Mitral regurgitation: anatomy is destiny

Summary

Mitral regurgitation (MR) occurs when any of the valve and ventricular mitral apparatus components are disturbed. As MR progresses, left ventricular remodelling occurs, ultimately causing heart failure when the enlarging left ventricle (LV) loses its conical shape and becomes globular. Heart failure and lethal ventricular arrhythmias may develop if the left ventricular end-systolic volume index exceeds 55 ml/m². These adverse changes persist despite satisfactory correction of the annular component of MR. Our goal was to describe this process and summarize evolving interventions that reduce the volume of the left ventricle and rebuild its elliptical shape. This ‘valve/ventricle’ approach addresses the spherical ventricular culprit and offsets the limits of treating MR by correcting only its annular component.

INTRODUCTION

Mitral regurgitation (MR) is an abnormal reversal of blood flow from the left ventricle (LV) to the left atrium. It is caused by disruption in any part of the mitral valve (MV) apparatus, which includes 6 components: left atrial wall, annulus, leaflets, chordae tendineae, papillary muscles and the LV wall. This report focuses on the ‘valve/ventricle’ interface.

The most common causes of MR are primary, due to MV prolapse, rheumatic heart disease, infective endocarditis and annular calcification or secondary, due to dilated cardiomyopathy from ischaemic or non-ischaemic causes. There are no anatomical lesions of the valves, annulus, chordae tendineae or papillary muscles among the secondary causes of MR, underscoring why valve/ventricle interaction is important. Until now, conventional surgery has addressed mitral replacement or repair linked to Carpentier’s innovative valvular techniques [1, 2]. Unfortunately, successfully restoring MV competence does not prevent a compromised long-term prognosis if the preoperative left ventricular end-systolic volume index (LVESVI) is >55 ml/m² [3]. This review emphasizes the importance of understanding and evaluating LV volume and shape in formulating a surgical approach. Adverse late outcomes in a variety of diseases are associated with ventricular dilation, a geometric alteration that can be avoided or treated surgically.

PATHOPHYSIOLOGY

Heart failure disrupts the ability of the ventricle to generate systolic ejection and to augment filling through diastolic suction. It is progressive and may stem from an index event such as damaged myocardium following infarction or ventricular volume overloading from primary valvular dysfunction. Left ventricular remodelling is the early ventricular response after the enlarged and more spherical LV chamber adapts by activating neurohormonal and cytokine systems.

Left ventricular remodelling is progressively maladaptive when the LV becomes more spherical. The increased afterload at end-diastole and early systolic wall stress accentuate specific signal transduction pathways to worsen remodelling [4, 5]. Myocardial fibre orientation is distorted by the abnormal spherical ventricular shape. Systolic torsion falls and there is an independent linear relationship between the left ventricular sphericity and peak systolic twist [6]. Progressive remodelling pulls the papillary muscles apart, altering the chordae tendineae’s natural vertical angulation, thereby tethering the leaflets, which accentuates functional mitral regurgitation (FMR).

A poor clinical prognosis follows progressive LV dilation, independent of its cause. LV size rather than ejection fraction (EF) more accurately predicts outcomes following myocardial infarction (Fig. 1) [7]. The Survival and Ventricular Enlargement Trial (SAVE) altered neurohumoral factors pharmacologically after myocardial infarction. More late heart failure and fatal ventricular arrhythmias developed if LV enlargement increased 3-fold [9]. Mortality was 100% when the left ventricular end-diastolic dimension (LVEDD) was >7.6 cm versus a 45% death rate if it was <7.6 cm in dilated idiopathic cardiomyopathy [10]. Carabello [3], Detaint et al. [11] and Bonow et al. [12] each studied patients with valve disease postoperatively and found higher late death rates from heart failure and ventricular arrhythmias if the preoperative LVESVI was >55–60 ml/m².

Bolognese and Cerisano [13] examined the impact of LV remodelling after successful percutaneous coronary angioplasty for acute myocardial infarction. Despite a patent infarct-related artery, the 80-month prognosis worsened when 15% of the patients developed >20% LV enlargement [13].

LV sphericity is the ratio of the long LV axis to the short LV axis. A sphere ratio is 1.0. Douglas et al. [14] found that 10 of 11 (91%) patients with non-ischaemic cardiomyopathy were dead at 48 months if the sphericity was >0.76, whereas the mortality rate at 72 months was 40% if the sphericity was <0.76 [14].

