of the thesis

(1)

Cover Page

The handle http://hdl.handle.net/1887/136091 holds various files of this Leiden University dissertation.

Author: Rosendael, P.J. van

Title: Cardiac computed tomography for valvular heart disease and non-coronary percutaneous interventions

Issue date: 2020-09-03

(2)

545024-L-bw-Rosendael 545024-L-bw-Rosendael 545024-L-bw-Rosendael 545024-L-bw-Rosendael Processed on: 22-7-2020 Processed on: 22-7-2020 Processed on: 22-7-2020

Processed on: 22-7-2020 PDF page: 9PDF page: 9PDF page: 9PDF page: 9

1 General Introduction and outline

of the thesis

(3)

10

Valvular heart disease is a major health problem associated with ageing of the population. Projections based on the British OxValve Population Cohort study indicate that the number of individuals aged 65 years or more with significant valvular heart disease will double by 2046 (from 1.5 million to 3 million in the United Kingdom).(1) Accurate diagnosis of the presence and severity of valvular heart disease is crucial to provide timely therapy. The development of transcatheter valve therapies during the last decade has revolutionized the treatment of patients with symptomatic severe valvular heart disease. Patients with symptomatic severe aortic stenosis who were deemed inoperable have now the possibility to improve their outcome if treated with transcatheter aortic valve implantation (TAVI). In addition, patients with symptomatic severe aortic stenosis at intermediate or high risk for surgical aortic valve replacement can now be treated with transcatheter aortic valve replacement which provides less invasive treatment, faster recovery and better cost-effectiveness as compared with surgical aortic valve replacement (when the transcatheter procedure is performed through a transfemoral approach).(2) Novel transcatheter therapies are being developed for mitral and tricuspid valve disease. Computed tomography (CT) is a 3-dimensional (3D), isotropic, high-spatial resolution imaging technique that permits accurate assessment of the cardiac anatomy and by integrating the anatomical data with echocardiographic data (functional imaging), it allows quantification of valvular heart disease. This imaging technique has been key to develop and refine transcatheter valve therapies and to improve the results of the treatment by accurately selecting the patients and the device size. This thesis provides a comprehensive overview of the role of CT in the diagnosis and treatment of patients with valvular heart disease.

VALVULAR HEART DISEASE: SCOPE OF THE PROBLEM

Valvular heart disease is a global clinical problem with an expanding prevalence that increases in parallel with the extending life expectancy of the general population.(3, 4) It has been estimated that worldwide more than 100 million patients are affected by valvular heart disease.(5) An echocardiographic community screening of 2500 individuals of 65 years or older, performed in 2016 in the United Kingdom, revealed a prevalence of significant (moderate or severe) valve disease of 3.3% at 65-74 years which increased to 11.9% in the individuals older than 75 years.(1) In addition, mild degrees of valvular heart disease, of

(4)

11 which the true clinical implications remain currently to be further clarified in many specific patient groups, was observed in 44.4% of the patients.(1)

If left untreated, severe symptomatic valvular heart disease is clearly associated with an impaired prognosis and with a significantly decrease in quality of life.(3, 6) As significant valvular heart disease is associated with an older age, it is frequently present in fragile patients who also suffer from other additional cardiac and non-cardiac diseases, resulting in an increased risk for conventional surgical treatment. The Euro Heart Survey on valvular heart disease, a registry that included 5001 patients who were enrolled in 25 European countries between April and July 2001, revealed that among 261 patients with symptomatic severe aortic stenosis (mean age 80.3±4.2 years), 33% of the patients were deemed not operable.(7) The main determinants of non-referral to surgery were a reduced left ventricular systolic function, (odds ratio of 2.27 for a left ventricular ejection fraction of 30-50% (95% confidence interval (CI) of 1.32-3.97) and older age (odds ratio of 1.84 (95% CI 1.18-2.89) for patients aged 80-85 years as compared to those aged 75-79 years).(7) After 1 year of follow-up, the overall survival was lower in the patients who were not operated as compared to those who were operated (84±4.8% vs. 90.4±2.6%, p=0.057).(7) For severe mitral valve regurgitation, which was observed in 396 patients (mean age 66±13 years), surgical intervention was not performed in 193 (49%) patients.

The principal reasons to consider the patient non-operable were a decreased left ventricular systolic function, a non-ischaemic instead of ischemic aetiology of mitral regurgitation, an older age and the clinical judgment for an increased surgical risk.(8) Also for these patients, the overall one-year survival was lower in those who were deemed non-operable as compared to those who were referred for surgery (89.5±2.3% vs. 96.0±1.4%, p=0.02).(8) Medical therapy temporarily improves the most severe symptoms but does not improve prognosis.(9) The increasing population of patients with severe valvular heart disease at an age or in a condition that does not allow them to safely undergo conventional heart surgery provided the grounds for the development of less invasive valvular interventions by transcatheter repair or replacement.

DIAGNOSIS OF VALVULAR HEART DYSFUNCTION

In current routine practice, valvular heart disease is initially diagnosed and monitored by using two-dimensional (2D) color Doppler echocardiography. This technique enables a global assessment of the severity of valvular heart disease.

1

(5)

12

Three-dimensional imaging techniques have further refined the diagnosis of valvular heart disease.(10)

Aortic stenosis. Current guidelines define severe aortic stenosis based on the following echocardiographic criteria: an indexed aortic valve area ≤0.6cm²/ m², a peak aortic valve velocity ≥4m/s or a mean aortic valve gradient of ≥40 mmHg.(6, 11) Ideally, there is concordance in all these parameters suggesting severe aortic stenosis. However, in 30% of patients these echocardiographic parameters do not always coincide.(12) In patients with a reduced left ventricular systolic function (ejection fraction <50%), the aortic valve area may be

≤0.6cm²/m² with a mean gradient <40 mmHg, the group of so-called classical low gradient severe aortic stenosis. (13) The reduced left ventricular systolic function leads to a reduced stroke volume and a low transvalvular gradient. In this group of patients, the key question is to define whether the narrow aortic valve area is caused primarily by degeneration (calcification) of the valve (true severe aortic stenosis) or if it is secondary to a reduced left ventricular function (pseudosevere aortic stenosis). According to current recommendations, low dose dobutamine stress echocardiography should be performed to normalize the stroke volume (flow reserve). In patients with flow reserve whose aortic valve area remains ≤0.6cm²/m², the aortic stenosis is truly severe. However, if the aortic valve area increases >0.6cm²/m² after normalization of the stroke volume, the aortic stenosis is pseudosevere. In patients in whom the stroke volume is not normalized (absence of flow reserve), additional imaging techniques can aid in the diagnosis of the aortic stenosis severity. Quantification of the aortic valve calcium burden with CT has shown good accuracy to identify patients with true severe aortic stenosis: an aortic valve calcium burden ≥2000 in men and ≥1200 in women indicate likely severe aortic stenosis.(13) Patients with an aortic valve area

≤0.6cm²/m², a good left ventricular systolic function (ejection fraction≥50%) and a mean aortic transvalvular gradient <40 mmHg, are classified as paradoxical low gradient severe aortic stenosis. In these patients, the left ventricle is frequently very hypertrophic, with a small ventricular cavity which results in a low stroke volume. Assessment of the aortic valve calcium load can be a good alternative to identify the patients with true severe aortic stenosis in this particular subgroup.

