Transcatheter Mitral Valve Repair and Replacement: Current Evidence for Intervention and the Role of CT in Preprocedural Planning—A Review for Radiologists and Cardiologists Alike

Abstract

The mitral valve is a complex structure with a three-dimensional saddle shape annulus. Mitral regurgitation occurs from leaflet coaptation failure that is either primary (a problem with the leaflets) or secondary (chamber dilatation in the setting of cardiomyopathy). There has been an increase in focus on transcatheter mitral valve interventions, for both mitral repair and replacement. These technologies have rapidly developed to provide treatment for a substantial number of patients with severe symptomatic mitral regurgitation who are at too high of a risk to undergo open heart surgery. CT assessment of the mitral valve has developed with equal rapidity, with regard to preprocedural planning for transcatheter therapies. This review will provide an overview of mitral valve anatomy, an update on the current transcatheter repair and replacement therapies, as well as a focused overview of the role of multislice CT in mitral assessment prior to intervention.

© RSNA, 2020

Summary

This review encompasses mitral anatomy, transcatheter mitral interventions (focused on edge-to-edge repair and transcatheter replacement), and CT assessment of the mitral valve for preprocedural planning.

Essentials

• ■ The mitral valve is a complex structure composed of the annulus, leaflets, chordae, and papillary muscles.

• ■ Mitral regurgitation may either be primary or secondary.

• ■ Evidence continues to grow for both transcatheter mitral repair and replacement for patients with severe symptomatic mitral regurgitation who are too high risk for open heart surgery.

• ■ Multislice CT is an excellent modality to allow assessment of the annular landing zone, neo-left ventricular outflow tract, and degree of mitral annular calcification prior to any intervention.

Introduction

Mitral regurgitation (MR) is the most common valvular pathologic condition in the Western world (1). There has been an increase in focus on transcatheter mitral interventions for symptomatic patients with severe MR who are unfit to undergo major cardiac surgery. This review aims to summarize the current literature with regard to transcatheter repair and replacement and also highlight the approach of CT in preprocedural planning.

Mitral Valve Anatomy

The Annulus

The mitral valve is a complex structure, which consists of the annulus, anterior and posterior leaflets, chordae, and the papillary muscles. While the term annulus denotes “a ring,” the mitral annulus has a dynamic three-dimensional (3D) saddle shape with a maximum height of the saddle in the middle of the anterior part of the annulus, another rise along the posterior leaflet, and the deepest part at the commissures or fibrous trigones (2,3).

The posterior peak is formed by the insertion of the posterior mitral leaflet. The anterior or aortic peak extends to the level of the aortic valve or annulus and is in part defined by the insertion of the noncoronary and left coronary cusp, thereby including the aortomitral continuity (4). A D-shaped mitral annular shape has been proposed using CT assessment and involves exclusion of the aortomitral continuity (4). Figure 1 demonstrates the traditional saddle shape versus the D-shaped annulus using CT assessment.

Figure 1a:

Diagram depicting the (a-c) traditional versus (d-f) D-shaped mitral annulus (by exclusion of the aortomitral continuity). CC = commissure to commissure, P.PE. = posterior perimeter, SL = septal to lateral, TT = trigone to trigone. (Reprinted, with permission, from reference 3.)

Figure 1b:

Diagram depicting the (a-c) traditional versus (d-f) D-shaped mitral annulus (by exclusion of the aortomitral continuity). CC = commissure to commissure, P.PE. = posterior perimeter, SL = septal to lateral, TT = trigone to trigone. (Reprinted, with permission, from reference 3.)

Figure 1c:

Diagram depicting the (a-c) traditional versus (d-f) D-shaped mitral annulus (by exclusion of the aortomitral continuity). CC = commissure to commissure, P.PE. = posterior perimeter, SL = septal to lateral, TT = trigone to trigone. (Reprinted, with permission, from reference 3.)

Figure 1d:

Diagram depicting the (a-c) traditional versus (d-f) D-shaped mitral annulus (by exclusion of the aortomitral continuity). CC = commissure to commissure, P.PE. = posterior perimeter, SL = septal to lateral, TT = trigone to trigone. (Reprinted, with permission, from reference 3.)

