Cardiac disease in patients with mucopolysaccharidosis: presentation, diagnosis and management

Cardiac findings were not prominent in the earliest reports of Hurler syndrome (Hurler 1919) and Hunter syndrome (Hunter 1917). Beginning in the 1940-50s, detailed post-mortem cardiac studies on individuals clinically diagnosed with MPS (previously referred to as “gargoylism”) were reported by several investigators (Emanuel 1954; Lindsay 1950; Strauss 1948). These early morphologic studies occurred well before the biochemical basis of MPS disorders was established. It was only in 1965 that Victor McKusick, trained as a cardiologist and the father of modern medical genetics, established the classification of the MPS disorders based upon patterns of mucopolysacchariduria, a system that continues, with few changes, to be in use today (Kaplan 1969; McKusick et al. 1965; McKusick 1969). The first MPS cardiac studies to include supporting biochemical data are those of Krovetz et al. (Krovetz et al. 1965), Renteria et al. (Renteria et al. 1976) and Brosius and Roberts (Brosius and Roberts 1981).

Because of the rarity of the disorders (Table 1), cardiac studies involving large numbers of individuals with a particular MPS are uncommon. Nevertheless, the prevalence and severity of cardiovascular disease in individuals with MPS (especially MPS I, II and VI) is strikingly high, occurring in 60-100% of those studied (Chen et al. 2005; Dangel 1998; Fesslová et al. 2009; Leal et al. 2010; Martins et al. 2009; Pastores et al. 2007; Wippermann et al. 1995). Cardiac valve involvement appears more common in those syndromes in which dermatan sulfate catabolism is deranged (MPS I, II and VI but not MPS III and IV) (Chen et al. 2005; Dangel 1998; Fesslová et al. 2009; Leal et al. 2010).

Within a particular type of MPS, cardiac pathology generally develops earlier in life for individuals with more rapidly progressing types of MPS (e.g. Hurler syndrome) (Brosius and Roberts 1981; Hirth et al. 2007) and may be delayed in slowly progressing forms (e.g. Hurler-Scheie and Scheie syndromes) (Soliman et al. 2007). Nonetheless, cardiac disease is a uniformly progressive process (Chen et al. 2005; Dangel 1998; Fesslová et al. 2009; Giugliani et al. 2007; Krovetz et al. 1965; Leal et al. 2010; Mohan et al. 2002; Muenzer et al. 2009b) with both incidence and severity increasing over time (Fesslová et al. 2009). Despite this, clinical signs and symptoms of cardiac involvement are uncommon leading to an underestimate of the true prevalence of cardiac disease. For example, only six of 26 patients with documented echocardiographic disease presented with cardiovascular signs or symptoms (Dangel 1998; Gross et al. 1988; Leal et al. 2010), a finding which underscores the importance of a complete cardiac evaluation after establishing a diagnosis of MPS.

Progressive cardiac valve pathology is the most prominent and uniform cardiac manifestation (60-90%) of patients with MPS (Fesslová et al. 2009; Wippermann et al. 1995). Cardiac valve thickening with associated dysfunction has been reported in more than 80% of patients with MPS I (including slowly progressing phenotypes) (Pastores et al. 2007; Thomas et al. 2010), 57% of patients with MPS II (Wraith et al. 2008) and in all individuals with MPS VI except for the most slowly progressing ones (Azevedo et al. 2004; Brooks et al. 2005; Scarpa et al. 2009).

Most studies have reported that valvular regurgitation, rather than stenosis, is more common and that the mitral valve, rather than the aortic, is most commonly affected. In general, left-sided valves (mitral & aortic) are more severely affected than those on the right side of the heart (tricuspid & pulmonary) (Dangel 1998; Fesslová et al. 2009; Krovetz et al. 1965; Rigante and Segni 2002; Thomas et al. 2010). The mitral valve leaflets are markedly thickened and cartilage-like, with particularly thickened edges (Fig. 1). The subvalvular apparatus of the mitral valve develops shortened chordae tendineae and thick papillary muscles resulting in dysmorphic and poorly mobile leaflets (Fig. 1) (Tan et al. 1992; Yano et al. 2009). Calcific deposits commonly occur in the mitral annular region (Johnson et al. 1981; Renteria et al. 1976). The aortic valve presents a similar picture of progressive valve thickening and dysfunction (Fig. 1).