Another factor that may explain the discrepancy between aortic valve area and mean transvalvular gradient is the assumption of a circular shape of the aortic annulus and left ventricular outflow tract by 2D echocardiography while in fact, these structures have a more elliptical shape as has been shown by 3D imaging techniques. Consequently, the left ventricular outflow cross sectional area derived from 2D echocardiography is frequently underestimated as

(6)

13 compared to the planimetered area measured with 3D modalities. Compared to 3D echocardiography, it has been demonstrated that the aortic valve annulus area is significantly underestimated (3.89±0.74cm² vs. 4.06±0.79 cm², p<0.001) by 2D echocardiography, resulting in a significant reclassification of patients with severe aortic stenosis into moderate stenosis if 3D echocardiography is used.(14) Similarly, the introduction of the planimetered left ventricular outflow tract area as measured with CT into the continuity equation for the calculation of the aortic valve area has resulted in a significant proportion of patients being reclassified from severe aortic stenosis into moderate aortic stenosis.(15)

Mitral regurgitation. Diagnosing the severity of mitral regurgitation relies on qualitative, semiquantitative and quantitative echocardiographic parameters.(10) The qualitative parameters include: the type of lesion causing the regurgitation (flail or papillary muscle rupture), morphology of the color flow regurgitant jet (very large central jet or eccentric jet adhering to the left atrial wall and reaching the pulmonary veins), dense and triangular continuous wave Doppler spectral signal and large flow convergence zone. The semiquantitative parameters that define severe mitral regurgitation are a vena contracta width of ≥7 mm, systolic pulmonary vein flow reversal and a dominant E-wave (≥1.5 m/s). Finally, the quantitative parameters indicating severe mitral regurgitation include an effective regurgitant orifice area of ≥0.4cm² or ≥0.2cm² and a regurgitant volume of ≥60 ml/beat or ≥30 ml/beat in primary or secondary mitral regurgitation, respectively. These values are based on the 2D proximal isovelocity surface area method. A regurgitant fraction ≥30% (a parameter that is usually obtained with cardiovascular magnetic resonance) is also a quantitative parameter indicating severe mitral regurgitation.

One of the limitations of assessment of the mitral regurgitation grade with the 2D proximal isovelocity surface area method is that the formula in this method assumes the regurgitant orifice area as circular, while especially in functional mitral regurgitation, the jet is frequently rather oval as has been demonstrated by 3D imaging. With 3D techniques, the regurgitant orifice area of the jet can be analysed on an “en face” plane that directly visualizes the anatomic regurgitant orifice area. Thereby, the assumption of the regurgitant jet as perfectly circular is avoided. Consequently, 2D echocardiography tends to underestimate the regurgitant volume of mitral regurgitation with the proximal isovelocity area method as compared to 3D echocardiography in which the true effective regurgitant orifice area is incorporated in the formula.

In a head to head comparison study between 2D and 3D echocardiography in 221 patients with mitral regurgitation (53% organic, 47% functional) and

1

(7)

14

cardiac magnetic resonance as reference standard, the regurgitant volume of the jet was significantly underestimated with 2D compared to 3D (52.4±19.6mL vs. 59.5±25.6 mL, p=0.005) echocardiography.(16) Especially in patients with severe mitral regurgitation (odds ratio 6.96, p<0.001), eccentric jets (odds ratio 3.82, p=0.017) and an asymmetrical regurgitant orifice area (odds ratio 11.48, p<0.001), significant differences (>15 mL) in regurgitant volume were observed.

(16) Furthermore, the 3D echocardiography derived regurgitant volume correlated overall more closely (r = 0.97 vs. 0.84) to the regurgitant volume as assessed by cardiac magnetic resonance. Cardiac magnetic resonance can be assumed as the most accurate imaging technique for the evaluation of cardiac volumes as it provides volumetric, in cine-loop images of the cardiac chambers and heart valves. However, in comparison with echocardiography, this modality is unpractical due to long acquisition times, associated costs, and the need for adequately trained staff. Nonetheless, combining different techniques as echocardiography and CT for the assessment of a fusion parameter analysing mitral regurgitation severity may also be appealing, as by the integration of both techniques, the individual limitations associated with each technique may be overcome. A dynamic, full beat acquired CT scan enables a detailed anatomical assessment of the true size and shape of the regurgitant orifice area of the leaking mitral valve leaflets in a spatial resolution that surpasses that of 3D echocardiography. In contrast, in CT, and also in dynamic CT data, analysis of the true haemodynamic flow characteristics of the regurgitation is not possible and this is where the Doppler flow data may fulfil an incremental contribute role.

Integrating the anatomical regurgitant orifice area from CT with Doppler data may provide a fusion regurgitant volume. Furthermore, the role of CT in the evaluation of patients with significant mitral regurgitation is further enhanced as this technique allows an accurate analysis of the coronary arteries, important for both the analysis of a potential ischemic cause of the mitral regurgitation as for the work-up evaluation in patients referred for mitral valve surgery. In addition, functional mitral regurgitation is frequently concomitant in patients with severe aortic stenosis who undergo CT for the evaluation of TAVI. These additional reasons for the performance of CT make the concept of an integrated echo-CT derived mitral regurgitation quantification parameter worth to explore.

Due to the current lack of available prognostic data of 3D parameters quantifying mitral regurgitation grade, current 2017 valvular guidelines still hold the cut-off values for the definition of severe mitral regurgitation that were based on 2D echocardiography.(6, 11) Whether these values are also applicable for the regurgitant orifice area and regurgitant volume as assessed by 3D techniques

(8)

15 and whether 3D techniques can provide advanced prognostic data warrant new outcome studies.

Tricuspid regurgitation. As for the other valvular diseases, tricuspid regurgitation is initially evaluated with 2D echocardiography. First, it is evaluated whether the tricuspid regurgitation results from organic tricuspid valve disease (e.g. infectious, rheumatic, carcinoid or myxomatous) as may be seen in younger patients, or that it has a functional aetiology secondary to right ventricular remodeling due to pressure or volume overload.(11)

Similarly to the mitral valve, it is recommended to grade the severity of tricuspid regurgitation using a multi-parametric approach by analysis of qualitative, semiquantitative and quantitative echocardiographic parameters.