Figure 1e:

Diagram depicting the (a-c) traditional versus (d-f) D-shaped mitral annulus (by exclusion of the aortomitral continuity). CC = commissure to commissure, P.PE. = posterior perimeter, SL = septal to lateral, TT = trigone to trigone. (Reprinted, with permission, from reference 3.)

Figure 1f:

Diagram depicting the (a-c) traditional versus (d-f) D-shaped mitral annulus (by exclusion of the aortomitral continuity). CC = commissure to commissure, P.PE. = posterior perimeter, SL = septal to lateral, TT = trigone to trigone. (Reprinted, with permission, from reference 3.)

A wide variation in mitral annular dimensions has been shown in healthy individuals, with men typically having greater dimensions than women. These differences largely disappear when indexed to body surface area. A positive association between mitral annulus size with left ventricular (LV) and left atrial (LA) size has also been demonstrated (5).

The Leaflets

The mitral valve separates the LA and LV and consists of two leaflets, traditionally termed anterior and posterior. The mitral leaflets are uneven in shape and size. The anterior leaflet occupies one-third of the annulus with the posterior leaflet occupying the remaining two-thirds (6). The coaptation line is a semilunar arc conformation, with leaflets converging at the anterolateral and posteromedial commissures. Both the anterior and posterior mitral leaflets have three scallops (A1–A3 for the anterior and P1–P3 for the posterior) (Fig 2). The most lateral scallop is P1 and is adjacent to the anterolateral commissure. P2 is central and varies in size considerably. P3 is medial and is adjacent to the posteromedial commissure (7).

Figure 2:

Mitral valve by CT assessment. Orthogonal planes at A1-P1, A2-P2, and A3-P3 may be generated from the short-axis view. Arrows depict leaflet tenting heights.

Subvalvular Apparatus

The subvalvular apparatus refers to the papillary muscles and the chordae tendineae. The papillary muscles are muscular components of the mitral apparatus to which tendinous cords attach. They arise from the apical and middle thirds of the LV wall (8). Pertaining to the commissure to which they attach, the traditional teaching is that there are two papillary muscles (posteromedial and anterolateral). The chordae are termed primary (attached to the leaflet free edges) and second order (attached to the ventricular surface of the anterior leaflet belly at the junction of the rough and smooth zones). The subvalvular apparatus has a dual role to maintain valvular competence and enhance LV systolic pump function (9,10).

MR: Classification and Pathophysiology

Mitral valve disease can be classified as primary or secondary. Primary MR is caused by a primary valvular defect. Most commonly this is myxomatous or degenerative change (ie, prolapse or flail). Prolapse is the term that indicates that chordal connections of the leaflet to the papillary muscle are intact so that, regardless of the severity of the prolapse, the tip of the leaflet still points toward the apex. Flail occurs after chordal rupture when the leaflet tip points toward the roof of the LA (11). Secondary or “functional” regurgitation refers to incomplete coaptation as a result of distortion of the supporting apparatus, resulting from an ischemic or dilated cardiomyopathy (12). Thus, in secondary regurgitation, the anatomy of the leaflets and tendinous cords is preserved. Inferolateral wall infarction and subsequent adverse remodeling results in posterior leaflet tethering and eccentric MR with a posteriorly directed jet. In a dilated cardiomyopathy, global LV remodeling results in mitral leaflet tenting and a central regurgitant jet. Figure 3 demonstrates varying mitral pathologic conditions.

Figure 3:

A, Demonstration of (i) posterior prolapse, (ii) anterior prolapse with partial flail, and (iii) anterior leaflet flail. B, Demonstration of the mechanism of ischemic mitral relapse (MR) left ventricular remodeling following myocardial infarction (shaded area) resulting in leaflet tethering and impaired coaptation. (Reprinted, with permission, from reference 13.)

Transcatheter Mitral Devices: Repair versus Replacement

While surgical repair or replacement is the reference standard treatment for severe and symptomatic MR, there are a substantial number of patients deemed too high risk to undergo surgery due to multiple comorbidities. Transcatheter techniques are increasingly being offered in such patients.

