Investigations and research Structural heart disease interventions: rapid clinical growth and challenges in image guidance

J.D. Carroll University of Colorado, Department of Medicine, Division of Cardiology, S.J. Chen Colorado, USA. M.S. Kim A.R. Hansgen A. Neubauer Philips Research, Briarcliff Manor, New York, USA. University of Colorado, Department of Medicine, Division of Cardiology, Colorado, USA. O. Wink Philips Healthcare, Bothell, Washington, USA. University of Colorado, Department of Medicine, Division of Cardiology, Colorado, USA.

Structural heart disease (SHD) interventions modification of anatomical structure and represent a broad category of percutaneous function, using balloon dilatation and tissue treatments for patients with both congenital ablation, to the deployment of various plugs, heart disease (CHD) and acquired heart disease valves, clips and cinching devices (Table 1). involving structural and functional abnormalities of heart valves, cardiac chambers The current status of SHD interventions ranges and the proximal great vessels [1, 2]. In the last from well-established procedures, such as five years, there has been an explosion in the percutaneous balloon valvuloplasty for stenotic number of innovative approaches to these valve conditions, which has been incorporated catheter-based treatments, ranging from the into clinical practice guidelines as the preferred

Time period of Percutaneous SHD interventions Image guidance modality F emerging into Table 1. The development of practice structural heart disease (SHD) interventions. Pre-2000 Balloon valvuloplasty Fluoroscopy Balloon septostomy Catheter ablation of SVT Last 5 years Device closure of PFO, ASD, VSD, PDA Fluoroscopy plus ICE and TEE Repair of paravalvular leaks Early mapping systems Catheter ablation of atrial fibrillation, VT Alcohol septal ablation for hypertrophic cardiomyopathy Next 5 years Mitral valve repair Fluoroscopy plus 2D and 3D TEE Aortic and pulmonic valve implantation and ICE Next generation devices for PFO and ASD Advanced mapping systems Left atrial appendage occlusion devices Fluoroscopy overlay on 3D CTA, MRA, and angio reconstructions Future Valve replacement with a variety of 3D imaging wedded to mechanical and biologic types robotic navigation Repair of all valves Advanced ICE imaging Biodegradable closure devices Myocardial regenerative therapies via intramyocardial delivery Device closure of all LV aneurysms and pseudoaneuysms Shunts for Complex CHD

SVT: supra-ventricular tachycardia; PFO: patent foramen ovale; ASD: atrial septal defect; VSD: ventricular septal defect; PDA: patent ductus arteriosus; VT: ventricular tachycardia; ICE: intracardiac echocardiography; TEE: transesophageal echocardiography; CTA: computer tomographic angiography; MRA: magnetic resonance angiography; CHD: congenital heart disease. MEDICAMUNDI 52/2 2008 43 E Lesion General Specific patient groups or Table 2. The frequency of SHD. defining statistic PFO 1:4 to 1:5 of population. • 36% - 59% of young adults presenting with cryptogenic stroke have a PFO.

Aortic Most common etiology is related • 6% of all people over age 90 have stenosis to aging. hemodynamically significant aortic stenosis. • In patients over age 80, operative mortality of surgical aortic valve replacement approaches 30%. Mitral Frequently accompanies heart • Affects as many as 9.3% of people age 75 regurgitation failure from all causes and surgical and older. therapy not feasible. • Of the 5 million people suffering from heart failure in the USA, 15% - 20% have moderate to severe mitral regurgitation. Atrial The vast majority of patients with • The prevalence of stroke associated with AF fibrillation AF currently require long-term full increases with age. (AF) anti-coagulation. The left atrial • AF is thought to be responsible for one-sixth appendage is the site of thrombus of all ischemic strokes in people over age 60. formation in 90% of patients with • The risk of stroke among all patients with AF non-valvular AF. is about 5% per year, which is about five to six times the risk of age-matched patients in sinus rhythm.