With qualitative analysis of the tricuspid regurgitation grade, the structural integrity of the tricuspid valve and the degree of systolic leaflet coaptation (flail, tethering) is evaluated, and the morphology of the color flow regurgitant jet (small or large and central or eccentric) and its signal (faint/dense and parabolic or triangular with an early peak) are analysed. Semiquantitative parameters that are recommended to define severe tricuspid regurgitation grade are a vena contracta width of the regurgitant jet of >7mm, a proximal isovelocity area radius of the jet of >9mm, the presence of systolic hepatic vein flow reversal, and a dominant E-wave larger than ≥1.5 m/s. Quantitative parameters that define severe tricuspid regurgitation are an effective regurgitant orifice area of ≥0.4cm² and a regurgitant volume of ≥45 mL/beat as analysed by the 2D proximal isovelocity surface area method.(11)

In addition to the echocardiography regurgitation grade, also the measurement of the tricuspid annulus has important therapeutic implications as initially shown by the surgical study of Dreyfus et al.(17) In that study including 311 patients undergoing mitral valve repair between 1989 and 2001, concomitant tricuspid valve repair was performed irrespectively of the echocardiographic tricuspid regurgitation grade, in the patients whose tricuspid annulus diameter was larger than 70mm as measured intra-operatively and this strategy was associated with improved functional outcomes.(17) The beneficial impact of repairing the tricuspid valve concomitantly during left-sided valve surgery according to the presence of tricuspid annulus dilatation instead of the presence of significant tricuspid regurgitation was also observed in a study comparing two historical cohort of patients, 102 operated during the year 2004 and 80 operated during the year 2002 (18). Patients operated during the year 2004 received a tricuspid annuloplasty based on a tricuspid annulus dimension >40 mm (regardless of the grade of tricuspid regurgitation) whereas

1

(9)

16

patients operated during the year 2002 received the tricuspid annuloplasty if the tricuspid regurgitation was at least moderate. Patients who received the tricuspid annuloplasty in 2004 had smaller dimensions of the right ventricle at follow-up as compared to their counterparts suggesting that preventive tricuspid annuloplasty based on the tricuspid annulus dimensions (even if there is not yet significant tricuspid regurgitation) halts right ventricular remodeling and dysfunction.(18)

Currently, the importance of tricuspid annulus dilatation is acknowledged and it is recommended to repair the tricuspid valve concomitantly during left- sided valve surgery also if the tricuspid annulus is larger than 40mm on 2D echocardiography.(11) However, more refined insights and reference values may be obtained with 3D imaging. The tricuspid valve annulus has a complex, 3D saddle shaped configuration which is challenging to comprehensively evaluate with 2D echocardiography. Also for this valve, 2D tends to underestimate annulus dimensions and regurgitation grades compared to 3D echocardiography. In a study including 90 patients with chronic tricuspid regurgitation who underwent 2D and 3D echocardiographic assessment, the 2D effective regurgitant orifice area and regurgitant volume were underestimated compared to these values as assessed by 3D echocardiography: 0.30±0.22 cm² vs. 0.38±0.30 cm² and 30.2±23.2 vs. 38.7±29.2 mL (no comparative p-value was provided).(19) CT may also be a valuable imaging tool for the evaluation of patients with significant tricuspid regurgitation who are potential candidates for surgical or transcatheter intervention. With the appropriate settings of contrast media injection and timing, CT provides a comprehensive evaluation of the right side of the heart, the tricuspid valve annulus and the coronary arteries. CT may be the imaging tool of choice to define suitability for upcoming transcatheter tricuspid valve therapies.

ROLE OF MULTIMODALITY IMAGING IN THE SELECTION OF PATIENTS FOR AND PLANNING OF TRANSCATHETER THERAPIES

In contrast to open heart surgery where the operator has a full direct vision on the diseased valve, transcatheter procedures fully rely on periprocedural imaging techniques. Multimodality imaging analyses using echocardiography, CT and fluoroscopy have proven to be essential for patient selection, procedural planning, accurate prosthesis sizing, therapy guidance and follow-up.(20) The insights and research data derived from these imaging techniques have

(10)

17 significantly contributed to the rapid technological developments in prosthesis design and delivery systems enabling procedural simplification and the striking reduction of complications.(20)

Transcatheter aortic valve implantation. For the TAVI procedure, CT is the imaging modality of first choice for prosthesis sizing and procedural planning. The high spatial resolution images that this technique produces, display the aortic valve annulus and aortic root in great detail which enables optimal prosthesis sizing.(21) Accurate analysis of the aortic valve annulus dimensions for preprocedural sizing of the prosthesis is crucial. An undersized prosthesis could result in paravalvular regurgitation or device embolization. In contrast, an oversized prosthesis could lead to rupture of the aortic annulus. CT measurements of the aortic annulus are very reproducible with low inter- and intraobserver variability.(21)

In the selection of patients for TAVI, several anatomical aspects should be carefully considered and may be comprehensively evaluated with CT(22): First (1), the overall degree of aortic valve calcification can be assessed on a non- contrast calcium scan using the Agatston calcium score and the calcium volume.

Thereafter (2), the exact location and distribution of calcification deposits on the aortic cusps can be analysed on the contrast-enhanced scans. An unequally large bulk of calcium might result in gaps between the prosthesis frame and the host annulus that lead to paravalvular regurgitation. Furthermore (3), the exact anatomy of the aortic valve (tricuspid versus bicuspid) can better assessed with CT than with 2D transthoracic echocardiography. Initially, a bicuspid valve anatomy was a relative contra-indication for TAVI with a significant lower success rate compared to patients with a tricuspid valve annulus. In a relative large, worldwide multicenter registry of patients with severe bicuspid aortic valve stenosis who were treated with TAVI (n=561), the procedural success rate was lower than in patients with a tricuspid aortic valve, (85.3% vs. 91.4%; p=0.002).(23) However, these procedural differences disappeared with the new generation prostheses.(23)

Next (4), the aortic valve annulus dimensions and eccentricity are assessed in the double oblique reconstruction of the aortic valve annulus in order to accurately determine the most optimal size of the TAVI prosthesis. In addition to the aortic valve annulus (5), the dimensions of the sinus of Valsalva, sinotubular junction and of the ascending aorta should be evaluated. In patients with severe aortic stenosis and at a low and intermediate risk for surgery who also have significant dilation of the aortic root and ascending aorta, surgical aortic valve replacement together with aortic root and/or ascending aorta replacement may

1

(11)

18

be preferable over TAVI. The following step (6) is the analysis of the distance between the aortic annulus and the ostia of the coronary arteries. In patients with a low take-off of the coronary arteries, the risk of obstruction of the coronary ostia with the TAVI is increased and this needs to be prevented. Subsequently, the CT can be used to determine the optimal access strategy (transfemoral or transapical) by analysis of the thoracic and abdominal aorta and the femoral arteries for diameters of the lumen, tortuosity and degree of atherosclerosis. For the current SAPIEN 3 (Edwards Lifesciences, Irvine, California, USA) prosthesis, the lower cut-off value for transfemoral artery lumen diameter is 5.5 mm for the 23 and 26 mm size prostheses and 6.0 mm for the 29 mm size prosthesis.