These pathologic alterations result in regurgitation (insufficiency) and/or stenosis, both of which are easily identified by two-dimensional echocardiography and Doppler examination (Johnson et al. 1981; Maganti et al. 2010) (Fig. 2). Valvular stenosis or regurgitation may lead to left atrial and/or left ventricular volume overload, left ventricular dilatation, left ventricular hypertrophy (LVH), and ultimately to systolic and diastolic dysfunction. In a retrospective case series of children with various forms of MPS, LVH and diastolic dysfunction emerged at an early stage, whereas left ventricular dilation and systolic dysfunction occurred at older ages and later disease stage (Leal et al. 2010) (Fig. 3).

Coronary artery narrowing and/or occlusion has been described in individuals with all types of MPS (Braunlin et al. 1992; Brosius and Roberts 1981; Danks et al. 1972; Factor et al. 1978; Kettles et al. 2002; Kurihara et al. 1996; Renteria et al. 1976; Shimamura et al. 1976; Tan et al. 1992), but is most common in MPS I and MPS II. Diffuse intimal proliferation from GAG deposition within large epicardial coronary arteries can occur early, especially in rapidly progressing MPS I (Brosius and Roberts 1981), causing high-grade narrowing (Fig. 4). The lumens are nearly always central in location. Occlusion within coronary arterioles (rather than in the large epicardial coronary arteries) and a large left ventricular apical aneurysm have been described in an individual with MPS VI suggesting that different types of coronary involvement may occur in the various MPS syndromes (Oudit et al. 2007). A similar apical aneurysm has been reported in MPS II (Kettles et al. 2002), although a description of the coronary arterioles was not provided.

The great vessels in individuals with MPS may have increased wall thickness and may either be narrowed (Taylor et al. 1991; Wraith et al. 2008) or dilated (Hinek and Wilson 2000). Diffuse narrowing of the thoracic and abdominal aorta (Taylor et al. 1991) has been reported to occur in 30% of a small series of individuals with MPS I and II. Narrowing may occur at the aortic isthmus, a location typical for coarctation (Braunlin et al. 2000), and may require surgical intervention for relief of significant obstruction. Although systemic hypertension due to arterial narrowing is common among individuals with MPS I and II (Taylor et al. 1991; Wraith et al. 2008), it has been less frequently noted in other types of MPS (including MPS VI) (Giugliani et al. 2007).

Dilation of the ascending aorta and markedly reduced aortic elasticity has been reported in MPS I (Nemes et al. 2008). This could be attributed to the downstream effects of GAGs on the assembly of tropo-elastin, resulting in elastin that is both decreased in content and abnormal in structure (Hinek and Wilson 2000). Studies in MPS I and MPS VII animals have revealed that aortic dilation is likely due to degradation of elastin by matrix metalloproteinase-12 and/or cathepsin S (Ma et al. 2008; Metcalf et al. 2010). Aortic root replacement has not been reported for MPS patients to date.

Conduction abnormalities and sinus tachycardia have been reported in 44% of the patients with MPS VI and in 7% of a subset of the Hunter Outcome Survey of MPS II patients, respectively (Azevedo et al. 2004; Wraith et al. 2008). Several case reports have documented the development of complete atrioventricular block (MPS II, III and VI) with subsequent need for cardiac pacing and sudden death (Dilber et al. 2002; Hishitani et al. 2000; Misumi et al. 2010). Fibrosis of the conduction system with GAG infiltration was seen in one individual with MPS VI who died suddenly (Keller et al. 1987).

The heparan-, dermatan-, chondroitin- and keratan-sulfated GAGs are normal components of cardiac valves and the great vessels (Grande-Allen et al. 2003; Gupta et al. 2009; Latif et al. 2005). Recent studies have implicated alterations in proteoglycan (GAG) metabolism in pathological processes such as myxomatous mitral valves (Grande-Allen et al. 2003; Rabkin et al. 2001), aortic aneurysm (Theocharis et al. 2003), and atherosclerotic vasculature (Nakashima et al. 2008). Dermatan-sulfated GAGs are a prominent component of normal cardiac valve tissue (Latif et al. 2005) so it is not surprising to find a strong association between MPS I, II and VI (that result in the accumulation of dermatan-sulfated GAGs) and cardiac valve disease (Dangel 1998; Leal et al. 2010). Additionally, deep within the cardiac valves of all MPS syndromes, lie GAG-laden cells that have variously been called ‘clear’ cells, ‘Gargoyle’ cells or ‘Hurler’ cells. Recently, these cells have been identified in MPS I as being activated valvular interstitial cells engaged in attempted, but ineffective, valve repair (Braunlin et al. 2010). The mechanisms by which the accumulated heparan-sulfated proteoglycans and attendant vascular interstitial cells affect the vasculature of the great vessels and coronary arteries in MPS I, II, and III (the heparan-sulfate accumulating MPS disorders) remains unknown. One hypothesis for the role of GAGs in cardiac disease could be that GAGs induce inflammation by activating the Toll-like receptor 4 pathway, leading to upregulation of degradative proteases (Simonaro 2010).

