Imaging guidance is of major importance for successful outcomes in the interventional treatment of structural, congenital and valvular heart disease. Cardiac transcatheter treatment of these diseases is a rapidly growing field and, as a result, there has been substantial progress in image-guidance technology.
During interventions, the major imaging focus is currently on echocardiography, which fulfils needs by providing a realtime imaging solution. The cost–value ratio is very reasonable, it is globally used and most interventionalists are familiar with this imaging format.
In the past, in most catheter laboratories percutaneous interventions were mainly guided by 2D transoesophageal echocardiography (TEE) or intracardiac echocardiography (ICE). This combined approach led to a reduction of the dose of ionising radiation. However, 2D imaging modalities have limitations due to the fact that there are only two spatial dimensions. It is not possible to get an idea of the 3D character of the structures in one single view and thus several images have to be acquired, which can lead to prolonged imaging times. Furthermore, the defects often have a more complex morphology, which is suboptimally characterised with 2D imaging.
A new generation of transoesophageal echocardiography probes, with a novel matrix array technique, was recently introduced. Data from a pyramidal ultrasound beam are acquired, allowing 3D visualisation of cardiac structures in realtime, which overcomes some of the major limitations of 2D TEE. The pathomorphology of defects, neighbouring structures, wires, catheters and devices can be visualised quickly and accurately while presenting the 3D character in a single image.
This article focuses on early experiences with this novel technique in the guidance of miscellaneous percutaneous cardiac interventions using a matrix array 3D TEE probe (X7-2t, Philips Medical Systems) connected to a 3D-capable echocardiographic system (iE33, Philips Medical Systems).
Clinical Examples
Closure of Patent Foramen Ovale
Since the first percutaneous patent foramen ovale (PFO) closure, performed by Bridges et al. in 1992,1 transcatheter closure of PFO has become a routine procedure for patients suffering from cryptogenic stroke, with low complication and recurrence rates.2–4
Multiplanar 2D TEE and ICE are widely used as imaging techniques for guidance of PFO procedures in most centres worldwide. However, limitations are evident: wires, catheters and devices cannot be fully imaged in relation to the surrounding structures. With a 2D imaging modality, scanning of several views is necessary to mentally reconstruct the 3D anatomy. 3D TEE provides spatial orientation in one view, which allows monitoring of the entire procedure with convincing imaging quality. 3D perspectives enable monitoring of all steps of the intervention: from the passage of the guidewire and delivery catheter through the PFO to the final assessment before device release. Martin-Reyes et al.5 confirmed 3D TEE to be feasible for guiding transcatheter patent foramen ovale closure.
Figure 1 illustrates the guidance of a PFO closure procedure with realtime 3D TEE. Figure 1A demonstrates the passage of the guidewire through the PFO track. The wire stretches the PFO tunnel and the tenting of the channel within the inter-atrial septum is clearly visualised. After balloon sizing, the specific introducer sheath is advanced. In Figure 1B crossing of the PFO with the delivery sheath is shown. Figure 1C shows an Amplatzer occluder with opened left atrial disc in the left atrium. The occluder is then pulled back towards the septum and the right atrial disc is deployed. Before device release, the correct and secure position of the occluder has to be ascertained. One single 3D perspective shows the definite device position, which is a major advantage of realtime 3D TEE. When the device position is satisfactory it can be released, as demonstrated in Figure 1D with a perpendicular perspective. In addition, complications can be detected immediately and accurately. Figures 1E and 1F show device misplacement. This patient had a large atrial septum aneurysm. First, a 25 mm Amplatzer PFO occluder was implanted. The entire occluder slipped into the PFO tunnel. Figure 1E shows an en face 3D perspective of the left side of the atrial septum where both atrial discs can easily be identified on the left atrial side. This occluder was retrieved and replaced by a 30mm Amplatzer PFO occluder.
On this occasion both discs of the device embraced the septum secundum adequately. Figure 1F shows clearly that, in 3D TEE, only a single disc can be visualised on the left atrial side. Conventional 2D TEE enables the assessment of appropriate device position but requires scanning of several image planes. By contrast, 3D TEE offers exact visualisation of the device position and its relation to the inter-atrial septum in a single view.