Myocardial fibre orientation and ventricular shape were used to mathematically predict the left ventricular ejection fraction (LVEF) [15]. A normal EF is 60–70% despite isolated heart muscle preparations exhibiting <20% shortening. The helical fibre orientation that defines the normal LV shape produces systolic twisting (torsion) (Fig. 2). Its natural 60° fibre angulation generates an EF of 60%. Alternately, a spherical LV shape orients the fibres more horizontally, resulting in an EF of only 30%. Ventricular twisting in the healthy heart is due to a clockwise rotation of the inner helical arm and counterclockwise rotation of the outer helical arm. Sphericity deforms this natural shape and worsens twisting by weakening the contractile properties because fibre orientation in dilated hearts becomes transverse rather than maintaining the natural helical orientation.

A unified architectural alteration occurs during ventricular dilation [17] as the normal helical fibre orientation [18] is distorted by flattening the ventricular helix. The resultant fibre angulation change impairs ventricular twisting [16, 19, 20] (Fig. 2). Deformation of the helical and transverse fibres also changes to limit LV mid-wall circumferential strain. The LV sphericity index is the strongest independent predictor of adverse peak apical LV systolic twist in patients with non-ischaemic dilated cardiomyopathy [18].

FUNCTIONAL MITRAL REGURGITATION

FMR develops from restricted leaflet systolic motion back toward the annular plane. The leaflets are tethered to the displaced papillary muscles, which results in their downward systolic tenting, a movement that causes poor systolic coaptation and produces MR. Tethering and impaired coaptation worsen as the spherical LV and mitral annulus widen.

FMR also follows abnormal ventricular conduction (left bundle branch block). Heterogeneous ventricular activation follows prolonged QRS duration, because delayed activation produces myocardial systolic dys-synchrony. The septum billows, creating FMR by tenting the adjacently connected papillary muscles. Cardiac resynchronization produces more homogeneous electrical activation to sometimes reverse remodelling [21] and return torsion [22].

Evaluation of FMR is aided by exercise testing. Echocardiography semiquantitates the extent of the FMR. Without this preferred ambulatory evaluation, inotropes can be delivered intraoperatively to increase blood pressure and afterload.

Current surgical treatment of FMR aims to reduce annular size and the septal-lateral dimension using partial or full rings. Partial banding proponents believe minimal dilation exists between the trigones despite ventricular dilation. This observation differs from de Oliviera’s observation that annular dilation was proportional in ischaemic and non-ischaemic dilated hearts [23]. Rigid rather than flexible bands or rings are preferred in ischaemic cardiomyopathy [24] because they pull the lateral wall medially and fix it in a systolic position. Downsizing the annulus still yields a 5-year mortality rate of 50% because of failure to correct the ventricular component of the mitral apparatus [25].

To achieve leaflet coaptation in FMR, the secondary chords of the anterior mitral leaflet have been cut to relieve tenting caused by posterior papillary muscle displacement, implying no adverse effects on the LV [26, 27]. This approach disrupts the natural papillary muscle leaflet connection that maintains an elliptical ventricular shape and contradicts Lillehei’s 1964 emphasis on preserving the subvalvular apparatus [28]. Experimental studies confirm that this surgical technique consistently increases LV sphericity [29, 30].

In 1 study, preoperative LV dimensions were recorded to determine if restrictive mitral annuloplasty with coronary revascularization reverses remodelling in 100 consecutive patients with cardiomyopathy. The 5-year mortality rate was 49% when left ventricular end-systolic dimension (LVESD) was >51 mm in 28 of these patients, whose LVEDD was >65 mm; these values reflect the cut-off levels for expecting reverse remodelling [31, 32]. Similarly, DeBonis reviewed annuloplasty in dilated hearts and found recurrent FMR and symptom worsening in half of the patients when the LV volume and sphericity index increased [26, 33].

The importance of LV volume and prognosis is shown in the Cardiothoracic Surgical Trials Network (CTSN) trial that randomized 301 patients with moderate ischaemic MR to coronary artery bypass grafting (CABG) or MV repair plus CABG [34]. LVESVI was the primary 1-year end point, and more cases of fatal congestive heart failure and lethal ventricular arrhythmias developed if it was >55–60 ml/m² [3, 11, 12]. Normal LVESVI is 25 ml/m² [35]. LVESVI was reduced by approximately 16% in both study groups, reaching a final index of 46 ml/m². LV remodelling did not reverse this abnormal stretch since postoperative LVESVI still remained almost twice the normal value [34]. This marginal reduction of LVESVI attained statistical significance, but the retained large LV volumes resulted in poor long-term outcomes.

