Surgical Management of Valvular Heart Disease in Mucopolysaccharidoses: A Review of Literature

Mucopolysaccharidoses are extremely rare diseases that are frequently presenting with structural heart problems of the aortic and mitral valve in combination with myocardial dysfunction. In a substantial proportion, this leads to heart failure and is a leading cause of death in these patients. As this glycosaminoglycan degradation defect is associated with other conditions strongly influencing the perioperative risk and choice of surgical technique, multidisciplinary planning is crucial to improve short- and long-term outcomes. The extensive variance in clinical presentation between different impaired enzymes, and further within subgroups, calls for personalised treatment plans. Enzyme replacement therapies and bone marrow transplantation carry great potential as they may significantly abrogate the progress of the disease and as such reduce the clinical burden and improve life expectancy. Nevertheless, structural heart interventions may be required. We reviewed the existing literature of the less than 50 published cases regarding surgical management, technique, and choice of prostheses. Although improvement in therapy has shown promising results in protecting valvar tissue when initiated in infancy, concerns regarding stability of this effect and durability of biological prostheses remain.


Introduction
Mucopolysaccharidoses (MPS) describe a family of rare inherited lysosomal storage diseases characterised by specific deficiencies in lysosomal enzymes necessary for the degradation of glycosaminoglycans (GAGs). Despite cardiac involvement being common in MPS, leading to structural heart disease in 49-90% of cases [1], current evidence for surgical management in this cohort is limited to approximately three-dozen case reports. This review discusses the existing literature, focussed on surgical considerations, therapy options including choice of prostheses, timepoint of intervention, and results in patients with this rare disease.

Incidence
MPS subtypes are differentiated biochemically by associated lysosomal enzyme deficiency and resulting GAG accumulation. Currently, seven distinct types of MPS have been identified (Table 1); MPS III and IV are comprised of four and two more subtypes. Additionally, MPS I is classified into three subtypes, representing the spectrum of severity of the clinical manifestation (Hurler syndrome [most severe], Hurler-Scheie syndrome, and Scheie syndrome [least severe]) [2]. With the exception of MPS II (X-linked recessive), all MPS follow an autosomal recessive inheritance pattern [1].

Pathophysiology
The deficiency or malfunction of enzymes present in MPS are insufficient for a correct degradation of glycosaminoglycans, long linear polysaccharides with highly hydrophilic characteristics produced in the Golgi apparatus or by integral membrane synthases. This disturbance in the catabolism leads to intra-lysosomal, thus intracellular accumulation of GAG in the cardiac tissues [11,24]. Dermatan sulphate is regularly found in heart valves, tendons, blood vessels, pulmonary tissue and skin, and a disbalance of this type of GAG is especially associated with structural heart disease, which is reflected by higher occurrence of structural heart disease in MPS types I, II, VI, more than types III and IV [25].

Valve Disease
Progressive cardiac valvular disease is a prominent cardiac feature of MPS and was first described in 1960 [26]. Although all cardiac valves may be affected, left-sided valve issues are the most common ( Table 2) with mitral valve and aortic valve being the predominantly affected structures [11]. Further, these abnormalities can present as combined, complex structural disease and may additionally be associated with left ventricular hypertrophy [27].
In valvular interstitial cells, which normally are responsible for valve growth and repair, MPS leads to activation and development of large, GAG-laden "Hurler" cells [28]. Additionally, an inflammatory and macrophage response is induced, causing further tissue dysfunction [29]. These GAG-triggered processes lead to valve thickening, diminished leaflet mobility, and coaptation deficiency. Valve leaflet thickening has been reported to affect 60-90% of MPS patients [27]. Severe cases can also present with shortened chordae tendineae [30]. The combination of these processes can lead to mitral regurgitation, mitral stenosis, aortic regurgitation and aortic stenosis and combinations. Increased dermatan sulphate accumulation further is associated with myxomatous mitral valve degeneration [31].

Myocardium and Large Vessels
The intracellular accumulation of GAGs leads to formation of large Hurler cells containing clear vacuole of mucopolysaccharides which also can be found in the myocardium. The triggered inflammatory response leads to increased collagen reaction of these cells with corresponding tissue fibrosis, thus, leading to a restrictive impaired ventricular function, predominantly in the left ventricle. The resulting echocardiographic findings besides diastolic dysfunction are left ventricular hypertrophy and left ventricular remodelling [36]. The structural changes in the myocardium further can lead to alteration in the conduction system clinically presenting with progressive AV block.
The same Hurler cells can be found in large vessels. Frequent concomitant appearances are enlarged aorto-ventricular junction, valsalva and sino-tubular junction, which is found in 30-41% of patients with MPS [37].