Closure of Atrial Septal Defects
Since the first transcatheter closure of an atrial septal defect (ASD) by King and Mills in 1976,6 device closure of ASD has become an accepted alternative to surgical closure.7,8 Normally, 2D TEE or ICE is used in combination with fluoroscopy to monitor the interventional procedure.9 However, due to the limitations of 2D imaging it is difficult to obtain optimal visualisation of the complex anatomy of ASDs. The dynamic variations of the defects are insufficiently appreciated by 2D imaging. 3D TEE provides en face views of the inter-atrial septum and, therefore, can clearly identify the morphology of the defect as well as its relation to surrounding structures, which cannot be achieved with any other available imaging technology.10 Lodato et al.11 recently showed that real-time 3D TEE is feasible to guide transcatheter ASD closure.
The examples shown in Figure 2 demonstrate that, in our experience, 3D TEE guidance of ASD closure procedures is helpful by adding unique new views of the defects. Figure 2A shows a 3D TEE stop-frame image of an ASD displaying an en face view from the left atrium. This view allows an assessment of the complete circumference of the ASD. The size, shape and rim of the defect can be judged precisely in this single view. The exchange wire passes through the ASD. A tissue bridge can be clearly identified – this was not viewable with 2D TEE. In Figure 2B the delivery sheath is positioned. The left atrial disc is then delivered and pulled back towards the interatrial septum. The final position of the occluder is shown in Figure 2C. It is also possible to close multiple defects in one procedure. In Figures 2D and 2E an example of a patient with two ASDs is shown. Preferably, the smaller defect is closed first, with the second, larger defect closed second. In Figure 2D we see two Amplatzer occluders; the smaller one is placed in a more caudal located defect.
This occluder is sandwiched by the larger Amplatzer occluder located in the cranial defect, which covers the larger ASD. Both occluders are still attached to their delivery cables. 3D TEE shows, in a single view, that both occluders are positioned correctly in relation to each other and to the inter-atrial septum. Figure 2E illustrates the final result in an en face view from the left atrial side.
Percutaneous Mitral Valve Procedures
Balloon Mitral Valvuloplasty
3D TEE provides en face views of the mitral valve from left atrial and left ventricular aspects. Detailed assessment of the mitral valve anatomy is possible. Ring size, precise valve morphology and pathomorphology of the leaflets can be visualised in detail. This is extremely useful in assessing the mitral valve before, and during, interventional and surgical procedures.
Ning et al.12 recently reported very promising data from their experiences with live 3D TEE in mitral valve surgery, including pre-operative diagnosis and evaluation of the function of both native and prosthetic valves. They found additional defects of the mitral valve, or of the valvular apparatus, in three out of 24 patients who were not previously diagnosed by 2D TEE. According to these findings, eligible patients for specific mitral valve procedures can be identified more accurately and the likelihood of procedural success is more predictable. Since its introduction by Inoue et al.13 in 1984, percutaneous mitral comissurotomy (PMC) has been successfully and safely performed in large series of patients at various centres in different countries.14–16
Our experience shows that it is feasible to monitor balloon-mitral valvuloplasty with 3D TEE. By receiving a 3D impression of the valve’s less favourable anatomy, valve calcification (especially calcification of the comissures) and severe subvalvular disease can be assessed accurately. The procedure described was performed in a 42-year-old patient with a history of rheumatic fever. Figure 3A shows the valve anatomy with typically thickened leaflets. There was no severe calcification and no relevant subvalvular disease. The thickened leaflets and the narrowed mitral valve area are clearly visualised in an en face view from the left atrium. The Inoue balloon is positioned in the left atrium before passing the mitral valve. Figure 3B demonstrates the correct position of the inflated Inoue balloon in the stenotic mitral valve during valvuloplasty. After valvuloplasty the mitral valve area, as well as the degree of mitral regurgitation, have to be assessed to ensure a good procedural result.