Edge-to-Edge Repair

MitraClip (Abbott, Abbott Park, Ill) is a device delivered to the mitral valve via a transvenous approach with a transseptal puncture. It was designed to mimic or recapitulate the surgical or edge-to-edge repair, which was almost exclusively used in degenerative (or primary) MR (14). The original surgical technique involved suturing the free edge of the mitral leaflets at the site of regurgitation to create a double-orifice mitral valve. In the percutaneous approach, a clip is applied to the A2-P2 cusps. Concerns regarding the technique included failure to address the leaflet pathology and a risk of mitral stenosis (15). Furthermore, benefit of surgical mitral repair was thought to be a result of the frequent concurrent completion of an annuloplasty.

The Endovascular Valve Edge-to-Edge Repair Study II (EVEREST II) compared MitraClip to conventional surgery, with the MitraClip arm demonstrating superiority in the primary endpoint of survival, surgery, or recurrent MR with surgery at 5 years. The primary safety outcome at 30 days (any major event) was also lower in the MitraClip group, largely due to lower frequency of blood transfusion in the percutaneous group (16,17). Subsequent registries assessing MitraClip in high-risk populations have provided further reassuring data (18,19).

The majority of patients in the EVEREST trials had primary MR. Historically, repair for secondary or functional MR was not thought to be beneficial as restoration of valve competence does not address the underlying problem. Progressive adverse LV remodeling causes worsening ventricular function and adds to strain on the mitral apparatus, with subsequent worsening MR. This in turn drives a cycle of further adverse remodeling and worsening MR (“MR begets MR”) (20). The Multicenter Study of Percutaneous Mitral Valve Repair MitraClip Device in Patients with Severe Secondary Mitral Regurgitation (MITRA-FR) trial supported this rationale. This trial randomized patients with severe MR with LV dysfunction to MitraClip versus optimal medical therapy, with no differences in the composite primary endpoint of unplanned hospitalization or all-cause death at 1 year (21). However, the subsequent Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy for Heart Failure Patients with Functional Mitral Regurgitation (COAPT) study demonstrated a 2-year hospitalization rate of 35.8% versus 67.9% and mortality rate of 29.1% versus 46.1% in those receiving a MitraClip device and optimal medical therapy compared with medical therapy alone, respectively (22).

The opposing outcomes may be explained by differences in baseline medical therapy and entry criteria (Table 1). Additionally, an interesting hypothesis of proportionate versus disproportionate MR has been raised. Compared with COAPT, MITRA-FR enrolled patients with larger ventricles but relatively modest MR. Put another way, in patients enrolled in COAPT, “the effective regurgitant orifice area was approximately 30% higher but their LV volumes were approximately 30% smaller” (23). A subsequent editorial highlighted that in a posthoc subgroup analysis of COAPT (presented at the 2018 Transcatheter Cardiovascular Therapeutics conference), patients with an effective regurgitant orifice area of less than or equal to 30 mm² and LV end-diastolic volume of more than 96 mL/m² demonstrated no change in all-cause mortality or hospitalization due to heart failure 1 year after MitraClip implantation (24). Proportionate patients with MR (ie, those in MITRA-FR) may not benefit from MitraClip as their regurgitant flow more relates to LV dilatation rather than a reparable defect in mitral coaptation (23). On the basis of COAPT, the Food and Drug Administration expanded approval for MitraClip in secondary MR.

Table 1:

Summary of the COAPT and MITRA-FR Trials

System maneuverability and strict anatomic criteria (see following section) are drawbacks of the original MitraClip design. The MitraClip XTR is the updated version of the device and aims to remedy these issues with longer clip arms and a wider grasping width. Additionally, the Edwards PASCAL (Edwards Lifesciences, Irvine, Calif) system is reported to address the limitations of the original MitraClip design by simplifying navigation in the LA, improving MR reduction using a central spacer and allowing for independent leaflet grasping. A first-in-human study using PASCAL demonstrated positive outcomes in terms of MR reduction in a high-risk population (25).

