other vasoactive substances such as neuropeptide K and bradykinin. Owing to first-pass clearance in the liver via the portal venous system, the tumors in the intestine usually do not produce secondary effects. However, metastatic tumors in the liver deliver secretory products into the hepatic vein, resulting in high levels of tumor-derived serotonin exposure to the right side of the heart. This situation is thought to produce endocardial damage leading to thickening, retraction, and regurgitation of the right atrioventricular valve and pulmonic valve (Moller et al 2003). Carcinoid plaques on the endocardium and valves are characteristic of carcinoid syndrome and consist of focal proliferation of subendothelial cells and fibromyxoid matrix without the elastin fiber component. Although levels of endogenous serotonin are significantly higher in patients with carcinoid heart disease, the exact mechanism of serotonin-induced endocardial disease is unclear (Zuetenhorst et al 2003). It has been shown that serotonin exerts a mitogenic effect on endothelial cells via 5-hydroxytryptamine-1B (5-HT_1B) receptors, which have further been shown to be specifically overexpressed in carcinoid-affected valves (Rajamannan et al 2001a).

Systemic Lupus Erythematosus and Antiphospholipid Antibody Syndrome

Original reports of a unique verrucose endocarditis including involvement of the valves in systemic lupus erythematosus (SLE) patients have been followed by recognition of a syndrome of SLE-related endocarditis, pericarditis, and embolic phenomenon (Libman and Sachs 1924). Mitral or aortic regurgitation with valvular thickening is the most common finding in SLE patients.

Valvular disease has been reported in SLE patients regardless of the presence or absence of antiphospholipid antibodies. However, approximately 50% of patients with antiphospholipid antibody syndrome (APLAS) have SLE. Regardless whether APLAS is primary or secondary to SLE, autoimmune diseases, malignancy, or drug abuse, it manifests as circulating antibodies to negatively charged membrane phospholipids with thrombocytopenia, and arterial or venous thrombosis (Veinot and Walley 2000). The pathogenesis of valvular lesions associated with SLE, with or without APLAS, is not clear, but it is postulated to be a primary immunologic insult to valvular endothelial cells causing surface thrombosis, interstitial inflammation, fibrosis, and calcification (Lev and Shoenfeld 2002). Anticardiolipin antibodies have been implicated in the pathogenesis of the heart valve component of the SLE (Leszczynski et al 2003).

Marfan Syndrome

Marfan syndrome is an inherited connective tissue disorder involving the FBN1 gene, which encodes for fibrillin-1 protein, an essential component of elastin fibers. Additionally, fibrillin-1 protein normally binds inactive TGF-β, a powerful mediator of connective tissue growth and remodeling, as part of the overall regulation of TGF-β. In Marfan syndrome, fibrillin-1 dysfunction results in excess TGF-β available for conversion to the active form, producing excessive and atypical connective tissue growth (Kaartinen and Warburton 2003). In addition to connective tissue problems throughout the body, Marfan syndrome individuals often develop mitral regurgitation and MVP, the latter being a peculiar syndrome of heart valve disease with myxomatous degeneration of the mitral valve (Robinson and Booms 2001). As described above, not all patients with MVP have Marfan syndrome. Other underlying connective tissue or extracellular matrix biology issues, such as non-fibrillin-1-related dysregulation of TGF-β, may be present (Weyman and Sherrer-Crosbie 2004).

4. Drug-Induced Valvular Disease

Fenfluramine/Phentermine

The prescription drug combination of fenfluramine and phentermine (fen-phen) was marketed as an anorexigenic in 1996. The drugs, approved by the FDA separately, were used in combination and were widely prescribed, with over 18 million prescriptions written in the first year. In 1997, the first report of heart valve disease in patients taking the combination was published (Connolly et al 1997). The report covered 24 cases in women with primarily echocardiographic diagnoses and provided detailed description of lesions in three patients. In these three patients, and many others subsequently described, left- and right-sided heart valves were shown to have surface plaques composed of myofibroblasts in a myxoid matrix that encased the valves and often the chordae, while sparing the original valve structure (McDonald et al 2002b).