The physical examination not only of young children, but also of adults with MPS is challenging due to both physical and, sometimes, intellectual patient limitations. A complete examination includes right upper as well as lower extremity blood pressures (to exclude aortic obstruction), pulse rate, palpation of pulses in all extremities, as well as careful auscultation of the heart and lungs. Other aspects of a complete cardiac examination, evaluating jugular venous distention or hepatosplenomegaly, may be unrewarding due to shortened necks and the presence of GAG-laden and enlarged visceral organs. Dysostosis multiplex and joint stiffening may complicate blood pressure measurements. Auscultation is often suboptimal due to pronounced upper airway noises. While the cardiac sounds are generally easily heard, the absence of cardiac murmurs does not exclude the presence of significant cardiac disease. Mild to moderate degrees of valvular stenosis and/or regurgitation can be present without significant murmurs.

Because of its versatility, ease of performance and cost-effectiveness, transthoracic two-dimensional echocardiography is the primary imaging modality used to evaluate cardiac anatomy and function in patients with MPS (Aguilar et al. 2008). Echocardiography provides reliable measurements of ventricular chamber dimensions (both systolic and diastolic) and left ventricular posterior and septal wall thicknesses, which can be normalized for body surface area. Accurate measures of ventricular function (shortening fraction & ejection fraction) can be determined. Doppler interrogation of cardiac valves allows accurate grading of valve stenosis and regurgitation as well as estimation of cardiac output and right ventricular/pulmonary artery pressures (Aguilar et al. 2008). Despite its benefits, echocardiography provides only limited information in the assessment of coronary artery anatomy and pathology. Transthoracic echocardiography can usually be performed without sedation if the child is allowed to remain in his parent’s arms or if he can be distracted by videos or toys. In some instances, however, transthoracic echocardiography may be suboptimal due to limited tolerance for examination, abnormal cardiac position, scoliosis, or hyper-aerated lung fields. Transesophageal echocardiography is not routinely employed in those with MPS as it requires monitored sedation or general anaesthesia, which may be poorly tolerated because of the airway complications that are not infrequent in this patient population.

Magnetic resonance imaging (MRI) and computed tomography (CT) are advanced techniques not routinely used for cardiac evaluation in children. Although MRI is a reliable tomographic technique for assessing cardiac anatomy and function (with quantification of volumetrics and regurgitation), its application is limited by heart rate constraints, slow acquisition times, and the need for sedation, and may be contraindicated in individuals with implanted devices (Aguilar et al. 2008). Cardiac CT angiography, advantageous with respect to acquisition speed and resolution, is limited by relatively large radiation exposure and suboptimal assessment of valve morphology and hemodynamics (Aguilar et al. 2008).

Evaluation for the presence of coronary arteriopathy in the MPS disorders is problematic because the diffuse nature of coronary involvement differs substantially from typical atherosclerotic coronary disease. While the latter is believed to be reliably excluded by a normal CT angiogram (Taylor et al. 2010), this may not be true in MPS coronary disease. Coronary angiography is therefore indicated in those suspected of having coronary artery pathology (Winchester et al. 2010), but may miss significant disease due to the diffuse nature of the process in MPS (Braunlin et al. 1992). Intravascular ultrasound has successfully identified significant coronary artery narrowing in adults with Fabry’s disease (Kovarnik et al. 2008), but has not been reported in the MPS disorders and remains primarily a research tool. The value of non-invasive dobutamine or adenosine stress testing combined with echocardiography or MRI imaging is currently under study in individuals with MPS.

Standard 12-lead electrocardiography (ECG) is the second most important technique used to evaluate the hearts of individuals with MPS. Assessment of heart rate and rhythm as well the determination of conduction times (PR interval, QRS duration, corrected QT interval) are all important clues to the health of the underlying conduction system. In instances where symptoms suggestive of either conduction block or arrhythmia are present, more prolonged monitoring (Holter or event monitor) is required for identifying arrhythmias and heart rate variability. Although rarely detected on routine ECG, one should pay attention for signs of ischemia in this patient population.