Transcatheter Mitral Valve Replacement

For high-risk surgical patients with severe symptomatic MR, there has been an increase in focus on transcatheter mitral valve replacement (TMVR), particularly for those who have unsuitable anatomy for repair. Challenges of TMVR include the complex mitral anatomy, heterogeneity of pathology, and annular dynamism. In diastole, the annulus increases in size by more than 20% (26). It also changes shape, from circular in diastole to oval in systole. Thus, any prosthesis must be strong enough to withstand such radial forces without fracture, while not being so rigid as to compress surrounding structures (namely, the left ventricular outflow tract [LVOT], coronary sinus, and circumflex artery) (27). Such anatomic and physiologic challenges have led to a variety of device anchoring designs, which are highlighted in Figure 4.

Figure 4:

Varying anchoring mechanisms of percutaneous mitral valves. A, Native mitral valve. B, Anchoring via tabs. C, Anchoring via paddles. D, Anchoring via atrial and ventricular barbs. E, Tethering via the left ventricular apex. AML = anterior mitral leaflet, LA = left atrium, LV = left ventricle, PML = posterior mitral leaflet. (Reprinted, with permission, from reference 3.)

Examples of the many valves under review include the EVOQUE (Edwards Lifesciences) mitral valve replacement system (transseptal delivery with anchoring via the annulus, leaflets, and chords), Intrepid (Medtronic) transcatheter mitral replacement system (transapical delivery with dual nitinol stent design that allows conformation to native annulus), Tendyne (Abbott) (transapical delivery with a self-expanding nitinol stent attached to an apical pad, via a tether) (28), and Tiara (Neovasc, Richmond, British Columbia, Canada) (transapical delivery with self-expanding D-shape nitinol frame and a combination of atrial portion and ventricular anchor design that allows secure fixation) (29). These valve designs are shown in Figure 5.

Figure 5:

A selection of percutaneous mitral valves currently under review. A, Intrepid (reprinted, with permission, from Medtronic). B, Edwards EVOQUE Mitral Valve Replacement System (reprinted, with permission, from Edwards Lifesciences, Irvine, Calif). C, TIARA (Neovasc, Richmond, British Columbia, Canada; reprinted, with permission, from reference 29.) D, Tendyne valve (Abbott Vascular, Santa Clara, Calif; reprinted, with permission, from reference 28). AC = atrial cuff, EP = epicarial pad, T = tether.

From a practical perspective, TMVR requires the delivery of large prostheses via delivery systems (between 34–40 F) into the LV cavity. A transapical approach has therefore typically been adopted. Emergingly, the transfemoral approach with transseptal puncture is being performed. While less invasive, this approach has the disadvantages of reduced intracardiac maneuverability and large sheath sizes (30). In terms of complications of TMVR, the most feared is LVOT obstruction, which is further discussed in a later section.

Perhaps unsurprisingly, and for the aforementioned reasons, the reported rate of screening failure for TMVR is approximately 60%. This is due to a combination of clinical factors (ie, high-to-prohibitive risk population) and anatomic factors (eg, risk of LVOT obstruction, too large or small a native mitral valve, severe mitral annular calcification) (27). The wide variety of devices currently undergoing trials is beyond the scope of this review. Instead we report two of the most recent publications, one using a transapical approach and the other transfemoral.

Sorajja et al reported the largest experience with TMVR to date, which included 100 patients who underwent insertion of a Tendyne mitral prosthesis (31). In short, with the patient under general anesthesia, a left anterolateral thoracotomy was performed with a 5-cm incision to allow transapical access to the ventricle. A total of 97 of 100 patients enrolled underwent prosthesis implantation, with all but one having acute elimination of MR (success rate of 96%). One of the four patients with failed implantation developed a paraprosthetic leak and subsequently underwent repair with an Amplatzer device. The remaining three unsuccessful cases had technical issues around the procedure and occurred early in the series. The major bleeding rate was low (1%), but thrombus was reported (involving the leaflets in most cases) in 6%, with the investigators subsequently adapting their protocol to include 3 months of anticoagulation after implantation. Of note, while survival at 12 months was 72.4% and another 20 patients were hospitalized for heart failure, none of these patients had echocardiographic evidence of MR at more than 1-month follow-up. This morbidity and mortality rate may therefore reflect the frailty and comorbidities of such a patient group. In addition to MR elimination in most, other improved echocardiographic parameters included LV end-diastolic volume reduction and less pulmonary hypertension. A trend in ejection fraction drop was noted; however, this is not unexpected given the increase in afterload following restoration of valve competency. Most striking from the results was the improvement in patient quality of life, with 88% reporting New York Heart Association I or II symptoms at 1 year, in comparison to 34% at baseline (31).