The drug combination was withdrawn from the market in 1997, and reports have followed that further clarify the fen-phen valvulopathy and detail the dose relationships, time course, severity, predispositions, comorbidities, and both clinical and pathophysiologic outcomes. The incidence of heart valve abnormalities associated with fen-phen was studied in large populations who had taken the drug combination or either drug alone. Data from these studies suggested that fenfluramine was most likely to produce valvulopathy, especially in patients treated for four months or longer (Jick et al 1998; Khan et al 1998). In affected patients, the lesion was observed more commonly in the left side of the heart, usually producing aortic regurgitation and sometimes mitral regurgitation (although there was considerable individual variability). The overall prevalence of fen-phen valvulopathy was low, and affected individuals were usually asymptomatic (Gardin et al 2000; Wadden et al 1998; Weissman et al 1998). Histologically, valvulopathy was confirmed as a plaque-like condition that consisted of proliferations of myofibroblastic cells in a myxoid stroma often with vascular channels, CD3 positive lymphocyte and CD68 positive macrophage inflammatory foci, and fibroelastic tissue in deep areas of the plaques in close proximity to the original valve surface (Steffee et al 1999; Volmar and Hutchins 2001). The lesion was seen as having similarity to other plaque-like valvulopathies, including those associated with carcinoid disease and other serotonergic drugs such as ergotamine and methysergide (Seghatol and Rigolin 2002). Indeed, the most widely accepted pathogenesis of the fenfluramine-related valvulopathy revolved around plasma serotonemia and/or 5-HT₂ receptor agonism. Fenfluramine and its metabolite norfenfluramine bind to 5-HT receptors, with nor-fenfluramine having high affinity for 5-HT_2B and 5-HT_2C receptors. Binding to these receptors activates the mitogenic pathways associated with 5-HT_2B receptors that have been shown to be present on human heart valves (Fitzgerald et al 2000; Rothman et al 2000).

Ergotamine and Methysergide

Ergot alkaloid drugs such as ergotamine and methysergide are used for the treatment of migraine headaches. Both drugs have agonist activity at 5-HT_2B receptors and produce valve lesions similar to the lesions seen with carcinoid disease. Characteristic features of the lesion were reported many years prior to the emergence of the fen-phen valvulopathy (Hauck et al., 1990; Hendrikx et al., 1996; Redfield et al., 1992). Ergotamine and methysergide primarily affect the mitral valve, producing histologic lesions consisting of plaques of proliferative myofibroblast cells in an avascular and noninflamed myxoid or collagenous stroma, often resulting in valvular stenosis and/or regurgitation (Hendrikx et al 1996).

Pergolide, Cabergoline, and Bromocriptine

Several other ergot-derived dopaminergic compounds are used for indications such as Parkinson’s disease, restless leg syndrome, and hyperprolactinemia. These include pergolide, cabergoline, and bromocriptine. All of these drugs have agonist activity at 5-HT_2B and have been reported to be associated with mitral and/or tricuspid regurgitation with histologic valve lesions similar to ergotamine and methysergide, and carcinoid syndrome (Baseman et al 2004; Pinero et al 2005; Schade et al 2007; Serratrice et al 2002; Waller and Kaplan 2006; Zadikoff et al 2006; Zanettini et al 2007). Pergolide (Permax®, Valeant, USA) has been voluntarily withdrawn from the US market because of concerns over the growing incidence of cardiac valvular disease with features highly consistent with the fen-phen and ergotamine-related valvulopathies (FDA News 2007).

Based on the pattern of lesion incidence associated with drugs of this type, the activation of 5-HT_2B receptors remains the leading hypothesis for the development of drug-induced valvulopathy. Indeed, another ergot-derived dopamine receptor agonist, lisuride, which is used in the treatment of migraine headaches and Parkinson’s disease, is a known 5-HT_2B antagonist that has been shown to have no clinical association with valvulopathy, further supporting the 5-HT_2B agonism pathogenesis (Hofmann et al 2006).

Other Drug-Induced Valvulopathies

Besides fenfluramine, the ergot-derived agonists are not the only drugs with potential for heart valve effects. The anorexigenic agent benfluorex (Servier, France) has been associated with at least two cases of valvular heart disease, which may be linked to a benfluorex-associated serotonergic mechanism (Noize et al 2006; Rafel Ribera et al 2003).

The amphetamie derivative 3,4-methylenedioxymetham-phetamine (MDMA, “Ecstasy”) and its metabolites preferentially bind 5-HT_2B receptors and produce mitogenic activity in human heart valve interstitial cells in vitro (Setola et al 2003). Although clinical reports of valvulopathy occurring in MDMA users have not been made, MDMA and many drugs in other classes have been screened for 5-HT_2B activity and are under watchful clinical surveillance.

Comparative Valvular Heart Disease

The embryology, anatomy, and histology of the heart is well conserved across mammalian species, with most differences occurring as a result of economy-of-scale issues in the physiology of heart function. Heart rate and localized hemodynamic pressures within the cardiac cycle vary widely according to body mass of the species. Response to injury also appears to vary by species with differences in the occurrence of spontaneous valve diseases and in sensitivity to induced valvular lesions; however, from the perspective of tissue pathology, the final common pathways of valve injury seem to be conserved.