Mitral Valve Repair
For patients who suffer from mitral valve regurgitation, a percutaneous non-surgical repair with the MitraClip® can be an alternative. Results from studies so far show that the MitraClip can provide a successful reduction of mitral regurgitation. Furthermore, studies have showed that with resumption of proper valve function, left ventricular remodelling significantly improved at 12-month follow-up.17–19
Evalve has developed catheter-based technology that, by apposing the edges of a regurgitant mitral valve, results in edge-to edge repair. The MitraClip grasps affected parts of the anterior and posterior leaflet, similar to the Alfieri surgical procedure. The challenge of the technique is to position the clip perpendicularly to the co-aptation line of the mitral leaflets in the A2 and P2 scallops of the leaflets. The intention is not to favour one side of the two created mitral orifices. Figure 3C shows the perpendicular positioning of the MitraClip to the line of co-apation from a left-atrial view before translation through the mitral valve. The leaflets are grasped from the ventricular side in the middle (A2 and P2 parts) of the anterior and posterior leaflet. The two created mitral orifices have a similar size, which is considered a good result, as shown in Figure 3D, where the detached clip can be identified from the left ventricular side. In this patient, mitral regurgitation declined from grade IV before the procedure to grade II after the procedure.
Paravalvular Mitral Leak Closure
Valve replacement surgery is the second most common cardiothoracic operation after coronary artery bypass grafting.20,21 A potential sequel to surgery is the development of a paravalvular leak due to incomplete apposition of the sewing-ring to the native tissue. This may be a consequence of suture dehiscence. Detection has increased as a result of improved techniques, particularly transoesophageal echocardiography.22–25 Percutaneous transcatheter closure techniques have also been applied to paravalvular leaks.26–28 A routine 2D TEE may document the presence and severity of regurgitation and allows visualisation of the correct deployment of the device, detection of normal prosthetic function before device release and immediate diagnosis of complications in realtime, as Cortes et al.29 reported recently.
However, 2D TEE is often not sufficient to accurately assess the location, exact size and shape of the defect, the course of the leak and, hence, the likelihood of successful percutaneous repair. The irregular 3D character of these defects has to be taken into account, and this cannot be imaged entirely by 2D TEE. Thus, guidance of paravalvular leak closure is very challenging, even for experienced echocardiographers and interventionalists. If the defect itself cannot be adequately visualised echocardiographically, percutaneous closure is less likely to be successful.
3D TEE is able to provide an en face view of the mitral valve and surrounding structures, which allows complete and detailed assessment of the 3D character of paravalvular leaks. This information is crucial, particularly for the determination of the technique and the choice of device size and shape. The 3D TEE images of a paravalvular closure procedure shown in Figure 3 emphasise these facts. Figures 3E and 3F show an example of a 51-year-old woman who had a mechanical mitral valve replacement (SJM 29mm) because of severe mitral regurgitation after a bout of endocarditis. A few months later she developed severe mitral regurgitation due to a paravalvular leak. Figure 3E shows a very clear image of the size and the oval shape of an anterior-medial located paravalvular leak. A 14.5mm Amplatzer-Vascular-Plug III (AVPIII) occluder was implanted. This device is especially designed for paravalvular leak closure and has an oval shape. The final device position is shown in Figure 3F, where the occluder is properly aligned to the defect.
Percutaneous Aortic Valve Procedures
Percutaneous Aortic Valve Replacement
Degenerative aortic valve disease is the most common valvular heart disease and its prevalence increases with age.30 The morbidity and mortality of surgical aortic valve replacement are increased in elderly patients with multiple high-risk co-morbidities.31–33
Percutaneous aortic valve replacement has become an alternative therapeutic option to surgical valve replacement in selected elderly patients with an unacceptably high risk for surgery.34,35 Early improvements in left ventricular ejection fraction and mitral regurgitation following pecutaneous aortic valve implantation have been reported.36 Recently, it has been shown that echocardiography is important in case selection, for guiding valve placement and in detecting complications during and after the procedure.37
Figures 4A and 4B show an example of a 82-year-old woman with a severe symptomatic aortic stenosis. Figure 4A shows the aortic valve in a 3D TEE image, giving a spatial impression of the aortic valve and its surrounding structures. Figure 4B shows the completely debilitated valve in 3D TEE imaging. An advantage is the 3D view of the aortic valve, the left ventricular outflow tract (LVOT) and its surrounding structures. The relation of the anterior leaflet of the mitral valve to the LVOT can be imaged particularly well. This is of major importance for the final positioning of the self-expandable CoreValve® prosthesis to ensure that the movement of the anterior mitral leaflet is not constrained.