Webb et al reported a first-in-human experience with a transseptal transcatheter heart valve (32). The novel aspect of this technique is the initial deployment of a nitinol “dock” which encircles the chordae below the annulus. Following release of the dock, a balloon-expandable bioprosthesis (Sapien M3 valve; Edwards Lifesciences) is deployed with rapid ventricular pacing, securing the leaflets between the dock and the transcatheter heart valve frame (Fig 6). A total of 10 patients were included in the study, of whom nine achieved the primary endpoint of procedural technical success. One patient underwent successful periocardiocentesis for tamponade during the dock insertion, with the procedure then aborted. At 30 days, 70% of patients achieved the secondary endpoint which was successful device implantation and survival, stroke, and device dysfunction (MR grade > 1, mitral gradient > 6 mm Hg, LVOT gradient > 20 mm Hg). This was a landmark trial as it demonstrated the feasibility of the transfemoral approach for TMVR, which is more desirable for both operator and patient. The technique was proven to be effective across varying pathologies.

Figure 6:

The deployment of the transseptal transcatheter mitral valve replacement system. A, A four-chamber view and B, an en face view from the left ventricular apex. (Reprinted, with permission, from reference 32.)

Current Indications for Transcatheter Mitral Intervention

Indications for percutaneous mitral repair are listed in Table 2. With regard to TMVR, this is currently an emerging area with limited data and thus is not currently in major guidelines.

Table 2:

Indications for Percutaneous Mitral Intervention as per European and U.S. Guidelines

Cardiac CT for Preprocedural Assessment in Severe MR

Echocardiography is currently the primary imaging technique to assess MR (Fig 7) (33). Advantages of echocardiography are widespread availability, lower cost (in comparison to CT and MRI), and portability (patients can be scanned at the bedside). Using 3D transesophageal echocardiography, an en face view of the mitral valve is created to mimic that seen by the surgeon, looking from the LA downward. High-resolution images of the leaflets are generated and, additionally, the site at which the regurgitant jet appears largest can be located. Transthoracic echocardiography is used for MR quantification and to assess for concurrent signs of severe regurgitation (LV or atrial dilatation, pulmonary hypertension, etc). Problems with MR quantification using echocardiography include inconsistency between measures and noncircular, eccentric, and nonholosystolic jets (35). When quantification is proving problematic, cardiac MRI may be used to assess MR severity given its ability to accurately calculate LV volumes and also through the use of phase-contrast imaging. The evidence for MRI assessment of MR severity was recently summarized (36). MRI is additionally advantageous given it does not require intravenous contrast material.

Figure 7:

Echocardiographic criteria for classification of moderate versus severe mitral regurgitation (MR) (33). EROA = effective regurgitant orifice area, LA = left atrium, LV = left ventricle, PISA = proximal isovelocity surface area.

Cardiac CT currently is not validated to assess chamber volumes and also cannot assess hemodynamics, and in these regards is disadvantageous to MRI and echocardiography, respectively. While CT may be used to assess mitral leaflet pathology (discussed in a later section) and has also been used to quantify the effective regurgitant orifice area (37,38), its main strength and current clinical use are to assess the mitral annular landing zone and the prediction of LV outflow obstruction after valve deployment.

Protocol

The mitral apparatus, LVOT, and LV have dynamic changes in configuration and thus CT data acquisition should preferably span the entire cardiac cycle (39). Retrospective or prospective electrocardiographic gating with full cardiac cycle coverage is recommended, without dose modulation. Imaging isolated to the heart is acceptable; however, in the case of transapical access one should use extended z-axis coverage (whole thorax) to allow better assessment of the intercostal space. Other recommended parameters include slice thickness of 0.625 mm and tube voltage of 120 kV. Iterative reconstruction is recommended to reduce noise. The contrast material injection rate is 5.5–6.5 mL/sec and total contrast material volume is 70 mL.