Spontaneous Valve Disease in Laboratory Animals

Unless available precise tissue trimming guides for the heart are used, systematic examination of the heart valves may not be a component of the toxicologic pathology assessment in preclinical studies. Histologic visualization of heart valves is likely to vary in standard sectioning, and the published documentation of valvular pathology, whether background or induced, has been relatively limited. A far greater understanding of the background and natural history of heart valve disease in standard laboratory animal species is needed to ensure that spontaneous valve lesions are not misinterpreted as treatment related. Specific inclusion of heart valve examination in preclinical toxicology studies for drug development is warranted to increase awareness of typical variations in histologic sectioning, features of normal valve histology, and spontaneous lesions. Several recent studies have provided more information on spontaneous valvular disorders in mice, rats, dogs, and monkeys and are described below.

Mouse

Spontaneous valve disease has been described in the Swiss CD-1 mouse. In an examination of 215 male and female mice up to 2 years of age, 96 had evidence of endocardial myxomatous change (EMC) in heart valves (Elangbam et al 2002). The lesions appeared microscopically as nodular or segmental thickening of valve leaves as a result of superficial subendocardial or interstitial fibromyxoid change. The expanded interstitium consisted of loose, lightly staining myxoid extracellular matrix with a moderately dense population of fibroblastic cells. The alteration was usually on the free edge of the leaf and occasionally formed flat plaques. The pulmonic valve was commonly affected, followed by mitral, aortic, and tricuspid valves in order of incidence. Thrombi were occasionally present, but no morbidity or mortality could be attributed to the presence of valve changes.

Rat

Laboratory rats have previously been described with age-related valve changes, including myxomatous change (Figure 2) and valvular endocardial proliferation (Mohr et al 1992). Indeed, EMC has been classified in the nomenclature of the Society of Toxicologic Pathology (Ruben et al 2000). In a more recent examination of 220 male and female Sprague-Dawley (SD) rats up to 2 years of age, 188 had lesions typical of EMC (Elangbam et al 2002). In the rat, the incidence was greatest in the mitral valve, followed by the aortic, tricuspid, and pulmonic valves. EMC was not implicated in the morbidity or mortality of any rats in the study. However, EMC was correlated with the incidence of the common age-related murine progressive cardiomyopathy syndrome, although the nature of an interrelationship or potentially mutual pathogenesis is unknown. Overall, the microscopic features of EMC in rats and mice can be compared to myxomatous valvular degeneration in humans, and to valvulopathy induced by carcinoid syndrome, ergotamine derivatives, and fenfluramine. Generally, all have subendocardial or superficial fibromyxoid change of the valve leaves.

Dog

Although multiple primary or secondary disease conditions of heart valves may occur in individual dogs, a general spontaneous age-related atrioventricular valvulopathy occurs in the dog. The condition is referred to as myxomatous atrioventricular valvular degeneration, chronic valvular fibrosis, or endocardiosis. Incidence is reported between 11% and 60%, with a direct relationship to age. Lesions are seen in dogs as early as 2 to 3 years of age and are most common in the mitral valve, although lesions may occur in the tricuspid valve (Rush 2002). The valve leaves are grossly thickened and shortened, with nodular margins. The leaves, and often the chordae, contain increased fibrous connective tissue and an abundance of acid mucopolysaccharide ground substance (Figure 3). Hemorrhage and mineralization may be present, but inflammation is rare. The myxomatous degeneration of the interstitium of the leaves is associated with a phenotypic conversion of the valvular interstitial cells to a mixed myofibroblast or smooth muscle cell phenotype, with interstitial cells clustered in closer association with the valvular endothelium (Black et al 2005). The histologic appearance of the lesion is most similar to mitral myxomatous valvular degeneration, “floppy valve” disease, in humans (Pedersen and Haggstrom 2000). However, the lesion in dogs is more fibrotic, and the nodular thickening with curling and shortening of the valve leaves does not fully emulate the billowing and prolapse seen in the human condition. Dogs gradually incur valve insufficiency and AV regurgitation that may produce ventricular hypertrophy/ dilation and chronic passive congestion of the lungs and/or abdominal viscera (Rubin 1992).

Nonhuman Primates

A general pattern of spontaneous valvular disease in populations of laboratory nonhuman primates has not been reported. In a study of the hearts from 120 wild-caught or purpose-bred cynomolgus macaques (Macaca fascicularis) aged 2–7 years, there were no observations of valve disease (Keenan and Vidal 2006). Additionally, a study of hearts from 2462 purpose-bred nonhuman primates, including cynomolgus macaques, rhesus macaques (Macaca mulatta), and common marmosets (Callithrix jacchus), aged 1–3 years, revealed no spontaneous valve diseases (Chamanza et al 2006). The paucity of evidence for spontaneous valvular disease in nonhuman primates is likely artificial and the consequence of only a few published studies in cohorts of young animals.