Furthermore (7), the presence of significant coronary artery disease requiring a concomitant or staged percutaneous coronary intervention should be evaluated in advance and for some patients, evaluation of the coronary arteries with CT may avoid the necessity of a diagnostic angiogram. In addition (8), concomitant cardiac pathology as severe left ventricular dilatation, the presence of a thrombus in the left ventricular apex or left atrial appendage, and significant mitral and tricuspid valve remodeling can be studied on a pre-TAVI CT. Finally (9), a proposal for the most optimal angiographic projections of the aortic valve annulus can be provided.(22)

In the development of TAVI prostheses, insights from CT data acquired after TAVI have permitted modification of existing prostheses with a sealing cuff to ensure good sealing of the aortic annulus, thereby reducing the incidence of significant paravalvular regurgitation. Analysis of the implantation depth and spatial relationship of the prosthetic frame with the atrioventricular node and left and right bundle branches has permitted refining the recommendations for implantation height to avoid conduction disturbances requiring pacemaker after TAVI. Accordingly, CT is an important imaging tool in TAVI.

Transcatheter mitral valve repair/replacement techniques. Regarding transcatheter mitral valve repair or replacement, the required imaging modalities depend on the specific interventional technique. For transcatheter mitral valve repair using the MitraClip repair system® (Abbott Vascular Inc, Santa Clara, California, USA), the periprocedural imaging analysis is mainly performed with 3D echocardiography.(20) This technique visualizes the mitral valve in real time and provides parallel en face views of the exact location and nature of the regurgitation jet. According to these real time 3D echocardiography images, the best position of the clips is determined.(24)

Although that CT may provide a detailed understanding of the distortions in the geometry of the mitral valve and left ventricle, this technique is not

(12)

19 systematically used in patients with mitral regurgitation who are referred for the MitraClip procedure.

For transcatheter repair techniques that pursue mitral annulus size reduction such as the Edwards Cardioband^TM Mitral Valve Reconstruction System (Edwards Lifesciences, Irvine, California, USA) or inter-connected pledging systems (e.g.

Mitralign [Mitralign® Inc., Tewksbury, Massachusetts]), a preprocedural CT provides a good, visual overview of the mitral annulus dimensions and of nearby anatomical structures such as the coronary sinus and left coronary circumflex artery which may be damaged during the intervention. For transcatheter mitral valve replacement, firm anchoring of the new mitral prosthesis within the fibrous, saddle shaped mitral valve complex is very critical and complex.(25) Therefore, accurate prosthesis sizing for transcatheter mitral valve implantation is essential.

Currently, CT provides the highest spatial resolution to measure the mitral valve annulus. However, 3D transesophageal echocardiography may be a very valuable alternative avoiding the use of iodinated contrast and ionizing radiation.

During the procedure, combination of 3D transesophageal echocardiography and fluoroscopy, just side-by-side or with fusion imaging, is essential to achieve an effective reduction in mitral regurgitation and to monitor the occurrence of complications such as impingement of the circumflex coronary artery or migration of the device.

Transcatheter tricuspid valve repair. Transcatheter tricuspid valve repair techniques have been adopted from the experience on the mitral valve. Again, accurate assessment of the anatomy and function of the tricuspid valve is crucial to select the most effective transcatheter therapy (targeting the leaflets with MitraClip or the annulus with Cardioband [Edwards Cardioband^TM Tricuspid Valve Reconstruction System, Edwards Lifesciences, Irvine, California, USA]

or Trialign devices [Trialign® Inc., Tewksbury, Massachusetts]). In addition to geometrical information of the tricuspid annulus and the degree of leaflet tethering, particularly the course of the right coronary artery and its relation to the annulus should be carefully considered to avoid periprocedural arterial damage. This thesis describes how CT can help in the selection of patients with severe tricuspid regurgitation for these therapies.

1

(13)

20

TRANSCATHETER VALVE THERAPIES: CURRENT STATE-OF- THE ART

Transcatheter aortic valve implantation. Currently, more than 100.000 patients with symptomatic severe aortic stenosis have been treated with TAVI worldwide.(26, 27) Continuing advances and refinements of the technique have resulted in improved outcomes. The 30-day mortality rate of 5% observed in the initial PARTNER (Placement of Aortic Transcatheter Valves) trials published in 2010 and 2011 in inoperable and high surgical risk patients(28, 29) has decreased to 1.4% in the high surgical risk patients of the PARTNER II trial published in 2016.(30) In parallel, the frequency of other complications such as paravalvular leakage, cerebral stroke and vascular injury has also significantly reduced.(30)

The less invasive nature of TAVI as compared to open heart surgery results in shorter lengths of hospitalization and post-procedural recovery and, accordingly, TAVI has also been proposed as an attractive treatment option for patients who are at a lower surgical risk. After the initial results of safety and efficacy of TAVI in inoperable and high surgical risk patients, the expansion of TAVI indication to patients with severe aortic stenosis who are at an intermediate surgical risk was evaluated in the PARTNER II trial using the balloon-expandable Edwards prosthesis (n=1011) (30) and in the SURTAVI (Surgical Replacement and Transcatheter Aortic Valve Replacement) trial using the self-expanding CoreValve prosthesis (n=864).(31) In both studies, the TAVI procedure appeared non-inferior to surgical aortic valve replacement for the two-year rates of mortality and cerebral stroke.(30, 31) The two-year incidences of the composite endpoint of mortality and stroke were 19.3% and 21.1% in the respective TAVI and surgical cohorts of the PARTNER II trial (p-value of non-inferiority = 0.001)(30) and 12.6%

and 14.0% in the respective TAVI and surgical cohorts of the SURTAVI trial.(31) In low surgical risk patients, the first data were provided by the Scandinavian NOTION trial (Nordic Aortic Valve Intervention Trial). In this trial, 280 patients who were 70 years or older and had severe aortic stenosis were randomized to TAVI or surgical aortic valve replacement, irrespective of their predicted surgical risk.

(32) Applying this strategy of enrolling all-coming patients, 81.8% of the patients with severe aortic stenosis in this population were considered as low surgical risk and this was the first trial to include those patients for treatment with TAVI.(32) No significant difference in the frequency of the composite endpoint of death, stroke or myocardial infarction was observed between the patients treated with TAVI and those with surgery (p=0.43 for superiority).(32) The characteristics and results of the main randomized trials comparing TAVI and surgical aortic valve replacement in intermediate and low surgical risk patients are summarized in Table 1.