AF: atrial fibrulation; PFO: patent foramen ovale; USA: United States of America. therapeutic approach in specific clinical situations, leading to embolism are ongoing. Should these E The volume of patients to investigative technologies still in development, clinical trials demonstrate either superiority or undergoing SHD such as percutaneous valve implantation [3]. equivalence of device therapy versus chronic interventions is rapidly While many pivotal, industry-sponsored trials anti-coagulation in preventing embolic stroke increasing. are currently enrolling patients, and will have in patients with atrial fibrillation, another major results available in the next 2 - 5 years, other expansion of patients eligible for SHD technologies remain in the very early phases of interventions will occur. Finally, catheter based concept and design development (see the treatments for valvular aortic stenosis and mitral “Future” category in Table 1). As Tables 1 and regurgitation have already shown preliminary 2 illustrate, the potential for growth in this results that will likely lead to the treatment of unique clinical area is staggering. patients who had previously been ineligible for traditional surgical valve repair or replacement. The volume of patients undergoing SHD In addition, many patients will potentially be interventions is rapidly increasing, and may even switched from open surgical to catheter-based surpass the number of many vascular treatments if comparative studies show benefit interventions performed within the next decade. with less risk. The rate of growth, however, depends heavily on the outcome of several ongoing clinical trials SHD interventions show a significant departure investigating pathological conditions that are from the inherent nature of the two prior waves very common in adult cardiology practices of new interventional treatments: percutaneous (Table 2). For example, device closure of patent coronary intervention and non-coronary vascular foramen ovale (PFO) in adults to prevent disease interventions, such as carotid stenting. embolic stroke and reduce migraine frequency Unlike these vascular therapies, where over-the- are the subject of several trials that, if positive wire technologies in the well-defined space of (i.e. demonstrate that device closure is superior small branching vascular trees are used, SHD to medical therapy), will make tens of thousands interventions frequently involve navigation in of patients immediately eligible for treatment open 3D space, defined by relatively large cardiac [4, 5]. In addition, the incidence of atrial chambers, interaction with moving targets, such fibrillation continues to increase and requires as heart valves, and deployment of devices, such lifelong anti-coagulation to prevent embolic as occluders and heart valves, that function stroke. Clinical trials investigating the ability of quite differently from traditional vascular left atrial appendage (LAA) occlusion devices scaffolds. These differences subsequently impact 44 MEDICAMUNDI 52/2 2008 to prevent thrombus formation in the left atrium procedural performance by relying heavily on 1 F Figure 1. These two panels show a model of the heart of a patient who has a secundum atrial septal defect (ASD) with a deficient aortic rim. A viewing window was created in the right atrial wall to allow examination of the defect. The right panel shows the enlarged view of the defect. This model was made from cardiac computed tomography angiography (CTA) data and the transformation into a rapid prototyping file (stl) was done by the 3D Research Lab at the University of Colorado. This model is made of the operator’s knowledge (both structural and simulators are designed to familiarize operators a semi-translucent and soft material spatial) of cardiovascular anatomy and with various aspects of catheter-based closure that mimics heart tissue. Rapid physiology, training with unique navigational (i.e. anatomy, imaging modalities, etc.). This prototyping allows