Models of Valvulopathy

Animal models of acquired valvular disease, particularly drug-induced, are uncommon and have not produced robust test mechanisms for studying the precise pathogenesis of specific etiologies or final common pathways of valve injury. The importance of the discovery of the fenfluramine-phentermine valvulopathy, and subsequent recall of those compounds, has been accentuated by the more recent recall of the pergolide compound owing to sufficient evidence of clinical risk for valve injury (Schade et al., 2007). As a consequence, research into the mechanism and modeling of valvulopathy has taken on a new emphasis, and new models are being developed. The following is a summary of the preeminent published animal models of valvulopathy, as listed in Table 1.

Fenfluramine-Phentermine Administration

Attempts at a logical, direct toxicologic model of fen-phen valvulopathy in laboratory animals have met with limited success. Whereas in vitro effects of fen-phen administration on cultured valvular interstitial and endothelial cells via serotonin receptors have been established, consistently reproducing the lesion in vivo has been elusive in adult rodents. The attention has switched to pre-natal or neonatal animals. Bratter et al (1999) treated pregnant rats with continuous subcutaneous infusion of phentermine and dexfenfluramine at doses approximately 10 times the human clinical dose from day 3 through 17 of gestation. In addition to transient anorexigenic effects (control animals were pair fed), there was reduced density of serotonergic axons in the brains of the treated mothers but not in the 21-day-old neonatal pups. However, 25% of the pups had grossly visible mitral valve changes consisting of a glossy, white thickening and increased rigidity (Bratter et al 1999). Follow-up studies have not been reported. Rayburn et al (2000) studied the effects of antenatal exposure of fenfluramine and dexfenfluramine on the cardiac development of CD-1 mice. Pregnant mice were given either compound in a feed formulation producing dose levels similar to or higher (approximately threefold) than the human clinical dose associated with valvulopathy. Gross and histologic examination of the hearts from the mothers and the pups at postnatal day 120 showed no evidence of changes in the valves of any group (Rayburn et al 2000). Key differences in the preceding studies may account for differential results and include differences in biology of rats and mice, route of administration, dose level, selected day for postnatal examination, and the administration of the fenfluramine-phentermine combination compared to fenfluramine or dexfenfluramine alone.

There are no other reports of specific studies using direct administration of fenfluramine or phentermine in laboratory animals that have successfully produced changes in heart valves. This result likely attests to the failure rate of reproducing the valvulopathy in vivo.

Pergolide Administration in Rats

Droogmans et al (2007) investigated the induction of pergolide-related valvular heart disease in Lewis rats. Eight male rats were given daily intraperitoneal injections of 0.5 mg/kg Pergolide for five months, and an additional eight rats were given daily subcutaneous injections of 20/kg serotonin; all were compared to a group of 14 sham-treated rats. At the end of the treatment period, echocardiographic studies revealed aortic and/or mitral regurgitation in the pergolide- or serotonin-treated rats with none in control rats, and although pulmonic and tricuspid regurgitation was present in the treated and control rats, there was greater severity in the pergolide- and serotonin-treated groups. Additionally, histologic analysis revealed a positive correlation between echocardiographic findings and valve thickening. This finding corresponded to a myxoid change occurring diffusely in the interstitium of valve leaves of mitral, tricuspid, and aortic valves of both pergolide- and serotonin-treated rats. Control rats also displayed myxoid change in the valve leaves, but to a lesser magnitude and generally as nodular foci in the free ends (Droogmans et al 2007). This small study is the first report of an in vivo model of pergolide-induced valvulopathy in the rat and a corroboration of the findings by Gustafsson et al (2005) in serotonin-administered SD rats discussed below.

dl-amphetamine Administration in Rats

Elangbam et al (2006) examined the incidence of age-related spontaneous valvulopathy (endocardial myxomatous change) in SD rats and made comparisons to SD rats treated with dl-amphetamine for two years. Having characterized the histomorphology and histomorphometry of EMC in SD rats as having segmental to nodular thickening consisting of fibromyxoid tissue with increased interstitial glycosaminoglycans (Movat’s pentachrome stain), the evaluation of control Fischer 344 rats and Fischer 344 rats treated with dl-amphetamine for two years revealed small but statistically significant increases in the incidence and severity of EMC in mitral valves (Elangbam et al 2006). Although there is in vitro evidence for mitogenic stimulation of human valvular interstitial cells by amphetamine derivatives such as 3,4-methylene-dioxymethamphetamine (MDMA) and 3,4-methylenedioxyam-phetamine (MDA) (Setola et al 2003), there is no clinical evidence of cardiac valvular disease among large populations of human patients prescribed marketed amphetamine drugs over long periods of time. Likewise there is no published animal toxicology or animal models of amphetamine-related valvulopathy.