Individuals with MPS should undergo physical examination, echocardiography and 12-lead ECG at the time of diagnosis, followed by regular monitoring (every 1 to 2 years for MPS I and VI, or 1 to 3 years for MPS II) thereafter in order to provide appropriate diagnosis and management (Giugliani et al. 2007; Muenzer et al. 2009a, b). Additional cardiac evaluations should be considered before major operative interventions. Complete evaluation includes right arm and leg blood pressure measurements, careful auscultation, performance of a full transthoracic two-dimensional and Doppler echocardiogram, 12-lead ECG, and Holter monitoring as indicated (Giugliani et al. 2007; Muenzer et al. 2009a, b). Performance and interpretation of echocardiography should be standardized (in terms of equipment and settings) and provided by individuals familiar with expected pathological findings in patients with MPS (Fig. 6).

In the adult population, standard pharmacological and surgical management has been driven by cardiac pathophysiology regardless of MPS type. One longitudinal cohort study reported 33% of 48 patients with MPS received standard drug therapy for hemodynamic complications caused by valvular lesions or heart failure (Fesslová et al. 2009). In the pediatric population, due to the absence of large-scale, long-term pharmacological studies, clear recommendations regarding pharmacologic management have not been defined (O’Connor et al. 2010). Afterload reduction with angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers has been reported as useful in small series of non-MPS children with aortic or mitral regurgitation, hypertension or heart failure (O’Connor et al. 2010). Although these drugs are frequently used, formal recommendations have yet to be made regarding their use in children, either with or without MPS (O’Connor et al. 2010).

Progressive deposition of GAGs within the cardiac valves can lead to regurgitation or stenosis severe enough to require surgical intervention. Cardiac valve surgery is among the procedures frequently performed in patients with MPS (Arn et al. 2009). Successful valve repair or replacement has been reported in individuals with all types of MPS (Table 2). Mitral valve replacement was performed in MPS I (Kitabayashi et al. 2007; Kraiem et al. 2001), MPS II (Antoniou et al. 2009) and MPS VI (Marwick et al 1992); a 6-year-old with MPS III (Muenzer et al. 1993) has undergone mitral valvuloplasty. Successful aortic valve replacement has been reported in adults with slowly progressing MPS I (Masuda et al. 1993), MPS IV (Barry et al. 2006; Nicolini et al. 2008; Pagel and Almassi 2009) and MPS VI (Wilson et al. 1980). Due to pathological abnormalities in the native pulmonary root, the Ross procedure, where the aortic valve is replaced by the patient’s own pulmonary valve, is contraindicated in individuals with MPS (Barry et al. 2006). Combined aortic and mitral valve replacement was successfully performed in adults (Butman et al. 1989; Fischer et al. 1999; Minakata et al. 1998) and in teens as young as 12 years of age (Goksel et al. 2009) with MPS I and in adults with MPS II (Joly et al. 2004) and MPS VI (Hachida et al. 1996; Tan et al. 1992). Fesslová et al. have reported favorable mitral and aortic valve replacement in three patients with Scheie syndrome and one with slowly progressing Hunter syndrome ranging in age from 13–53 years (Fesslová et al. 2009). Other successful cardiac procedures include coronary artery bypass surgery in a 56-year-old male with slowly progressing MPS I (Minakata et al. 1998), closure of an outlet type of ventricular septal defect in a child with MPS III (Kourouklis et al. 2007), and repair of ‘coarctation’ of the aorta after HSCT in a 3-year-old boy with rapidly progressing MPS I (Braunlin et al. 2000).

While the long-term outcome is generally favorable (Hachida et al. 1996; Marwick et al. 1992; Tan et al. 1992; Wilson et al. 1980), candidacy for surgical intervention must be based on general health status and potential risks posed by intubation and post-operative care. Craniofacial abnormalities, excessive oral pharyngeal and upper airway tissue, a short stiff neck, and unstable atlantoaxial joint (with risk of spinal cord injury) (Giugliani et al. 2007) may combine to complicate intubation and airway management in select patients (Walker et al. 1994). Intubation with use of fiber-optic bronchoscope and expertise in dealing with uncommon airway issues in this patient population is often necessary. Endocarditis prophylaxis is currently recommended only in patients with a history of endocarditis, certain residual postoperative lesions, or a prosthetic valve (Maganti et al. 2010; Wilson et al. 2007).