Coronary Artery Disease
In addition to valvular and myocardial infiltration, deposition of GAGs can also occur in the vascular smooth muscle cells, leading to diffuse, concentric narrowing and consecutive stenosis of the coronary arteries, and additional intima hypertrophy has been described [38]. Consequently, an accelerated course of coronary artery stenosis possibly leading to ischaemic heart failure has been reported to occur in MPS patients. Clinical symptoms of this process may remain absent until a later stage. Contradictory reports on association with different types of MPS have been published. While historically coronary artery disease is known to be associated with MPS I, one report found histopathological alterations predominantly in non-Hurler MPS [39]. Even more, exclusion of coronary artery disease is recommended prior to surgery [40].

Haematopoietic Stem Cell Transplantation
First performed in patients with MPS I-H in 1981, haematopoietic stem cell transplantation (HSCT) is considered standard of care for these patients [41]. HSCT has been shown to improve symptoms in MPS I-HS, II, IV-A, VI, and VII, though there are high associated risks, particularly from graft-versus-host disease [42]. Therefore, HSCT is generally reserved for the more severe forms of MPS.
Despite clear improvements in life expectancy and neurosensory outcomes post-HSCT, long-term follow-up has revealed that approximately one-third of MPS I-H patients still showed progressive mitral and aortic insufficiency despite HSCT [43]. Thus, valve disease remains a significant disease burden despite treatment.
However, the efficacy of ERT in stabilising valve disease in the MPS types is not as apparent. It is postulated that the avascularity of the cardiac valves diminishes the efficacy of ERT, making it ineffectual in halting the progression of valve disease [48]. Additionally, an antibody response to ERT has been shown to occur in some patients, reaching levels that could potentially affect the efficacy of ERT [50].
Ten-year follow-up of MPS I patients on ERT revealed that although two-thirds of patients achieved stabilisation of valve disease, approximately one-third of patients still had valvular deterioration [45]. Progression of mitral valve disease has also been noted in 10-year follow-up of MPS II patients on ERT [51]. Three-to-six-year follow-up in MPS IV-A patients suggested that ERT has little effect on valve function despite improvements in ventricular hypertrophy [19]. Finally, a two-year follow up of MPS VI patients showed significant progression in AR and no improvement in valve stenosis despite ERT treatment in the subset of patients who began ERT after childhood [48]. Current evidence suggests that ERT begun in infancy has a protective effect on valvar tissue [52][53][54] compared to treatment begun later in childhood [55][56][57]. However, the durability of these results has yet to be confirmed. As valvular dysfunction remains present and, in some cases, progressive despite ERT in patients across all MPS types, structural heart surgery eventually has to be considered.

Gene Editing Therapy
A variety of gene transfer and gene therapy approaches are currently in preclinical and clinical phase with 17 registered clinical trials, of which some have completed treatment with successful first safety endpoints and now progressed to a long-term follow-up phase. Further investigations on their clinical efficiency are to be expected.

Pre-Operative Considerations
As a rare, complex multisystemic disease, surgery should be performed in specialist centres experienced with MPS patients. A well-planned multidisciplinary approach is necessary with experienced anaesthetic input and a joint cardiac disease multidisciplinary team meeting is recommended. The potential risks of the procedure and anaesthesia should be weighed against the benefits and extensively discussed with the patient and relatives [58].
Other considerations include pre-operative ENT examination with naso-endoscopy or MRI assessment of the spine, brain, and airways. Additionally, coronary angiography should be performed prior to any cardiac procedure due to the increased prevalence of diffuse coronary disease in MPS patients [40].

Respiratory Tract
Respiratory symptoms are apparent across all MPS types and are important drivers of mortality [13]. Besides upper and lower airway obstruction, predisposing patients to frequent respiratory tract infections and obstructive symptoms [14], restrictive lung disease is also common, resulting from MPS manifestations such as a small thoracic cage, kyphoscoliosis, pectus carinatum, and displacement of the diaphragm superiorly due to hepatosplenomegaly [15]. As this adds to the perioperative risk, a preoperative lung function study is essential prior to cardiac surgery.