Vasoactive or Hemodynamically Induced Valve Disease in Dogs

Vasoactive pharmaceuticals have been associated with hemodynamically induced heart lesions, including valve injury, in dogs. Experimental positive inotropic substances, particularly those with additional vasodilatory properties, have been associated with acute mitral valve injury (Schneider 1990). At toxicologic doses for periods of weeks to months, dogs displayed mitral valvular injury such as focal swellings of granulation tissue, hemorrhage, myxomatous interstitial change, and fibrosis at the base of the mitral valve and along the valve closure margins (Schneider 1990). The extensive valvular injury in dogs is believed to have occurred with extensive hemodynamic derangements. The author reported similar findings with several compounds in this class, including Amrinone (Sterling-Winthrop, USA) and theophylline, and reference findings with other experimental cardiostimulatory and vasodilatory drugs. Preclinical toxicology reports of investigations in dogs with type III phosphodiesterase inhibitors indicate propensity to produce valve injury, usually in conjunction with other multiple cardiac toxicities (Sandusky 1997). Lesions reported range from acute hemorrhage in AV valve leaves to fibroplasia, proliferative endocardial cell lesions at the base of AV valves, and foci of granulation tissue in AV valve interstitium (Harleman et al 1986; Isaacs et al 1989).

The findings with these named classes of vasoactive pharmaceuticals have not been seen in humans using approved drugs of these classes, likely because of species differences or because at the commonly used lower therapeutic doses, extensive hemodynamic derangements do not occur.

Serotonin Administration in Rats

Attention has been focused on the likely active mechanisms of serotonergic drug-induced valvulopathy, namely 5-HT_2B receptor activation. Gustafsson et al (2005) studied the effects of exogenous serotonin in adult SD rats as a model for carcinoid disease. Daily subcutaneous injections of serotonin were administered to 10 SD rats for 5 or 90 days, the latter producing hyperserotonemia and clinical signs consistent with carcinoid disease. In vivo echocardiographic examination of the hearts at 90 days revealed reported findings of aortic and pulmonary valve insufficiency. In subsequent histologic examination, 5 rats had morphologic changes in the aortic valves that consisted of thickening and shortening of leaflets with increased myofibroblast cells in a collagenous matrix. Some atrioventricular valves and some areas of atrial and ventricular endocardium were reported to have surface plaque-like formations. Ki-67 positive cell proliferation was not present in the valve tissue of these rats but was detected in the aortic valves of a group of the rats treated with serotonin for only 5 days. Additionally, RT-PCR analysis on aortic valve tissue from untreated rats detected mRNA for the serotonin transporter 5-HTT and the serotonin receptors 5-HT_1A, 5-HT_2A, and 5-HT_2B, but not 5HT_2C. The authors concluded that the presence of receptors andevidence of tissue alterations in cell proliferation, functional alterations, and histomorphology suggested a direct role for serotonin in the heart valve changes of treated rats (Gustafsson et al 2005). However, although a good correlation existed between the echocardiographic data and the histologic findings, many valve structures could not be visualized histologically or were in a plane of section that prevented comparison between treated and untreated groups. Further, although no similar histologic findings were present in the concurrent control rats, the histologic findings in this study were similar to, and within the reported range of the variability of the spontaneous changes in the histomorphology of cardiac valves in SD rats (Elangbam et al 2002). Thus, given the small sample size, it is possible that the histologic lesions reported by Gustafsson et al (2005) may reflect spontaneous findings rather than the effects of increased serotonin. The valve lesions were not reproduced in a subsequent study by Hauso et al (2007) in which a 5HT_2B/2C antagonist, terguride, was coadministered to serotonin-treated rats. The terguride did appear to ameliorate the subcutaneous injection site tissue reactions associated with serotonin alone, but despite echocardiographic evidence of pulmonic regurgitation in some serotonin-treated rats, there were no histologic findings (Hauso et al 2007). No other repeat or follow-up studies have been reported, with the exception of a small group of serotonin-treated rats in the pergolide administration study by Droogmans et al (2007), described above.