interventionalists devices, incorporation of new procedural skills approach also allows for the added advantage of to create models prior to challenging and familiarity with novel image guidance enabling the early learning curve of physicians SHD interventions, as well as for technologies. to occur during simulation and not on real general training purposes [8]. patients, who could potentially be exposed to Interventional cardiologists performing SHD increased risk due to the inherent novelty of the interventions must understand anatomy to a procedures. degree similar to that of cardiac surgeons. Unlike surgeons, however, interventional Recently, there have been several technological cardiologists do not have the advantage of advances in imaging modalities used in both learning cardiovascular anatomy in the setting the evaluation and treatment of SHD [7]. E Recent technological of direct anatomic exposure during open-heart Ultrasound guidance has increasingly been advances in imaging surgery. Interventionalists instead rely heavily used in SHD interventions. The emergence of SHD include ICE, TTE on medical images produced by ultrasound, percutaneous closure of atrial septal defect and TEE. computed tomography angiography (CTA), and (ASD), PFO and ventricular septal defect magnetic resonance angiography (MRA), which (VSD) marked the routine incorporation of can be processed into 3D formats and are ultrasound imaging. Transthoracic somewhat useful in providing an understanding echocardiography (TTE) and transesophageal of patient-specific anatomy. Although 3D echocardiography (TEE) are routinely used in reconstructions and graphical display of children and adults to assess defect and device modalities, such as cardiac CTA, are infrequently sizes, guide device deployment and assess the used for traditional diagnostic purposes, these procedural result. Furthermore, the applications are becoming much more important development and incorporation of intracardiac to the interventionalist when planning both echocardiography (ICE) has provided image structural and vascular procedures. clarity equivalent to that achieved with TEE, without the burdens associated with prolonged Beyond their required understanding of esophageal probe placement (i.e. increasing the structural and spatial cardiovascular anatomy, depth of anesthesia, need for an expanded team interventional cardiologists performing SHD to perform the procedure, etc.). interventions must also learn new procedural skills and gain familiarity with novel The research group at the University of navigational and therapeutic devices. Within Colorado has taken the emergence of SHD the last five years, simulators designed to train interventions, requirements for training in an operators in the intricacies of catheter-based entirely new procedural skill-set, and the need interventions (i.e. increased hand-to-eye for more in-depth anatomical understanding as coordination, translating the manipulation of the impetus to develop the technical capability objects in 3D space with movements on a 2D to transform medical images to physical models screen, etc.) have been developed. In of patients’ hearts (see Figure 1) [8]. An interventional cardiology, simulation-based accurately sized physical model of the patient’s training has been used in both the investigative heart is a powerful and efficient tool for phase as well as the post-approval roll-out of a visualizing and simulating the sequential steps variety of SHD interventions [6]. These of a SHD intervention. While computer graphics MEDICAMUNDI 52/2 2008 45 2 physicians and staff, is clearly reflected internationally by major interventional cardiology meetings shifting from SHD being a small niche to having equivalent time and emphasis to that devoted to coronary interventional sessions [9-12]. It has also resulted in many cardiac surgeons training in catheter-based skills and image guidance, as well as the development of hybrid surgical suites equipped with advanced imaging equipment.