Cardiac Diagnostics and Symptom Assessment
It is important to understand that a large proportion of patients with MPS are limited in their exercise capacity due to musculoskeletal or spinal issues with a substantial number being wheelchair-bound. Therefore, it is often so that they only become symptomatic with advanced valve, myocardial or coronary artery disease. It is, therefore, important to closely monitor these patients in the cardiology clinic for subclinical advanced heart disease. MPS not only affects the heart valves, it also can lead to ventricular dysfunction with fibrosis, coronary artery disease, and electrophysiological symptoms. These three aspects of cardiac affection in MPS patients are important when planning a structural procedure, as concomitant severe coronary artery stenosis or restrictive left ventricular function will strongly influence the perioperative risk. Aortic abnormalities can also occur with MPS with a risk of dissection. Routine investigations with echocardiography and ECG should be standard of care. In situations where it is felt that the heart disease has advanced significantly so, detailed investigations are indicated. These include transthoracic echocardiography, coronary CT and CT angiography, ECG, and, if possible, cardiac magnetic resonance scanning. Coronary artery disease in these patients does not usually mimic atherosclerotic patterns, and, therefore, it is more useful to perform functional studies such as stress-echocardiography or myocardial perfusion scanning. Multidisciplinary discussions regarding consideration of intervention or surgery should always take place.

Anaesthetic Considerations
Abnormal anatomical features and tissue morphology due to GAG accumulation can pose challenges regarding airway control, intubation, intravenous access, and cervical spine stability. Although surgery is frequent in MPS (MPS I 81% at least one procedure) [59], patients are at increased risk when undergoing general anaesthetics procedures. Difficult and failed intubations are common [60] and should be carried out by experiences anaesthetists familiar with MPS patients with adequate intensive care and ENT backup.
Intubation and airway difficulties are due to narrowing of the upper airways with thickening of laryngopharyngeal structures, including macroglossia, tonsillar enlargement, and narrowed nasal passages [61]. Lower airway stenosis can involve malformed tracheal cartilage, abnormal vocal cords and airway oedema from recurrent respiratory infections [62]. Additionally, the short neck, raised maxilla, and small mouth opening due to stiff temporomandibular joints may make laryngoscopy difficult [61].
Additionally, cervical spine stability should be considered. Particularly in MPS IV, patients can present with odontoid hypoplasia, which is a considerable risk factor for atlantoaxial instability. If not properly managed, this can result in spinal cord compression, eventually leading to paralysis or death [63]. Therefore, minimising head and neck movement by manual stabilisation is advised during induction and intubation [58]. The head should maintain a neutral position; this can make direct laryngoscopy difficult. Thus, fibreoptic intubation, or an angulated video laryngoscope may be used [64].

Post-Operative Considerations
Careful post-operative planning is crucial; there is an increased rate of post-operative mortality with MPS patients [59]. Intensive care unit admission is necessary; extracorporeal membrane oxygenation or intra-aortic balloon pump therapy may be required for postoperative ventricular dysfunction [65]. Extubation should be performed only when the patient is fully conscious and coughing [58]. Post-extubation laryngeal oedema can occur, and steroid administration before planned extubation has been suggested to reduce this complication [66,67]. Additionally, steroid cream can be applied to reduce oral mucosal and tongue swelling [58]. In case of serious airway concerns, post-operative elective tracheostomy may has to be considered; this should be discussed with the patient preoperatively [68]. Furthermore, precautionary equipment for emergency tracheostomy and fibreoptic reintubation should be available at extubation should complications arise [66]. The best time point for heart valve replacement can be a challenging decision in MPS patient. Complicating factors, cardiac and non-cardiac MPS symptoms must be considered, best to be assessed in an interdisciplinary conversation considering all available medical history as well as recent cardiac and respiratory examinations. Clinical manifestation and progression show strong variation between different types of MPS, additionally individual differences between patients of the same type of the disease can be observed. Therefore, a personalised treatment plan is required for these patients.