The Cardiothoracic Surgical Trials Network investigators also randomized 251 patients to repair with annuloplasty versus MV replacement with chordal preservation [36]. The LVESVI change was marginal at 6.6 and 6.8 ml/m². Postoperative volumes were still large: LVESVI 55 ± 25 ml/m² (repair) and 61 ± 32 ml/m² (replacement). The 12-month adverse cardiac or cerebrovascular events, functional status and quality of life were similar, but the repair group sustained a higher incidence (33% vs 2%) of recurrent moderate or severe MR [36]. At 2 years, 69% of patients having repair developed moderate to severe MR or died. The mechanism for recurrent MR is increased MV leaflet tethering [37], emphasizing why the surgeon must focus on LV size.

A large multicentre report from Italy comparing MV replacement to repair [38] found similar 8-year survival rates (82% ± 3% for repair and 80% ± 5% for replacement), but the LVEF did not improve in either group. MV repair was a strong predictor of reoperation. The dilated LV continued to damage postoperative function.

An alternate approach for FMR involves supplementing annuloplasty by suturing the centre of the anterior leaflet to the posterior leaflet, thereby creating a double mitral orifice (the Alfieri repair) [39]. Reverse LV remodelling occurred in 52% of 111 patients with ischaemic or non-ischaemic dilated cardiomyopathy, but remodelling remained unchanged in 48% of patients [33]. Reverse LV remodelling improved repair durability and clinical outcome, yet persistent and progressive remodelling developed in more dilated hearts.

The goal of improved leaflet coaptation was addressed by pericardial patching of the anterior MV leaflet in patients with FMR with tented leaflets from LV dilation [40]. Despite achieving a competent annulus, this approach does not address the dilated ventricle or document long-term reversed remodelling.

Conventional use of the term ‘remodelling’ may create confusion because it describes increased LVEDD, spherical LV shape and reduced EF; reversed remodelling describes the return of the LVEDD, shape and EF toward baseline. Bockeria and Averina found that preoperative remodelling correlated with ventricular fibrosis, measured by MRA, but that postoperative fibrosis was unchanged or continued despite reversed remodelling [41]. LVESVI fell from 59 ml/m² to 45 ml/m², but fibrosis was initiated and progressed with enlarged ventricular volumes. Consequently, any mention of reverse remodelling should also include the extent of LVESVI reduction. This will avoid initiating the confusion of relating remodelling to improvement.

ORGANIC MITRAL REGURGITATION

Primary leaflet or chordal disease results in organic MR, leading to the development of several repair methods [1]. Excision of redundant leaflet tissue now includes use of neochords to promote coaptation by resuspending the leaflets [42]. Minimally invasive and robotic approaches are being used more frequently [43, 44]. The Alfieri method of fixing anterior and posterior mitral leaflet coaptation with edge-to-edge central mattress sutures is also done with the transcatheter MitraClip [45]. The 6-month incidence of reverse remodelling depends upon the extent of LVESVI reduction. It was less at 25% in more dilated hearts (63–47 ml/m²) but fell more (41%) with smaller preoperative LV volumes (49–28 ml/m²) [42, 43].

The 2014 American College of Cardiology/American Heart Association valve guidelines point to earlier intervention for chronic MR because of the poor outcomes if LVEF <60%, LVESD >45 mm or LVESVI >60 ml/m² [3]. Currently, asymptomatic patients with preserved LV function (LVEF >60%) are considered for mitral repair (Class IIa) if the LVESD is <40 mm, provided there is a 95% chance that the repair can be done with a mortality rate <1% [46].

Early intervention for MR is the new standard to prevent LV dilation. The Mayo Clinic reported outcomes in >1000 patients undergoing repair or replacement and noted that return to normal LVEF was better if the preoperative LVEF was >65% and the LVESD was <36 mm [47].

THE VENTRICULAR CULPRIT

The aforementioned findings demonstrate that an atrial level repair of MR has consistently failed to prevent progressive LV dysfunction because the ventricle and papillary muscles of the mitral apparatus were not addressed. Annular correction introduces the important first step in modifying components of the mitral apparatus, which form the LV trilogy: the LV base, its mid wall and apex (Fig. 3). The dilemma is how the annulus, leaflets, chordae, papillary muscles and the underlying ventricular wall structures interact to produce leaflet coaptation, because a ‘tug of war’ seems to exist between them (Fig. 4).