(14)

21 Table 1. Characteristics (upper panel) and principle outcomes (lower panel) of the main randomized trials comparing transcatheter aortic valve implantation with surgical aortic valve replacement in low and intermediate surgical risk patients

Study Number of patients Age^a

(years) Male^a Risk score^a Transfemoral Type of TAVI TAVI SAVR

PARTNER II 2016 (30)

1011 1021 81.5±6.7 54.2% STS score 5.8±2.1

76.3% Sapien XT (100%) SURTAVI

2017 (31)

864 796 79.9±6.2 57.8% Log EuroSCORE 11.9±7.6

94% CoreValve (84%) Evolut R

(16%) NOTION

2015 (32)

145 135 79.2±4.9 53.8% Log EuroSCORE 8.4±4.0

96.5% CoreValve (100%)

Study Follow-up All-cause

mortality Major

stroke Pacemaker implantation

Significant paravalvular regurgitation TAVI SAVR TAVI SAVR TAVI SAVR TAVI SAVR PARTNER II

(30)

2-year 16.7% 18.0% 6.2% 6.4% 11.8% 10.3% 5.5% 0.6%

SURTAVI (31) 2-year 11.4% 11.6% 2.6% 4.5% 25.6% 6.6% 5.7% 1.2%

NOTION(32) 2-year 8.0% 9.8% 3.6% 5.4% 41.3% 4.2% 15.4% 0.9%

a characteristics describing the patients randomized to TAVI

Abbreviations: Log Eur.: Logistic Euroscore; SAVR, surgical aortic valve replacement;

STS: Society of Thoracic Surgeons; TAVI, transcatheter aortic valve implantation;

These studies emphasize the perspective of TAVI as a potential treatment alternative for surgical aortic valve replacement for the large majority of patients with severe aortic stenosis. Data from 141,905 patients who were included in the large database of the American Society of Thoracic Surgeons showed that currently >90% of the patients who are referred for surgical aortic valve replacement are considered as low or intermediate risk.(27) Expanding the NOTION trial, the value of TAVI compared to surgical aortic valve replacement in low surgical risk patients with severe aortic stenosis is currently evaluated in 3 ongoing randomized trials:(33)

1. the PARTNER 3 trial (NCT02675114) including 1328 patients who are treated in a transfemoral only approach with the SAPIEN 3 device.

1

(15)

22

2. the Evolut R low risk trial (NCT02701283) including 1200 patients treated with the self-expanding CoreValve or the iterated version hereof, the succeeding Evolut R device.

3. the NOTION 2 trial (NCT02825134) including 992 patients who are treated with a transfemoral only approach. Interestingly, there are no pre- specified restrictions for the type of used TAVI prosthesis, as long as the prosthesis has Conformité Européenne mark approval.

Furthermore, no lower limits of age are set for the enrolment of patients in these current trials. Therefore, as the majority of these patients is treated with a transfemoral approach, it is expected that a minimalistic approach with only conscious sedation instead of general anaesthesia is pursued in the majority of patients.(33) This approach is associated with even shorter hospitalization and recovery.(27)

As the studies of TAVI in low surgical risk patients are still ongoing, current valvular guidelines recommend (class I indication) TAVI for severe aortic stenosis in patients who are at a prohibitive, high and intermediate surgical risk after evaluation by a multidisciplinary heart team.(11)

Before that TAVI can also be regarded as the treatment option of choice in low surgical risk patients with severe aortic stenosis, certain important safety issues still remain to be addressed. The post-procedural pacemaker implantation rate after TAVI is higher than after surgical aortic valve replacement. A large meta-analysis including 11,210 TAVI patients reported a median post-procedural pacemaker implantation rate of 6% after TAVI with the balloon expandable Edwards SAPIEN prosthesis and of 28% after the self-expanding CoreValve prosthesis.(34) Compression of the surrounding atrioventricular conduction pathways and the left bundle branch by the implanted prosthesis is thought to be the main pathophysiological factor underlying the conduction disorders after TAVI. Therefore, one of the factors that may explain the higher pacemaker implantation rate after the self-expanding CoreValve prosthesis compared to the balloon-expanding SAPIEN devices is that the CoreValve has a longer stent frame that extends deeper into the left ventricular outflow tract and may therefore exert more radial pressure on the surrounding atrioventricular conduction pathways. Although the prognostic implications of a post-procedural pacemaker after TAVI remain debatable in old, high surgical risk patients with a reduced life expectancy, for young patients the implantation of a pacemaker is associated with a decrease in left ventricular function and a reduced survival.(35) On the other hand, with the new generation TAVI prostheses, low pacemaker implantation rates have been reported. A very low pacemaker implantation

(16)

23 rate of only 2.3% after TAVI with the new-generation ACURATE prosthesis was observed by Toggweiler et al. in 175 patients.(36) The inflow portion of this device exerts minimal radial force on the surrounding conduction system.

(36) Nonetheless, also for the SAPIEN 3 prosthesis lower pacemaker rates are achievable. By slight changes in implantation technique in order to deploy the prosthesis more proximal to the base of the aortic cusps, the SAPIEN 3 European approval trial showed a post-TAVI pacemaker rate of 4.0%.(37) This was a markedly reduction compared to the post-TAVI pacemaker rate of 13.3%

that was observed in the initial PARTNER II trial with this device.(38)

Another safety concern around TAVI that should still be further clarified is the occurrence of the phenomena of hypo-attenuated leaflet thickening and reduced leaflet motion as visualized during post-procedural follow-up on high resolution, full heart beat acquired CT scans. These CT findings may indicate the occurrence of prosthetic leaflet thrombosis and have been first described in 2015 by Makkar et al who evaluated the post-TAVI CT scans of 55 patients treated with TAVI from a prospective trial and of 132 patients from two registries who had undergone TAVI or surgical aortic valve replacement.(39) Hypo-attenuated leaflet thickening was observed in 40% of the TAVI trial patients and in 13% of the patients of the two registries. As the incidence of hypo-attenuated leaflet thickening was lower in patients using anticoagulant drugs, the association of these CT findings with prosthetic leaflet thrombosis was hypothesized (39) and the need for adjunctive anticoagulation is currently discussed. Nonetheless, the true clinical implications and pathophysiological mechanisms of these CT-derived observations remain currently unclear. In the recent CT and echocardiography analysis by Vollema et al.(40), hypo-attenuating leaflet thickening was observed in 12.5% of the 128 patients who had undergone a post-TAVI CT. Only in 1 of these patients, this finding was associated with abnormal echocardiographic haemodynamics of the prosthesis. For more conclusive insights, longer follow- up data from prospective clinical trials performing a routine post-TAVI CT are warranted to assess the true impact of this CT phenomena on valve durability, the risk for stroke and the resulting need for anticoagulation.(41) In the GALILEO (Global Study Comparing a rivAroxaban-based Antithrombotic Strategy to an antipLatelet-based Strategy After Transcatheter aortIc vaLve rEplacement to Optimize Clinical Outcomes; NCT02556203) trial a post-procedural regime of anticoagulant therapy with rivaroxaban is compared to antiplatelet therapy with clopidogrel. In the ATLANTIS (Anti-Thrombotic Strategy After Trans-Aortic Valve Implantation for Aortic Stenosis; NCT02664649) trial the direct anti-Xa inhibiting anticoagulant is compared to the standard of use of care: anti-platelet therapy

1

(17)

24

or a vitamin K antagonist if there is already an indication for anticoagulation. The results of these studies are expected to become available in 2019.