Prosthesis Type
Most reported valve replacements involved mechanical rather than bioprosthetic implants. This seems an obvious choice as mechanical valves are preferable in the relatively young MPS patient population; bioprosthetic prostheses have a limited durability due to leaflet deterioration, particularly true for younger patients [69]. Additionally, it is possible, given the pathophysiology of MPS, that bioprosthetic degeneration could be affected by GAG deposits leading to even quicker degeneration of the bioprosthetic leaflets. On the other hand, mechanical valve replacement requires patients to be on full anticoagulation with associated annual significant complication incidence of 1-2% in a non-MPS population, which conceivably could be higher in the MPS population.
Although current literature suggests mechanical valves as implants, under certain circumstances biological prostheses might be considered. Increasing experience in new therapy options such as ERT and optimised treatment protocols have the potential to enable prosthesis degeneration similar to that of non-MPS patients. This is particularly true for patients presenting at later stages in life, thus milder forms of MPS. For patients with neurological impairment due to MPS, or other reasons of impaired suitability for oral anticoagulation, biological implants might be considered. Furthermore, the possibility of an interventional valve-in-valve implantation has the potential to prolong the interval to reoperation in cases where a sufficiently sized valve can be used, as also mentioned by Dostalova et al. [70,71]. Most centres treating these complex patients would still shy away from using non-mechanical valve substitutes at present, due to a lack of supportive data for this strategy.
Regarding the effect of ERT on progress of natural valve changes, controversial experiences and considerations have been reported. During an elective aortic valve replacement, despite significant MS on echocardiography, Torre et al. elected not to perform double valve replacement on a 40-year-old MPS VI patient. The authors reasoned that, since the patient was on ERT for three years, the MS would not progress further [57]. Additionally, the patient presented with a small mitral annulus, which would warrant either an annular enlargement or a small prosthetic valve. The former is complicating the procedure significantly with corresponding effect on risk and morbidity, whereas the latter incurs risk of patient-prosthesis mismatch, full anticoagulation, and residual gradient. On the other hand, the patient was at risk of having to undergo a further valve replacement procedure, should the mitral valve disease progress. This scenario has been highlighted in several reports where progressive valve disease has led to a second valve replacement operation despite ERT [72,73]. Given the variable rate of disease progression and response to ERT between patients, deciding between conservative or aggressive valve replacement can be difficult. Risks and benefits of both courses of treatment should be discussed in a multidisciplinary setting and with patients. Further long-term analysis of valve disease progression under ERT is expected to provide further evidence. An overview on double valve replacement reports can be found in Appendix A Table A2.
The systemic nature of MPS can often cause multi-valvular incompetence. Poor annular tissue quality, often described as "friable", "hard", and "not pliable", has been reported to complicate valve implantation in MPS patients [56,65,74,75]. Kitabayashi et al. suggested the use of an equine pericardial patch between the atrial wall and the prosthetic ring to reinforce the suture line [65]. Similarly, Bell et al. found anchoring of the prosthesis on the annular tissue challenging and suggested the use of felt pledgets to reinforce the prosthetic sutures [56]. These adjuncts may aid in preventing valve dehiscence and paravalvular leakage, thereby reducing the risk of valve failure and redo surgery.

Aortic Valve
Regarding aortic valve replacement, one of the first MPS patients to undergo a mechanical valve replacement received a Braunwald-Cutter prosthesis and was reported to be well 19 years later, having undergone further elective aortic and mitral valve replacements [76,77]. Similar to the mitral valve, a majority of cases reported difficulties with small aortic root and annuli (Appendix A Table A1).
Nick's technique for root enlargement has been used successfully in several MPS patients [76], this is done by extending the inferior aspect of the aortotomy incision through the aortic annulus into the base of the anterior mitral leaflet. A teardrop shaped Dacron or bovine pericardial patch is used to enlarge the incision. The prosthetic aortic valve is then sutured into the enlarged annulus, with a segment attached to the patch.
Konno aorto-ventriculoplasty with root enlargement has also been reported in at least one case [72]. While this procedure offers the possibility of an extended root enlargement with an annular increase of 3-4 mm, it is a complex procedure involving creation of a ventricular septal defect and a double-patch closure [78]. The choice of method for root enlargement should be dictated by surgical experience and overall patient risk [79].