Subvalvular techniques have been developed to treat MR in the remodelled LV. Hvass addressed massively dilated LV dimensions (mean LVESVI 70 mm) in Class III and IV heart failure by implanting a slightly downsized ring with an intraventricular papillary muscle sling to bring them together. They are restored to a more normal position with a 4-mm Gore-Tex tube sling. The reduced mitral leaflet tenting resulted in improved LV dimension, LVEF and LV sphericity ratio at the 1-year follow-up examination [48].

Langer supplemented annuloplasty in patients with FMR by reducing leaflet tethering (when the latter was >10 mm) by passing a suture from the head of the papillary muscle into the mitral annular fibrosa. He exteriorized it in the area between the non-coronary and left coronary aortic cusps [49]. Repositioning the displaced posterior papillary muscle toward the fibrosa was accomplished by tying the suture under echocardiographic guidance in the beating heart.

Using papillary muscles for correcting severe MR supplements annuloplasty is done by suturing their bodies to the mitral annulus [50, 51]. Initial experimental followed by clinical studies demonstrate that this technique reverses LV remodelling, decreases the tenting area and improves the leaflet coaptation depth. At 5 years, only 3% of patients developed moderate recurrent MR [52]. A comparison of standard MV replacement with MV replacement supplemented with papillary muscle positioning was done in 50 randomized patients with FMR and LV dysfunction. The papillary muscle heads were sutured to the corresponding annulus. The 3-year follow-up data showed that only the papillary muscle positioning group sustained increased LVEF (34–45%), lower LVESVI and reduced sphericity ratio (0.64–0.49) [53].

Buffalo operated on 116 patients with secondary MR from advanced heart failure with ischaemic and non-ischaemic cardiomyopathy [54], including 11 patients on intra-aortic balloon counterpulsation. All received a bioprosthesis, and he addressed the ventricular component of MR by preserving all chords and by pulling the papillary muscle heads into the annulus for LV reshaping. The hospital mortality rate was 16%. LVEF, sphericity, functional class and stroke volume improved progressively. Patients with and without non-ischaemic cardiomyopathy received the same procedure, so that results in patients with ischaemia may have been impacted by the residual scar.

Calafiore et al. [55] described a ventricular approach for FMR in dilated ischaemic cardiomyopathy by creating a conical ventricular contour without addressing the mitral leaflets. He rebuilt the septum by anchoring an intraventricular patch to normal muscle or scar to restore the LV elliptical shape. In 77 patients, MR fell from grade 2.5–0.6, and the 2.5-year survival rate in patients in New York Heart Association (NYHA) Class I–II was 88% and 85%, respectively.

Menicanti et al. [56] addressed all mitral components in patients with ischaemic dilated cardiomyopathy and FMR. He used the standard surgical ventricular restoration (SVR) operation that included a ventriculotomy, patch placement to restore LV elliptical shape, imbrication of the papillary muscles and suture reduction of the posterior annulus. Operative mortality was 15%. The LVEF did not appreciably change but the LVESVI fell from 98 ± 32 to 63 ± 22 ml/m²; functional class improved and survival was 63% at the 30-month follow-up.

The Coapsys device treats MR by reshaping the spherical dilated ventricle without ventriculotomy. In the beating heart, anterior and posterior epicardial pads are drawn together by a transventricular chord that decreases the mitral annulus and the septal-lateral dimension by decreasing its short axis to reshape the ventricle. The randomized multicentre RESTOR-MV (Randomized Evaluation of a Surgical Treatment for Off-Pump Repair of the Mitral Valve) study of 165 patients with FMR [57] compared Coapsys ventricular reshaping with CABG to conventional mitral repair with CABG. Patients with reshaping had an improved 2-year survival (87% vs 77%) rate and significantly fewer adverse major outcomes. MR fell from grade 2.54 at baseline to 0.52 at 1 year and to 0.35 at 2 years. The Coapsys device is not available, but their findings confirm the importance of changing LV shape and volume by addressing the ventriculomitral complex rather than only employing conventional mitral repair in patients with FMR.