HSCT has been successfully employed for metabolic correction of MPS disorders for the past 30 years (Hobbs et al. 1981; Krivit et al. 1984). While this therapeutic approach has, in the rapidly progressing form of MPS I, significantly increased long-term survival and arrested neurologic deterioration (Orchard et al. 2007), its beneficial effects in the remaining MPS disorders have not been proven (Prasad and Kurtzberg 2010a, b). Associated risks of HSCT are high (Boelens et al. 2009; Prasad and Kurtzberg 2010a, b) and, as a consequence, HSCT is considered standard of care for only the most rapidly progressing forms of MPS I (Prasad and Kurtzberg 2010a, b).

Long-term metabolic correction by HSCT results in preservation of cardiac function and regression of cardiac hypertrophy (Braunlin et al. 2003; Gatzoulis et al. 1995; Guffon et al. 2009; Krivit et al. 1984; Viñallonga et al. 1992). Cardiac valvular thickening is less likely to resolve (Herskhovitz et al. 1999; Malm et al. 2008) and left-sided valvular stenosis or regurgitation may remain unchanged (Yamada et al. 1998) or progress (Braunlin et al. 2003). Coronary artery narrowing and occlusion from myointimal proliferation in rapidly progressing MPS I appears to be arrested after successful HSCT (Braunlin et al. 2001), a factor which may play a significant role in prolonged survival in these patients (Orchard et al. 2007). The progression of cardiac valve pathology, despite successful engraftment after HSCT, is similar to disease progression in other relatively avascular sites such as bone and joint (Taylor et al. 2008).

ERT, supplying exogenous human recombinant enzyme by regular intravenous infusion, is approved for three types of MPS; laronidase for MPS I (Wraith et al. 2004), idursulfase for MPS II (Muenzer et al. 2006) and galsulfase for MPS VI (Harmatz et al. 2004; Harmatz et al. 2005). Cross-correction with ERT is achieved in many organs and tissues, reflected by lower urinary GAG levels and symptomatic improvement. While galsulfase infusion has been shown to produce improvement in endurance (walk and stair climb test, sustained up to 5 years), pulmonary function, joint pain and stiffness, and growth and organomegaly (Harmatz et al. 2004; Harmatz et al. 2005; Harmatz et al. 2006; Harmatz et al. 2008), the outcome of ERT is less favorable for poorly vascularised or “privileged” sites (central nervous system, cornea, articular cartilage, and cardiac valves).

The effect of ERT alone on cardiac structure and function in patients with MPS is documented in only a few studies (Braunlin et al. 2006; Fesslová et al. 2009; Okuyama et al. 2010; Scarpa et al. 2009; Wraith et al. 2007). As with HSCT, long-term ERT preserves, and may even improve (Hirth et al. 2007), systolic ventricular function and results in resolution of ventricular hypertrophy in adults and children with MPS I and II (Braunlin et al. 2006; Okuyama et al. 2010; Wraith et al. 2007). However, ERT does not appear to prevent progression of left-sided cardiac valve thickening and dysfunction (stenosis, regurgitation) in MPS I, II and VI. Stabilization of valvular disease in select patients over a limited study period has been observed (Braunlin et al. 2006; Fesslová et al. 2009; Okuyama et al. 2010; Scarpa et al. 2009). ERT has been used as an adjunct in MPS I to lower the GAG burden prior to HSCT (Tolar and Orchard 2008) and, recently, galsulfase has been found to normalize mild elevations in GAG levels found in a single fully-engrafted patient with MPS VI 20 years after successful HSCT (Whitley and Utz 2010). Whether these protocols will have beneficial cardiac effects remains unstudied.

In contrast to the young age at which most individuals with MPS undergo HSCT, ERT is usually begun in the older child (often after 5 years of age) with established cardiac valvular disease. It is possible that earlier use of ERT may prevent or ameliorate cardiac valve thickening and dysfunction in MPS as has been shown by two recent sibling studies (Gabrielli et al. 2010; McGill et al. 2010) (Table 3). Each older sibling with either MPS I or MPS VI shared the same enzyme mutation as his younger sibling and differed only in the age at diagnosis. The older sibling in each instance, starting with ERT at 4 or 5 years of age, retained the classic phenotypical appearance of MPS as well as thickening, regurgitation or dysplasia of the mitral valve (Table 3). By contrast, each younger sibling, treated with ERT from 8 weeks and 5 months of age and onwards, respectively, appeared phenotypically normal and had virtually no cardiac pathology by echocardiography either before, or after, ERT (Table 3). These data underscore the importance of newborn or early screening for MPS as there appears to be a window in early infancy when ERT may prevent many of the end-organ manifestations of MPS.