Ross Procedure
The Ross procedure can bring the advantage of not requiring anticoagulation and expected good long-term freedom from reintervention in non-MPS patients (81% at 15 years) [80]. Two documented Ross procedures have been performed on MPS patients, both of which involved post-operative diagnoses of MPS (Table 3). One case reported visual loss due to optic nerve ischemia occurring three days post-op in a 15 year-old MPS II patient [81]. Stable cardiac function was reported at two-year follow-up despite refusal of ERT. In the second case, a MPS IV-B patient re-presented with severe autograft regurgitation and severe pulmonary homograft calcific stenosis twelve years after the procedure [82]; the patient died several days later. It is thought that GAG deposition in both the native autograft and already calcification-prone homograft precipitated the accelerated course of valve failure. Thus, due to risk of rapid valve degeneration, the Ross procedure should not be considered for MPS patients.

Aortic Valve Repair/Ozaki Procedure
When considering the disappointing results with the Ross procedure in these patients, as well as progression of valve disease under ERT, there is scanty support for aortic valve repair, and none for complex reconstruction such as the Ozaki procedure in MPS patients.

Transcatheter Aortic Valve Implantation (TAVI)
In the wider patient population, similarity in mid-term outcomes between TAVI and surgical aortic valve replacement has previously been suggested [83,84]. A successful TAVI (Edwards-Sapien 26 mm) was performed on a MPS I-S patient in 2014 with reported rapid recovery and resolution of symptoms (Table 4) [85]. However, given that TAVI valves are bioprosthetic and constructed with bovine or porcine material, the durability of these valves in MPS patients is currently unknown. There is one case report describing redo surgical aortic valve replacement in a 54 year old female patient with MPS of undescribed type twoand-a-half years after TAVI [86]. This was due to severe AS resulting from structural valve deterioration thought to be associated to GAG deposition. The patient unfortunately died of a "stuck mitral valve" nine days post-redo surgery. However, in patients with very high surgical risk due to advanced MPS or other reasons, TAVI can represent a treatment option in case a surgical replacement is not considered feasible. There would still be significant concerns regarding vascular access routes for these patients if so contemplated.

Mitral Valve
Regarding mitral valve replacement, the majority of cases in literature have reported challenges pertaining to small mitral annuli. Implantation of an appropriate valve size for body size may be difficult and, thus, prosthesis-patient mismatch may occur. As we know, this is associated with poor haemodynamic function, higher risk of cardiac events and increased mortality [87]. Therefore, with a small diameter annulus, it is especially important to select a prosthesis that provides the largest possible effective orifice area to minimise the postoperative pressure gradient across the valve [88]. Should the mitral annulus be too small to fit a mitral prosthesis, an reversely implanted aortic valve prosthesis needs to be considered [76]. However, an enlargement of the mitral annulus might be a superior technique for these cases as to minimise transvalvular gradient as well as improved valve durability in biological prosthesis of larger sizes [56,89]. It needs to be taken into context though of the small body size area and reduced activity levels that many of these patients can achieve though. A summary of published mitral valve replacements is given in Appendix A Table A3.

Mitral Valve Repair
Successful repair of the mitral valve in MPS patients has only been reported twice. Once in a 6-year-old MPS III patient and with very limited follow-up of one year. The repair comprised chordal shortening and a limited annuloplasty with pledget supported mattress sutures in both commissures [90]. Albeit a progression of the local disease must be expected as the clinical presentation already affected the valve tissue, it might be considered as bridging therapy to allow somatic growth facilitating a larger prosthesis at a later timepoint. It should remain a niche indication reserved for situations where no prosthetic valve is anatomically fitting. However, in case ERT, bone marrow transplantation or other treatments in the future are found to significantly slow down disease progression in milder forms of MPS, a mitral valve repair in patients presenting in adulthood might become more commonly considered. Further long-term results of ERT involving histopathological assessment would be required for this.

Conclusions
Surgical valve replacement is often required for the progressive, severe valvular disease that can occur in MPS patients despite advances in systemic treatment. Challenges associated with valve implantation include poor tissue quality and small valve annuli, which may necessitate annular enlargement or a small-diameter mechanical valve both for aortic and mitral position. Additionally, perioperative risks are increased in this cohort of patients, in particular due to difficulties with airway control and ventricular function. Careful multidisciplinary pre-operative planning is essential, personalised solutions for this broad spectrum of clinical presentation are necessary and procedures should be carried out by a clinical team experienced in managing MPS patients. The effect of ERT both on native valve tissue as well as bioprosthetic tissue is yet to be further investigated.