Although serotonin receptors, particularly 5-HT_2B, remain the focus of the likely pathogenesis of drug-induced valvulopathy and carcinoid syndrome in humans, a direct effect of circulating exogenous serotonin or 5-HT receptor agonists has not been repeatedly shown in a rodent model, which attests to the difficulty of reproducing and studying this lesion.

5-HTT-KO Mice

Outside the central nervous system, endogenous serotonin originates from the gastrointestinal tract and is stored in platelets. In addition to sequestration of serotonin in platelets, degradation of serotonin occurs in the lung, where 5-HT transporter (5-HTT) protein on the surface of the pulmonary vascular endothelial cells removes circulating serotonin. 5-HTT is also highly expressed on platelets and is present on cardiac endothelial cells. To examine the pathogenesis of 5-HT-related valvular injury, Mekontso-Dessap et al (2006) investigated the outcome of a mouse gene knockout for 5-HTT. Although investigations into the role of administered 5-HTT inhibitors have not shown a link to valvular heart disease (Mast et al 2001), it was hypothesized that 5-HTT gene knockout mice may emulate the valvular heart disease seen in humans with long-term exposure to endogenous serotonin or serotonergic receptor activators. Five- to ten-week-old male mice deficient in 5-HTT demonstrated left ventricular dysfunction and dilation, and myocardial and valvular fibrosis, sometimes with cartilage metaplasia, in all four valves compared to wild-type controls. Additionally, this study examined the effect of 5-HT_1B gene knockout in mice. In contrast to the 5-HTT knockout mutants, 5-HT_1B knockout mice had no changes in cardiac function or histomorphology compared to wild type. However, dual 5-HTT and 5-HT_1B knockouts had accentuation of cardiac functional deficits and similar histomorphology findings to 5-HTT knockouts. Overall, the results suggest that 5-HTT gene deficiency leads to cardiac dysfunction with myocardial and valvular fibrosis, but that the 5-HT_1B receptor does not appear to have a critical role in this process (Mekontso-Dessap et al., 2006). The histologic pattern of the cardiac changes in these mice does not fully emulate a specific human valvulopathy, such as fen-phen or carcinoid valvulopathy, owing to the more pure fibrotic nature of the changes, lack of increased mucinous ground substance, and the occurrence of focal myocardial fibrosis, a feature not typical of any valvulopathy syndrome.

Streptoccocal M Protein in Rats

Rheumatic heart disease is an autoimmune disease resulting from group A streptococcal infection. The specific antigen associated with inflammatory rheumatic heart disease is streptococcal M protein, which is structurally and immunologically similar to cardiac myosin, but not skeletal myosin. Quinn et al (2001) investigated the hypothesis that the immune reaction to streptococcal M protein could produce inflammatory valvular heart disease similar to those seen in rheumatic fever. Three of six Lewis rats immunized with recombinant type 6 streptococcal M protein (rM6) developed multifocal myocardial inflammation as well as valvulitis. Valvular lesions were primarily inflammatory and appeared to have initiated at the valve surface and spread into the valve. Surface verrucose-like lesions were also present. T lymphocytes from rM6-immunized rats proliferated in the presence of purified cardiac myosin but not skeletal myosin, suggesting a specific cellular immune reaction to heart tissue, including valve tissue (Quinn et al 2001).

BON Cell Transplant in Mice

Carcinoid heart disease occurs in most patients with carcinoid syndrome, a metastatic neoplasm of serotonin-producing gastrointestinal endocrine cells, and is characterized by fibrous thickening of cardiac valves. The role of serotonin in the pathogenesis of carcinoid valve disease was investigated by Musunuru et al (2005) using nude mice transplanted with human pancreatic carcinoid BON cells. Seventeen nude mice were given intrasplenic inoculation of approximately 10⁷ BON cells, and pathology examinations were performed 9 weeks later. Sixty-five percent of mice developed liver metastases. These mice had significantly elevated plasma serotonin levels and increased surface area of the right atrioventricular valve leaves, which corresponded histologically to fibrosis of the valves. These findings were consistent with human carcinoid valve disease (Musunuru et al 2005).

LDL Receptor-Deficient Mice

Age-associated valvular degeneration producing aortic valvular stenosis (AVS) in humans is characterized by lipid accumulation, collagen deposition, and calcification containing smooth muscle and osteoblast-like cells (Rajamannan et al 2003). Aged hypercholesterolemic, low-density lipoprotein receptor-deficient apolipoprotein B-100-only mice [LDLr(-/-)ApoB(100/100)] develop aortic valve calcification with functional aortic valvular heart disease compared to wild-type controls (Weiss et al 2006). These mice are a useful model for AVS.