Reconstruction

Data are reconstructed in 10% intervals throughout the cardiac cycle. When looking to assess mitral prolapse, smaller increments (eg, 5%) may also be performed to allow detailed assessment of the leaflets during systole (40). For annular assessment, mid- to late-diastolic image reconstructions are identified given larger annular dimensions at this stage of the cardiac cycle (26,39). The atrial dimensions may be assessed at 85%–95% of the cardiac cycle, depending on the presence of atrial contraction. The aim is to evaluate the annulus on the last reconstruction prior to atrial contraction (4).

CT Utility in Assessment of Primary versus Secondary MR

CT is useful as an adjunct to provide information regarding mitral leaflet pathology. Compared with echocardiography, CT was found to be robust in identifying prolapse, with three- and two-chamber reformations yielding the highest accuracy (sensitivity of 96%; specificity of 93%) (41). Furthermore, CT can confirm the absence of anatomic exclusion criteria, which are mitral valve area of less than 4 cm², flail gap greater than or equal to 10 mm, flail width greater than or equal to 15 mm, and coaptation length of less than 2 mm (14).

Mitral annular disjunction is a remodeling process associated with mitral prolapse. It is a detachment of the annular “roots” from the ventricular myocardium, which occurs in the area under the posterior leaflet (P1 and P2 scallops), but not the anterior due to the adjacent fibrous trigones (42). The result is a wide separation between the atrium–mitral valve junction and the LV junction (43). Mitral annular disjunction is diagnosed in systole when the separation is greatest (Fig 8). The magnitude of disjunction has been shown to correlate with the number of segments with prolapse/flail (44). Mitral annular disjunction causes “decoupling” of the annulus from LV deformation during systole, with a resultant paradoxical systolic expansion and “unsaddling” (43). Its presence therefore has implications for mitral repair techniques, adding emphasis to the need for concurrent annuloplasty in mitral prolapse. Cardiac CT is an excellent modality for the assessment of mitral annular disjunction, as full cardiac cycle data acquisition allows appreciation of both the degree of disjunction and also annular dynamics.

Figure 8:

Mitral annular disjunction (arrow) at CT imaging with image reconstruction to create left ventricular, A, two-chamber and, B, four-chamber imaging. Both mitral leaflets are thickened, consistent with myxomatous mitral disease.

With secondary or functional MR, CT can provide information regarding the remodeling process and mitral valve geometry, including leaflet tenting heights (coaptation depth) and leaflet angles. In patients with severe MR, the maximum tenting height and tethering of the posterior mitral leaflet has been found to be located at the central and posteromedial levels. Knowledge of this data has importance with regard to device anchoring. It also allows accurate procedural planning and may result in shortening of fluoroscopy and procedure times (40).

An additional feature of atrioventricular remodeling due to functional MR that is well assessed using CT is the posterior myocardial “shelf.” Naoum et al demonstrated that while functional MR is associated with a significantly increased LV size, the annular size is in fact more closely associated with the LA size (5). It appears that the linear relationship between the LV myocardium and LA is disturbed during progressive LV dilatation, which results in the development of the myocardial “shelf” (Fig 9). The shelf and, in particular, its presence in both systole and diastole, is of importance for capture and positioning of certain TMVR devices (45). CT can also provide further information regarding LV dimensions, which is pertinent given that a diameter of greater than 70 mm is often used as an exclusion criterion for TMVR.

Figure 9:

Echocardiographic and CT imaging of the myocardial shelf. A, Parasternal long-axis transthoracic echocardiogram demonstrates thinning/scar of the basal inferolateral wall (arrow). B, Same image with color Doppler imaging demonstrates a posteriorly directed mitral regurgitant jet due to posterior leaflet tethering. C, CT images with left ventricular short-axis and, D, three-chamber views demonstrate thinning and fatty metaplasia of the basal inferolateral wall, consistent with scar (arrows).