In addition, the possibility to further expand the indication of TAVI beyond symptomatic severe aortic stenosis is being investigated. In the EARLY-TAVR (Evaluation of Transcatheter Aortic Valve Replacement Compared to SurveilLance for Patients With AsYmptomatic Severe Aortic Stenosis; NCT03042104) trial the role of TAVI in patients with severe aortic stenosis but who are still asymptomatic is compared to a treatment strategy with active surveillance by randomization of the population according to their exercise capacity as analyzed by treadmill test.

Expansion of TAVI as a treatment also for moderate aortic stenosis in patients with heart failure and a reduced left ventricular ejection fraction is explored in the ongoing TAVR UNLOAD (Transcatheter Aortic Valve Replacement to UNload the Left Ventricle in Patients With ADvanced Heart Failure; NCT02661451) trial.

Transcatheter mitral valve therapies. For moderate and severe mitral regurgitation, the first randomized data supporting a transcatheter treatment option as an acceptable alternative to open heart surgery is derived from the EVEREST (Endovascular Valve Edge-to-Edge Repair Study) II trial in which the transcatheter MitraClip repair system® was evaluated.(42) With this technique, the mitral leaflets are clipped by the device at the place where the maximum effective regurgitant orifice area is located (Figure 1).

The EVEREST II trial randomized 279 patients (mean age 67.3±12.8 years, 63.8% male) 2:1 to transcatheter repair with the MitraClip system or to conventional surgical mitral valve repair. Importantly, 73% of the included patients had organic instead of functional mitral regurgitation while in current clinical practice in European countries, functional mitral regurgitation is the main underlying mechanism of patients treated with this device. In this trial, the 12-month frequency of the composite endpoint of death, redo surgery and of persistent moderate or severe mitral regurgitation was higher in patients who were treated with the MitraClip system compared to those who had undergone conventional surgery (45% vs. 27%, p=0.007).(43) Nonetheless, treatment with the MitraClip system resulted in similar 12 month improvements in clinical outcomes as assessed by scores of quality of life, functional exercise capacity and improvement in left ventricular function, whereas the frequency of major adverse events at 30 days was significantly lower in the patients who were treated with the MitraClip compared to the surgical group (15% vs. 48%, p<0.001).(43)

(18)

25 Figure 1. The MitraClip repair system. After trans-venous access and a right to left atrial transseptal puncture, the catheter with the mounted MitraClip system reaches the left atrium where it is pursued to clip the mitral leaflets at the position of the maximum effective regurgitant orifice area. Retrieved from https://www.vascular.abbott/us/

products/structural-heart/mitraclip-mitral-valve-repair.html, January 2018

The observational Pilot European Sentinal Registry that included 628 patients who were treated with the MitraClip system between 2011 and 2012 in 25 European centers confirmed the efficacy of MitraClip in mitral regurgitation.

(44) In contrast to the EVEREST II trial, in this registry 72% of the patients had functional instead of organic mitral regurgitation and the mean age of the population was also 7 years older, 74±10 years. Acute procedural success, defined as a reduction of mitral regurgitation to a non-severe grade without the occurrence of major adverse events, was achieved in 95.4% of all patients with no differences between the patients with organic and functional mitral regurgitation. Despite a high predicted periprocedural risk as reflected by a logistic EuroSCORE of 20.4±16.7%, the in-hospital mortality rate was low (2.9%).

After 1 year of follow-up, 84.7% of the patients were alive and 94% were free of the re-occurrence of severe mitral regurgitation.(44)

Longer follow-up data corroborating the durability of the MitraClip system are provided by the 5-year results of the EVEREST II trial. After 5 years, the mortality rate in the patients treated with the MitraClip system was comparable to the mortality rate observed in the patients who were treated by surgery (20.8% vs. 26.8%, p=0.36).(43) Interestingly, 26% of the patients who were initially treated with the MitraClip system had undergone a redo surgical procedure

1

(19)

26

during the follow-up. Currently, more than 50.000 patients have been treated with the MitraClip repair system (mitraclip.com, October 2017). To answer the question whether MitraClip is a safe and effective therapy in patients with functional mitral regurgitation, the COAPT trial is currently finalizing the inclusion of heart failure patients with moderate and severe mitral regurgitation who are randomized to optimal medical therapy (including cardiac resynchronization therapy) versus optimal medical therapy and MitraClip [NCT01626079].(45)

Other transcatheter treatments that target different underlying mechanisms of mitral regurgitation have been developed.(46) Some of these new techniques aim at improving mitral leaflet coaptation with mitral leaflet suture devices or by reducing the mitral annular dimensions with annuloplasty devices, while other techniques pursue an entire replacement of the mitral valve by the implantation of a new mitral prosthesis.(25) Figure 2 shows some examples of currently tested transcatheter mitral repair/replacement systems. The Edwards Cardioband^TM Mitral Valve Reconstruction System (Edwards Lifesciences, Irvine, California, USA), a device that is positioned on the posterior side of the annulus and that is subsequently constricted to the desired degree of annular reduction, and the Mitralign® Percutaneous Annuloplasty system (Mitralign Inc., Tewksbury, Massachusetts), a technique where interconnected pledgets are positioned around the mitral annulus that can also be constricted, are examples of treatments that improve mitral leaflet coaptation and reduce mitral regurgitation by reducing the mitral annulus dimensions.(25). For transcatheter replacement of the mitral valve, the Tendyne (Abbott Vascular Inc, Santa Clara, California, USA) and the Intrepid (Medtronic, Dublin, Ireland) devices are examples of currently tested transcatheter mitral valve prostheses.(25) In the global Tendyne feasibility study (n=30) there was a successful implantation with no residual mitral regurgitation in 27 patients and there were no deaths, myocardial infarctions or strokes.(47)

(20)

27 Figure 2. Examples of transcatheter mitral valve repair techniques. In the upper 3 panels, the Mitralign® system is depicted. In the proximity of both commissures, opposing interconntected pledgets are positioned at the anterior and posterior side of the mitral valve annulus. Subsequently the interconnection is constricted in order to obtain the desired level of annulus constriction. In the lower left panel, the transcatheter mitral annuloplasty Edwards Cardioband^TM Mitral Valve Reconstruction system is displayed. In the lower middle panel, the transcatheter mitral valve Tendyne prosthesis is displayed. The lower right image shows the Intrepid prosthesis for transcatheter mitral valve replacement.