Another model of age and atherosclerosis-associated aortic stenosis was recently investigated by Drolet et al (2006) using a diet-induced obesity model in mice. Wild-type C57BL/6J mice, which are genetically prone to diet-induced obesity and atherosclerosis, and low-density lipoprotein receptor knockout (LDLr-/-) 57BL/6J mice were fed either a normal diet or a high-fat/high-carbohydrate (HF/HC) diet for four months. Wild-type mice on a HF/HC diet became mildly hypercholes-terolemic, obese, and hyperglycemic, and as expected, LDLr-/-mice became severely hypercholesterolemic. Both groups on HF/HC diets had smaller aortic valve areas and higher trans-valvular velocities. Histologically, aortic valve leaves were thickened with infiltrations of lipids and macrophages, consistent with the histologic appearance of valves affected by AVS (Drolet et al 2006).

Apoliporotein E-Deficient Mice

In further investigation of the pathogenesis of AVS, Tanaka et al (2005) assessed valvular function and morphology in C57BL/6 mice (green fluorescent protein or β-galactosidase expressing strains) and apolipoprotein E-deficient (ApoE-/-) mice. Increased transaortic flow velocity correlated with increased age in wild-type and ApoE-/-mice, suggestive of aortic stenosis. The aortic valves of aged ApoE-/- mice had histologic evidence of sclerosis that resembled human AVS. Cells of the sclerotic valves of ApoE-/- mice were smooth muscle actin-positive, whereas most cells in the wild-type mice aortic valves were positive for endothelial cell or macrophage markers. Additionally there were bone-marrow–derived cells that were positive for osteoblast-related proteins near foci of ectopic calcification, and there was chemokine expression and frequent apoptotic cell death (Tanaka et al 2005).

Hypercholesterolemic Rabbits

Rabbits have been used in multiple models of cardiovascular disease. such as a spontaneous atherosclerosis model in Watanabe rabbits (Aliev and Burnstock 1998) and diet-induced hypercholesterolemic and atherosclerotic models in wild-type rabbits. Valvular disease, primarily aortic valvular stenosis, has been described in several models. Rabbits fed high-cholesterol diets develop changes in echocardiographic measurements and gross and histomorphologic changes consistent with human AVS (Cimini et al., 2005; Drolet et al., 2003). Rajamannan et al (2001b) discovered cellular apoptosis in aortic valves of diet-induced hypercholesterolemia in New Zealand white rabbits, suggesting that fatal cellular degeneration plays a role in the mechanism of atherosclerotic valvular disease (Rajamannan et al 2001b). Additionally, functional and morphologic features of AVS can be induced in rabbits using models of experimentally induced hypertension (Cuniberti et al 2006).

eNOS-Deficient Mice

Aortic stenosis may result from many factors, including a congenital defect in the morphogenesis of the aortic valve known as bicuspid aortic valve. This malformation is characterized by the presence of two rather than three cusps in the morphology of the aortic valve and predisposes the aortic valve to remodeling and fibrosis resulting in stenosis. Bicuspid aortic valve may be a result of alterations in endothelial nitric oxide synthase (eNOS) function during embryogenesis (Braunwald 1998). Studies into the embryologic development of the heart valves have shown an atypical, critical transformation of a group of overlying cardiac cushion endothelial cells into underlying mesenchymal tissue of the future valves (Eisenberg and Markwald 1995; Arciniegas et al 1992). The molecular mechanisms of this process with respect to the role of eNOS, a well-known participant in vascular embryogenesis of limbs and postdevelopmental angiogenesis/remodeling, were studied by Lee et al (2000) in gene knockout eNOS-deficient mice. Five of 12 mature eNOS-deficient mice (C57BL/6J background) had a bicuspid aortic valve anomaly, compared to none in 26 wild-type control mice. Immunohistochemical analysis revealed prominent eNOS staining of endothelial cells lining the aortic valve leaflets in the wild-type mice and no staining in the cohort eNOS knockouts. Although vascular casts did not reveal stenosis of the aortic outflow tract in bicuspid-affected mice, the study did not assess in vivo function of the aortic valve tract and did not follow adult mice for longer periods of time (Lee et al 2000).