Both mitral annular size and dynamism vary between different mitral pathologies. Mitral annular dimensions have been shown to be larger in patients with MR versus control participants, with variation in geometric changes. In those with moderate to severe MR, annular remodeling was more anteroposterior (measured using the intercommissural measurement) rather than lateral. (5). In functional MR, the annulus has reduced dynamism (due to it being dilated and stiff) whereas in primary MR, dynamism is increased (46). Full cardiac cycle CT imaging allows accurate measurement of such changes, thus providing useful information with regard to planning for TMVR procedures.

Annular and Landing Zone Geometry

Given the saddle complex nature of the mitral annulus and the variation in imaging planes for differing individuals, 3D imaging is superior to two dimensional for assessment of the annulus. Accurate assessment of annular geometry is imperative to provide accurate information regarding the landing zone (ie, where the prosthesis will sit).

The traditional approach to mitral annular geometric assessment involved manually plotting 16 segmentation points for cubic spline interpolation along the insertion of the posterior mitral leaflet and along the contour of the anterior peak comprising the fibrous intervalvular continuity by stepwise rotation of a long-axis view aligned to the LV long axis. A simplified D-shaped model has been proposed for mitral annular sizing (47). In this approach, the anterior horn is excluded by drawing a virtual line between the medial and lateral trigones. The insertion of the posterior leaflet defined the posterior margin (Fig 1).

The D-shaped approach better reflects the planar landing zone of transcatheter mitral valve implantation devices, and, additionally, it eliminates the difficulties associated with defining and segmenting the anterior horn (5). Furthermore, this approach has a lower risk of LVOT obstruction from a device, as the projected area (particularly the septal to lateral distance) is smaller than that measured using the strict anatomic term for the annulus. The trigone-to-trigone distance remains unchanged (4).

Predicting LVOT Obstruction (Simulating Transcatheter Heart Valve)

TMVR elongates the outflow tract into the LV, which is referred to as the neo-LVOT (48). The neo-LVOT is defined as the centerline through the middle of the residual space between the septum and simulated implanted transcatheter device (6). LVOT obstruction is a catastrophic complication of TMVR. Mechanisms are both fixed from the prosthesis itself obstructing the LVOT and dynamic, from systolic anterior motion of the anterior mitral leaflet due to displacement by the prosthesis. Predictors of LVOT obstruction are outlined in Figure 10 (3). Additionally, reduced area of the neo-LVOT after device implantation is a marker of obstruction due to anterior leaflet displacement into the LVOT by the covered device (45). Derived from studies in hypertrophic cardiomyopathy, a valve area of less than 2 cm² is associated with significant outflow obstruction (pressure gradient > 50 mm Hg) (49,50). Furthermore, a recent study of patients planned for TMVR due to failed mitral bioprosthetic valves (valve-in-valve), annuloplasty rings (valve-in-ring), and mitral valve calcification (valve-in-mitral annular calcification) demonstrated that a neo-LVOT area equal to or less than 1.7 cm² was a predictor of LVOT obstruction with a sensitivity of 96.2% and specificity of 92.3% (51).

Figure 10:

Predictors of left ventricular outflow tract (LVOT) obstruction following transcatheter mitral valve implantation. MA = mitral annulus, LV = left ventricle, SAM = systolic anterior motion of the mitral valve. (Adapted, with permission, from reference 48.)

Advanced software can be used to simulate device implantation by embedding a cylindrical or device-specific contour into the CT data set, with subsequent segmentation and planimetric assessment of the neo-LVOT cross-sectional area. An ongoing challenge is when in the cardiac cycle to assess the LVOT area. Unsurprisingly, a neo-LVOT will measure smaller at end systole; however, as the LV volume has been ejected, it may be of more use to measure the neo-LVOT at either early or midsystole (3).

Additional Metrics for Assessment

Mitral Annular Calcification

Mitral annular calcification is a fibrous, degenerative calcification of the mitral valve support ring (52,53). Risk factors for mitral annular calcification are similar to those for atherosclerosis, and mitral annular calcification is a strong and independent predictor of cardiovascular morbidity and mortality (54). The pathophysiology of cardiac calcification may relate to metabolic processes, given coronary calcification is commonly seen in patients with end-stage renal failure (55,56).