- The images of the Mitralign® system were retrieved from: http://www.mitralign.

com/mitral/global (4 February 2018)

- The images of the Edwards Cardioband^TM Mitral Valve Reconstruction System are of courtesy of Edwards Lifesciences, Irvine, California, USA and were kindly provided by Enid Shu.

- The images of Tendyne prosthesis are of courtesy of Abbott Vascular Inc, Santa Clara, California, USA and were kindly provided by Margot Zielhorst

- The images of the Intrepid are of courtesy of Medtronic, Dublin, Ireland and were kindly provided by Thom Derks.

Transcatheter tricuspid valve therapies. Tricuspid valve regurgitation is currently attracting attention since its prevalence is increasing along with heart failure prevalence, increasing number of pacemaker implantation and patients who underwent cardiac surgery and develop at follow-up right ventricular dilatation and tricuspid regurgitation.(10) Current transcatheter techniques for tricuspid valve regurgitation, tested in small cohorts of patients (n<100), include bicuspidalization of the tricuspid valve using sutures, the placement of annulus reducing cinching systems, the implantation of an inflatable spacing system that aims to fill the regurgitant gap, and the implantation of prosthetic valves in the vena cava (Figure 3).(48) Feasibility and safety of transcatheter tricuspid annuloplasty using the Trialign system ® (Mitralign Inc., Tewksbury,

1

(21)

28

Massachusetts) have been shown in the SCOUT (Percutaneous Tricuspid Valve Annuloplasty System for Symptomatic Chronic Functional Tricuspid Regurgitation) trial.(49) With the Trialign system, a device that has been adapted from the transcatheter Mitralign mitral valve repair system, bicuspidalization and annulus reduction of the tricuspid valve is pursued similarly to the surgical Kay procedure (49). After transvenous introduction of a catheter in the right ventricle, the tricuspid annulus is crossed at the level of the antero-posterior and at the level of the septo-posterior commissure. At both positions, pledget sutures are placed on the right atrial and on the right ventricular side of the tricuspid annulus (Figure 3). The atrial pledgets at the different commissural sides are interconnected with a drawstring that can be constricted to create from the tricuspid valve a bicuspid valve with the desired annular reduction (Figure 3).

In the initial report of the SCOUT trial, 15 patients (mean 73.2±6.9 years of age, 87% female) with symptomatic moderate or severe tricuspid valve regurgitation who did not have an indication for a concomitant left-sided valve intervention, were treated with the Trialign system. In all the 15 patients, the device was successfully implanted without the occurrences of death, tamponade, bleeding and strokes.(49) At 30 days, a single (of the four) pledget was detached from the annulus in 3 patients. In the remaining 12 patients, significant reductions in tricuspid annulus dimensions (from 12.3±3.1cm² to 11.3±2.7cm², p=0.019) and in effective regurgitant orifice area (0.51±0.18 cm² vs. 0.32±0.18 cm², p=0.020) were achieved. In addition, repair of the tricuspid valve with the Trialign system resulted in improvements in left ventricular stroke volume (from 63.6±17.9ml to 71.5±25.7ml, p=0.021), in the 6 minute walking distance (from 245±110m to 298±107m; p=0.008) and in improvements in at least one class of the functional capacity NYHA score (p=0.001).(49) The TriCinch^TM Transcatheter Tricuspid Valve Repair System (4Tech Cardio Ireland Ltd, Galway, Ireland) is another transcatheter tricuspid annuloplasty device (Figure 3) that is currently evaluated in an ongoing feasibility trial: the PREVENT (the Percutanous Treatment of Tricuspid Valve Regurgitation with the TriCinch^TM System) trial. With this technique, an anchor is positioned on the tricuspid annulus and an interconnected self-expandable stent is deployed in the inferior vena cava. Applying tension on the interconnecting band results in a septo-lateral reduction of the tricuspid annulus and after achievement of the desired degree of annulus reduction, the inferior vena cava stent is deployed (Figure 3). The enrollment of the 24 patients in this multicenter trial (Switzerland, France, Germany, the Netherlands and Italy) has recently been completed and the evaluated outcomes are acute safety at 30 days, at 3 and 6 months and improvements in quality of life.(50) Interestingly, also the use of the MitraClip device for the transcatheter treatment of severe tricuspid valve regurgitation

(22)

29 has been explored. Nickenig et al. showed improvements in functional and echocardiography outcomes in 64 patients (mean age 76.6±10 years) who had severe tricuspid valve regurgitation and excessive surgical risk (logistic EuroSCORE 27.8±16.7%).(51) The MitraClip could be successfully implanted on the tricuspid valve in 97% of patients. The mean tricuspid regurgitation volume decreased from 57.2±12.8mL/beat to 30.8±6.9mL/beat (p<0.001), the mean effective regurgitant orifice area from 0.9±0.3cm² to 0.4±0.2cm² (p<0.001) and the mean 6 minute walking distance increased from 165.9±102.5m to 193±116m (p=0.007). During the procedure no major adverse events as death, tamponade, conversion to open heart surgery, major vascular injury or strokes occurred.(51)

Another transcatheter tricuspid valve repair technique is the use of the Edwards FORMA^TM Tricuspid Valve Repair System (Edwards Lifesciences, Irvine, California, USA) (Figure 3). The procedure is similar to an implantation of a pacemaker: a subcutaneous pocket is created to fix the rail that is advanced through the left subclavian vein and attached to the right ventricular apex. The spacing device is positioned within the regurgitant tricuspid valve to fill coaptation gap between leaflets and to provide systolic leaflet coaptation. Feasibility and good 1 year safety data of this device have been provided by Perlman et al.(52) In 18 inoperable patients (mean age 76±10years) who had severe symptomatic tricuspid regurgitation, the device could be successfully implanted in 16 patients (mean procedural time 129±36minutes). In one of the other two procedures, device dislodgement due to insufficient right ventricular anchoring occurred while the other procedure was complicated with a right ventricular perforation that necessitated the conversion to an emergent but rescuing open surgical procedure. After 1 year of follow-up, there were no deaths or other serious complications while functionally, capacity class improved in 79% of the patients to a level where they experienced no or only slight limitations, e.g. NYHA class I or II. By echocardiography, severe tricuspid valve regurgitation was improved to non-severe in 11 of the 16 patients (69%) at 30 days, while after 1 year.(52)

Another transcatheter strategy that relieves the symptoms caused by severe tricuspid valve regurgitation is the heterotopic implantation of transcatheter valves in the superior and/or inferior vena cava. Functionally, the severe regurgitant tricuspid valve is replaced by a caval valve. As a result, the right atrium becomes part of the right ventricle. Feasibility of this technique and the first-in-man reports were shown by Lauten et al.(53), and currently, the HOVER (Heterotopic Implantation Of the Edwards-Sapien Transcatheter Aortic Valve in the Inferior Vena Cava for the Treatment of Severe Tricuspid Regurgitation, NCT02339974) trial is enrolling 15 high or inoperable risk patients with severe tricuspid valve regurgitation to explore this strategy.