Fibrillin-1 Deficient Mice

Mitral valve prolapse associated with progressive myxomatous valvular degeneration is a common, age-related, cardiac valvular disease. An all-encompassing pathogenesis is not known, but the condition is often part of well-understood genetic syndromes such as Marfan syndrome (MFS). MFS is a systemic connective tissue disorder caused by mutations in the structural matrix protein fibrillin-1 gene. Fibrillin-1 is involved in the regulation of the interaction of the cytokine TGF-β with TGF-β binding proteins. Deficiency of fibrillin-1 leads to enhanced TGF-β activity. As discussed above, TGF-β has a critical involvement in extracellular matrix modeling and may play a direct role in valvular remodeling pathways leading to valvulopathy. Ng et al (2004) studied the role of the fibrillin-1 gene in connection with increased TGF-β signaling in the multisystemic pathogenesis of MFS, including myxomatous valvular degeneration. Using a mouse model of MFS with a fibrillin-1 gene deficiency (heterozygous or homozygous Fbn1 ^C1039G ), an assessment of valvular morphology revealed echocardiographic changes in mitral valve architecture, including functional prolapse with atrial and ventricular enlargement and histomorphologic changes including lengthening and thickening of valve leaves with cell proliferation and decreased apoptosis of valvular interstitial cells. Additionally, affected valves had increased TGF-β activation and signaling, and moreover, TGF-β antagonism in vivo prevented the valvulopathy that occurred in the absence of treatment with a TGF-β antagonist (Ng et al 2004).

Discussion and Conclusion

Knowledge of the structural basis of the function and biology of cardiac valves has elucidated mechanisms of disease and fostered understanding of the processes underlying the final common pathways of manifestation of valvulopathy. The complexity of the heart valve, from embryology of the valve tissues to the changes developing in aging and to the adaptive hemodynamic forces influencing valve function, underscores the varied approaches to studying heart valves and heart valve diseases and the need for further understanding. Simple discoveries make compelling progress and open new ideas. For example, recent investigations suggest that the endothelial cells covering the heart valves may have distinct phenotypes that vary by location. Endothelial cells on the aortic side of the aortic semilunar valve appear to express fewer inhibitors of calcification, certain specific pro-inflammatory molecules, and enhanced antioxidant genes compared to cells lining the ventricular side, a difference which very likely contributes to differential disease processes (Schoen 2006). Key cellular players among the valvular interstitial cells remain to be fully investigated. For example, the role of CD34+ fibrocytes in the elaboration of matrix metalloproteinase-9 and collagen subtypes I and III offers new insight into myxomatous mitral valve degeneration (Barth et al 2005). Moreover, cellular molecular mechanisms represent the vast frontier of discovery in the pathogenesis of valvular heart diseases.

The widely recognized drug-related valvulopathies drive significant areas of research. The definitive role of the serotonin receptor 5-HT_2B and the serotonin transporter (5-HTT) proteins awaits full explanation in the fenfluramine, ergotamine, and other related serotonergic valvulopathies. The hypothesis that activation of 5-HT_2B may be directly responsible for valvular heart disease remains a hypothesis, and since no reproducible animal model exists, it has been impossible to precisely reproduce this condition under controlled circumstances. The epidemiology of fenfluramine-related valvulopathy, by far the most widely prescribed agent, showed a low-frequency occurrence and raised the issue of individual susceptibility, the role of pre-existing or comorbidities, and perhaps the overly simple assumption that activation of one receptor is sufficient for development of valvulopathy. Ongoing investigations into the activation pathways of 5-HT-coupled G-protein receptors and subsequent transduction events inside critical cells in the process of valve tissue remodeling, such as ERK 1/2 phosphorylation and up-regulation of TGF-β1, will have important implications. The evidence that TGF-β plays a pivotal role in the process of valve injury and repair is mounting, opening avenues to new hypotheses that tie many well-established and incipient observations about valvular heart disease together. Indeed the biological effects of TGF-β1 appear to include stimulation of collagen-producing cardiac fibroblasts for remodeling of myocardium after infarction, activation of processes related to dilated and hypertrophic cardiomyopathies, and cardiac valvular disease. In addition to understanding mechanisms of injury and repair, functional cellular-based investigations provide insight into possible new treatments. The frontiers for the treatment and management of cardiac valvular disease are exciting and challenging and include targeted therapies to prevent or slow valvular injury as well as varied approaches to valve repair and replacement, especially involving artificial/mechanical devices and bioengineered valve tissues.

This review presents a summary of the biology and pathology of the leading human heart valve diseases, correlated with known entities in laboratory animals, and the ongoing investigations and novel approaches to understanding the pathogenesis of valvular disease mechanisms in animal models. Animal models continue to be described, but most are only in the preliminary stages of characterization. Modeling of valvular heart disease has historically been under used, largely owing to lack of success in fully reproducing specific heart valve disease syndromes in a useful way and key species-specific differences in biology and pathology of valves. With greater emphasis on cardiac disease overall as the confirmed top cause of human mortality in the Western world, the attention to valve-related disease issues will continue to grow. Better animal models, increased surveillance in preclinical toxicology of pharmaceuticals, and a deeper understanding of the molecular mechanisms of injury and repair are needed.