The importance of addressing the ventricular component of MR in ischaemic cardiomyopathy has evolved because of marginal long-term outcomes with annular reduction alone. The RESTORE (Reconstructive Endoventricular Surgery returning Torsion Original Radius Elliptical Shape) study focused on the vessels, valve and ventricle [58] in a 1198-patient, multi-institutional international registry. Inclusion criteria were previous transmural anterior myocardial infarction, significant ventricular dilation (LVESVI ≥60 ml/m²), and an asynergic (non-contractile) regional scar of ≥35% of the LV circumference. During SVR, the anteroseptal, apical and anterolateral LV scarred segments were excluded by an intracardiac patch or by direct closure. CABG and mitral procedures were performed concomitantly (Fig. 5). MV procedures were performed in 22% of patients. They had larger ventricles and more reduced EF. The 30-day mortality rate was higher in the MV repair/replacement group (9% vs 4%), but the 5-year survival rate was similar (69 ± 4% vs 71± 3%) to that of the group without mitral intervention. MV replacement was rare (30 patients, <1%). Findings from this registry confirm the importance of addressing the vessels, valve and ventricular components when treating FMR. Suma extended the ventricular approach to patients with MR with non-ischaemic dilated cardiomyopathy by showing that SVR improved prognosis over repair alone when the LVESVI was >150 ml/m² [60]. Lethal late arrhythmias and fatal heart failure were reduced in sharp contrast to traditional reports of poor surgical outcomes with LVESVI >55 ml/m² [3].

The SAVE (Septal Anterior Ventricular Exclusion) or ‘Pacopexy’ [in recognition of Francisco (Paco) Torrent-Guasp] is an operation that restores the LV elliptical shape with an oblique intracardiac patch from the LV apex to the high septum beneath the aortic valve (Fig. 6). This technique excludes normal septal muscle to rebuild form and expands the seminal contributions by Dor [62] and Jatene [63] that only focus on scar exclusion [64, 65]. The widened papillary muscles are also imbricated as suggested by the RESTORE team [66] and Matsui et al. [67].

The extent of LV volume reduction on long-term outcomes was shown by Isomura in 90 patients with ischaemic dilated cardiomyopathy who underwent SVR. The 8-year survival rate was 82% in 50 patients undergoing mitral repair, and SVR was most effective when a >33% volume reduction achieved an LVESVI <90 ml/m² [68]. Rebuilding the shape of the LV was also a key goal: The 7-year survival rate was 72% with a conical shape versus 62% with a spherical shape.

The role of substantially reducing LV size was further documented by Calafiore’s recent report of 113 patients with ischaemic cardiomyopathy with FMR undergoing SVR and either mitral repair (2/3) or replacement (1/3) [69]. Ventricular volume was reduced >40% and repair versus replacement was determined by mitral leaflet coaptation depth: Repair was used if it was <10 mm and replacement, if it was >10 mm [70, 71]. The 5-year survival rate was 81% if preoperative MR was not severe and only 11% were in NYHA III/IV postoperatively.

Calafiore adhered to the initial inclusion and surgical performance criteria originally envisioned by the STICH (Surgical Treatment for Ischaemic Heart Failure) trial. Unfortunately, the STICH randomized trial was flawed because of its departure from its original design and from the SVR registry guidelines [72, 73]. However, the RESTORE and other similar registries became the benchmarks in establishing the guidelines for SVR by the European Society of Cardiology and the European Association of Cardiothoracic Surgeons [74], which included LV volume, documentation of scar and experienced centres.

The functional impact of restoring LV size and shape was highlighted by Cirillo who reported that rebuilding normal cardiac shape restored natural twisting or torsion in non-ischaemic cardiomyopathy [75]. Patients with FMR were not included, but achievement of an elliptical form was critical because restored torsion persisted for 7.9 years without hospital deaths or late heart failure deaths. These findings echoed those of Isomura and emphasize that correction of shape is vital to address FMR’s ventricular culprit. Volume reduction is also critical because mortality after SVR is increased with a postoperative LVESVI >60 ml/m² [76].

The significance of reshaping the LV and of reducing volume in patients with cardiomyopathy has prompted the development of innovative procedures without open-heart surgery. One device anchors cables between the LV free wall and the septum to reduce volume and reshape the ventricle. LVESVI was reduced 36% and 40% from baseline at 6- and 12-month follow-up examinations in 31 patients, respectively [77].

The LV Parachute device achieves LV reduction non-invasively. An umbrella-shaped nitinol frame covered with an expanded polytetrafluoroethylene occlusive membrane is expanded following percutaneous passage across the aortic valve into the LV apex. Its anchors engage the myocardium to seal off the distal akinetic LV chamber. Currently, 65 US centres are running trials comparing the Parachute with optimal medical therapy in patients with a dyskinetic or aneurysmal LV apex and <35% LVEF [78].

CONCLUSIONS

MR is due to a disturbance of any of the 6 components of the mitral apparatus. LV remodelling progresses as the LV contour becomes more spherical. The scarred or dysfunctional LV is a vital culprit in this process. Addressing MR surgically by reducing the LV volume and restoring the shape together with annular reduction can change the outlook for patients with dilated and ischaemic cardiomyopathy.

Conflict of interest: none declared.

REFERENCES