1

(23)

30

Figure 3. Examples of transcatheter tricuspid valve repair techniques. In the upper panel including 6 images, the Trialign® system is depicted. After entering the right ventricle, the tricuspid annulus is crossed at the level of the antero-posterior and at the level of the septo- posterior commissure. In both positions, pledget sutures are placed on the right atrial and on the right ventricular side of the tricuspid annulus. Between both pledgets there is an interconnected drawstring on the atrial side that can be constricted thereafter to create from the tricuspid valve a bicuspid valve with the desired annular reduction.

In the middle panel, the Tricinch^tm Transcatheter tricuspid valve repair system is displayed. With this technique, an anchor is positioned on the tricuspid annulus and a connected, self-expandable stent is deployed in the inferior vena cava. Tension on the interconnecting band is applied until the desired septo-lateral annulus reduction has been achieved.

In the lower left panel, the Edwards Cardioband^TM Tricuspid Valve Reconstruction system is displayed. The lower right panel shows an image of the Edwards FORMA^TM Tricuspid Valve Repair System. After creating a subcutaneous pocket and attachment of a rail into the right ventricular apex, this inflatable spacing device is positioned within the regurgitation orfice area of the leaking tricuspid valve to provide a surface for systolic leaflet coaptation.

- The images of the Trialign® system were retrieved from: http://www.mitralign.com/

tricuspid (18 February 2018)

- The images of the TriCinch^TM system were retrieved from: http://4techtricuspid.com/4tech- tricinch-transcatheter-tricuspid-valve-repair-ttvr-system/ (18 February 2018)

- The images of the Edwards Cardioband^TM Tricuspid Valve Reconstruction System and of the Edwards FORMA^TM Tricuspid Valve Repair System are of courtesy of Edwards Lifesciences, Irvine, California, USA and were kindly provided by Enid Shu.

(24)

31 Concomitant transcatheter treatment options. In the vast majority of patients who are referred for transcatheter valvular repair or replacement, the degenerated valve is not the only manifestation of cardiac disease. Related also to the ageing process, and to shared interrelated pathophysiological mechanisms, diseases as significant coronary artery atherosclerosis, atrial and ventricular remodeling and rhythm disorders as atrial fibrillation are frequently coexisting in patients with significant valvular heart disease. The severity of these co-morbidities may have important implications for the pre-procedural evaluation of the patient with valvular heart disease and may determine whether the patient should undergo a combined surgical procedure or combined transcatheter treatments.

One of these transcatheter therapies is the left atrial appendage closure for patients with atrial fibrillation who have contraindications for anticoagulant treatment (Figure 4).

Figure 4. Transcatheter left atrial appendage occlusion with the WATCHMAN device.

Image retrieved from: http://www.watchman.com/how-watchman-device-works.html on 4 February 2018

The non-inferiority of transcatheter occlusion of the left atrial appendage compared to systemic anticoagulant therapy for stroke prophylaxis in patients with atrial fibrillation was initially demonstrated in the PROTECT AF (WATCHMAN Left Atrial Appendage System for Embolic Protection in Patients with Atrial) trial.(54) In this trial, 707 patients with non-valvular atrial fibrillation underwent transcatheter implantation of the WATCHMAN (Boston Scientific, Marlborough, Massachusetts, USA) left atrial appendage occlusion device and were then randomized to discontinuation or to continuation of systemic anticoagulant drugs (warfarin). After a follow-up of 1065 patients-years, the event rate of the composite endpoint including cerebral stroke, systemic embolism and

1

(25)

32

cardiovascular death was 3.0% in the group that discontinued warfarin vs. 4.9%

in the group that continued warfarin and this implicated non-inferiority of the WATCHMAN left atrial appendage occlusion system.(54) These initial findings of safety and efficacy of transcatheter left atrial appendage occlusion for stroke prevention in patients with atrial fibrillation were corroborated in the real-world EWOLUTION registry that included 1025 patients (73.4±9 years) who were treated with the WATCHMAN device. In this registry, the procedural implantation success was 98.5%; the serious device or procedure related adverse event rate was 3.6% of which major bleeding (22.5%) and pericardial effusion (16.1%) were the most frequent complications.(55) Interestingly, the patients in this registry were at a serious risk for ischemic embolic events, as indicated by a mean CHA₂DS₂-VASc score of 4.5±1.6 which reflects an estimated annual stroke risk of 5% in patients with atrial fibrillation who are not treated with anticoagulation. In this registry, the 1-year stroke rate after the WATCHMAN implantation was as low as 1.1%.(56) Besides the WATCHMAN device, the Amplatzer Cardiac Plug and its iterated version, the Amulet (St. Jude Medical, Saint Paul, Minnesota, United States of America) are a second type of transcatheter left atrial appendage occlusion devices. Compared to the WATCHMAN, the Amplatzer Cardiac Plug has a wider lobe while its length is shorter. Therefore, these occlusion devices may also be suitable for shorter appendages with a wider proximal orifice(57) The largest published experience with this device is provided by a European and Canadian multi-center registry that included 1047 patients (mean age 75±8 years). For this device, the reported procedural success rate was 97.3%. Major adverse events occurred in 52 (4.97%): major bleeding n=13 (1.24%), cardiac tamponade also n=13 (1.24%), in-hospital mortality n=8 (0.76%).(58) At one-year follow-up, 2.3% of the patients had experienced a thrombo-embolic event and this provided a 59% risk reduction compared to the estimated annual stroke risk by the CHA₂DS₂-VASc score.(58)

For the preprocedural work-up analysis of these patients, CT may be very valuable as the images can be used for a comprehensive analysis of the valvular disease and the left atrial appendage anatomy.

OUTLINE OF THE PRESENT THESIS

Of all 3D imaging techniques that are used and investigated in the field of cardiology, this thesis focuses on the potential role that CT may have in the diagnosis and evaluation of valvular heart disease and in the periprocedural