Mitral annular calcification is of importance when considering percutaneous mitral interventions. It was found to be an independent predictor of an elevated mean diastolic gradient after MitraClip procedure (57) and in its most severe form, is a contraindication to TMVR (as is heavy leaflet calcification). CT is advantageous over echocardiography for providing information regarding the anatomic extent and severity of mitral annular calcification. A particular benefit of CT is the ability to differentiate mitral annular calcification from caseous annular calcification, a rare variant, in which bulky space-occupying lesions occur along the posterior annulus (58). CT imaging allows tissue characterization to differentiate caseating from the more typical mitral annular calcification (Fig 11).

Figure 11:

Cardiac CT images of caseating mitral annular calcification. A, Noncontrast image demonstrates calcification of the anterior annulus and, B, contrast-enhanced image demonstrates the central liquefaction of the mitral annular calcification, which is clearly seen to extend into the left ventricular outflow tract and onto the anterior mitral leaflet.

The presence of mitral annular calcification may be problematic for annular segmentation. A clear, highly reproducible technique has clear advantages and may be achieved by defining the annulus as a “harmonious, smoothly marginated structure.” Excluding protuberant calcification during segmentation will avoid individualized or nonstandardized segmentation to such variant anatomy (6).

Fluoroscopic Angle Prediction

Transcatheter valve implantation involves “the implantation of a quasicylindrical device inside a highly variable anatomical structure.” Minimization of parallax is achieved through the identification of a projection angle that has the source-to-detector direction orthogonal to the axis of symmetry of the anatomic feature of interest (59).

Coplanar fluoroscopic projection angles appear important to ensure coaxial device deployment, thus achieving optimal device positioning and minimizing the risk of complications. CT can provide such projection angulations, based on the mitral annular plane, to provide the optimal projection curve for coplanar projections. C-arm projections orthogonal to the mitral annulus and additionally, aligned with specific anatomic structures, allow the interventionalist to visualize important device components during deployment (60).

Imaging the Coronary Sinus and Circumflex Artery

Coronary sinus imaging using multidetector CT is of importance in preprocedural planning for both percutaneous valve repair and replacement. Anatomic studies exploring the spatial relationship between the coronary sinus and great cardiac vein raised concern regarding the potential clinical efficacy of the annuloplasty technique and also the risk of complications (61). A complication of particular concern is coronary artery compression, which occurred in 17% of patients in the CARILLON Mitral Annuloplasty Device European Union Study (AMADEUS) trial (62). Coronary vessels (typically the circumflex artery) frequently travel between the coronary sinus and great coronary vein and the mitral annulus. Such anatomy has been shown in an artery of clinical relevance in 66%–68% of patients (63,64). Tops et al demonstrated that the minimal distance between the coronary sinus and circumflex artery is approximately 1 mm, and also that in 90% of cases of functional MR, the coronary sinus is superior to the annulus (ie, adjacent to the LA) (64). Clearly these patients would not benefit from a percutaneous annuloplasty procedure. The aforementioned anatomic studies highlight the advantages of multislice cardiac CT to create 3D renderings of the heart, to allow the operator to gauge patient suitability in preprocedural planning.

With regard to TMVR, multidetector CT can be used for segmentation to simulate a guidewire in the coronary sinus, to allow assessment of its relation to the mitral annular plane. In addition to 3D reconstruction to demonstrate the anatomy of the coronary sinus and great cardiac vein and their relation to surrounding structures, appreciation of how close or distant the coronary sinus wire is from the true annular plane provides additional information to guide device deployment (60).

Conclusion

Evidence continues to grow for transcatheter mitral intervention in selected patients. Multislice cardiac CT offers a 3D detailed assessment of the mitral anatomy prior to transcatheter repair or replacement. For repair, CT can provide anatomic information with regard to suitability for a MitraClip device. For replacement, landing zone geometry can be assessed, as well as features that increase procedural risk, in particular the potential for LVOT obstruction and the degree of mitral annular calcification.