As the technical aspects and clinical understanding of this technique continue to evolve, echocardiography and especially 3D TEE will have a crucial role in the future development of the treatment of aortic valve disease.
Paravalvular Aortic Leak Closure
Although significant data from larger patient cohorts are missing, it could be shown that transcatheter treatment of paravalvular aortic leaks is a technically feasible, but challenging, procedure.38,39
We report on the history of a 50-year-old man who had surgical aortic valve replacement with a mechanical valve prosthesis (ATS Medical, 25mm) due to aortic insufficiency of a bicuspid aortic valve in July 2007. A paravalvular leak with severe aortic regurgitation was diagnosed four months after surgery. Dyspnoea on exertion had increased since then. Figure 4C shows the posterior-located paravalvular leak in 2D TEE with colour Doppler. Figure 4D shows the shape of the defect in a 3D en face view of the aortic valve. In Figure 4E the irregular and complex shape of the paravalvular leak can be identified even more clearly in the lateral view of the prosthetic ring.
According to the information added by 3D TEE images we chose an oval-shaped device, and Figure 4 shows the final result after placement of an Amplatz-Vascular-Plug III occluder (5x14mm), which is properly aligned to the defect without constraining the prosthetic valve.
Left Atrial Appendage Closure
The left atrial appendage (LAA) is the most common source of cardiac thrombus formation in patients with atrial fibrillation. More than 90% of all cardiac thrombi forming in the left atrium in patients with non-rheumatic atrial fibrillation originate in the LAA.40 Occluding the LAA would therefore seem to be a logical approach to minimise the risk of thrombus formation and subsequent embolisation.
Surgical attempts40,41 at the LAA have drawn attention to percutaneous interventional approaches and led to the development of devices specifically designed for this purpose. The earliest device, the percutaneous Left Atrial Appendage Occluder (PLAATO),42 is no longer available. The Watchman Left Atrial Appendage System43 and, most recently, the Amplatzer Cardiac Plug III (ACPIII) are currently used in clinical practice.
To determine eligible patients for the procedure and device selection, measurements of the LAA orifice and length are of major importance. Using 2D TEE, several views have to be scanned to get an idea of the size and shape of the LAA orifice. 3D TEE enables us to get an en face view of the LAA orifice, which allows us to clearly identify the dimensions in one view. Shah et al.44 recently showed that realtime 3D TEE is feasible for the visualisation and quantitative analysis of the LAA orifice area and correlated well with a 64-slice cardiac CT.
Figure 5 demonstrates the guidance of a LAA closure procedure with a Watchman device in a patient with long-lasting atrial fibrillation. Figure 5A shows the LAA, which is assessed for pre-procedural thrombi. Figure 5B provides an en face view of the LAA orifice and the left upper pulmonary vein, which is a unique view available with 3D TEE. In Figure 5C the guidewire is advanced into the LAA. Figure 5D shows the delivery sheath placed in the LAA and Figure 5E demonstrates the final position of the delivered Watchman device. Fixation barbs around the mid-perimeter engage and fix the occluder to the wall of the LAA. The major advantage of 3D TEE in this intervention is that it provides an en face view of the LAA entrance, which gives a clear impression of the orifice of the LAA – the landing zone of the finally positioned occluder.
Conclusion
Realtime 3D TEE, as a novel technique to guide cardiac interventions, adds previously unavailable anatomical information of cardiac structures and defects online. Information on the pathomorphology and the relationship of defects to neighbouring structures can be provided immediately and accurately. Our clinical examples underline the enormous potential benefit 3D TEE can provide. We conclude that high-quality realtime 3D TEE imaging with good spatial and temporal resolution has great potential for guiding cardiac interventions and, therefore, may be remarkably clinically valuable in the future.