The need for 3D imaging to guide structural heart disease (SHD) interventions

Because traditional 2D imaging modalities remain limited in their ability to represent the complex 3D relationships present in SHD, the growing number of SHD interventions performed worldwide has heightened the need for advanced 3D imaging modalities. Although complex moving 3D structures, such as heart valves and chamber defects, can be imaged with 2D cross-sectional ultrasound images, they require the physician to mentally integrate the slice images into the context of a 3D object. G These challenges in imaging, both acquisition Figure 2. A challenging structural allow 3D visualization, they are limited by an and interpretation, are compounded when heart disease (SHD) intervention is inherent lack of realism and cannot be held in performing a SHD intervention. The delivery closure of a ventricular septal defect the physician’s hands, turned and studied using catheter, device and target are often very difficult (VSD). There is complex anatomy direct visualization. In addition, the planned to visualize simultaneously in single a cross- of both the defect and surrounding pathway of catheters to the SHD target is better sectional image. As a result, a significant amount tissues that must be understood understood in a 3D spatial representation, which of time is spent searching for the optimal view and visualized during the procedure. is more realistic than the 2D projection images to evaluate and guide the procedure, integrating The delivery system can come from obtained by fluoroscopy. Likewise the multiple images to make clinical and technical several routes, including the superior deployment of a device can be simulated with an decisions, and assessing the procedural results vena cava, the inferior vena cava, immediate and clear understanding of its after device deployment. These challenging and transseptally through the mitral potential impact on surrounding structures visual-spatial and technical tasks are the primary valve. The upper two panels show a (Figure 2). reasons underscoring the steep learning curves model made from the CTA of a associated with many SHD interventions. Such patient with a VSD and helped plan Given the growing number of SHD complexities are also the reason why SHD the procedure. Note the blue interventions being performed worldwide, and interventions often require a team of physicians, catheters in the model. This the uniqueness of the training and practical including experts in echocardiography and simulation of possible catheter experience required to safely perform these advanced imaging. pathways to the VSD proved that procedures compared to vascular interventions, the best approach to close this VSD it was an inevitable reality that a sub-specialty Alignment of delivery catheters and devices to was from the superior vena cava. within interventional cardiology is now being the 3D target, a technical aspect that is common The bottom left panel shows the created. Traditionally, pediatric interventional to many SHD interventions, is likely facilitated device (AGA Medical Corporation) cardiologists have developed many of the by 3D imaging. Two examples are instructive: and the bottom right panel shows techniques employed in SHD interventions. In Firstly, large ASD devices are deployed in a the implanted device, visualized by the last decade, however, the increasing number sequential fashion with the left atrial disc 3D TEE, after successful placement of adult patients undergoing SHD interventions, deployed first. Before the center and the right guided by 3D TEE. and the growing number of adult interventional atrial disc can be deployed, however, the left cardiologists becoming adequately trained and disc must be aligned parallel to the 3D curved experienced, has shifted the paradigm in the plane of the defect. Device misalignment may direction of many adult interventionalists, not be recognized due to the limitations of 2D providing the leadership for the innovation, imaging and subsequent complete device design and execution of the next generation of deployment may result in failure of the right SHD interventions. The growing interest in SHD and left atrial discs to be properly positioned, 46 MEDICAMUNDI 52/2 2008 interventions, and therefore the need to train leading to the need to recapture the device and redeploy following appropriate device 3 repositioning. In situations such as this, the superior anatomic representation of the interatrial septum offered by 3D imaging may further enhance optimal device positioning, thereby avoiding the repeated need to recapture and redeploy. The second example involves the investigative Evalve clip used to repair regurgitant mitral valves by clipping together the center portions of the two valve leaflets [13]. The deployment of a mitral valve clip can only be attempted after the delivery catheter is coaxially aligned with the center of the mitral valve orifice so that the clip system can be advanced in a pathway that permits proper alignment of the clip with the two leaflets. Given the non-planar orientation of the mitral valve annulus, achieving coaxial alignment of the delivery catheter with the center of the mitral valve orifice is challenging when limited to 2D fluoroscopic or echocardiographic views. By simultaneously providing both anatomic and spatial representations of the mitral valve, 3D imaging should facilitate the manipulation and alignment of the delivery catheter to the G mitral valve orifice, thereby increasing the odds Figure 3. An image from the workstation is shown here, demonstrating the combining of of achieving procedural success. pre-procedure CTA data with live fluoroscopy during an ASD closure procedure. The delivery cable is seen attached to a partially deployed device in the real-time X-ray image. The CTA Currently, there are two methods of using 3D contains the soft tissue data that helps identify the location of the ASD, right atrium (yellow) image data to guide percutaneous interventions. and left atrium with pulmonary veins (blue-green). The CTA data are first segmented to isolate The first uses a 3D CTA- or MRA-derived these important structures, transferred into the workstation in the procedure room, registered image set obtained pre-procedurally [14]. The with the fluoroscopy, and then actively used during the SHD intervention. image is segmented to illustrate important anatomical features (i.e. location of pulmonary data in a variety of graphical display modalities. veins and valve annuli, relationships of major The Division of Cardiology at the University of E RT 3D TEE image data vascular structures to surrounding cardiac Colorado has used RT 3D TEE in over fifty of offer the ability to chambers, etc.). The segmented image is then the more difficult SHD interventions including navigate in the heart transferred to a workstation in the procedure complex ASD and PFO closure, mitral balloon using live 3D imaging. area, which allows the image to be displayed valvotomy, mitral repair using the Evalve clip and during the intervention. The image is then alcohol septal ablation of obstructive hypertrophic registered, scaled and localized in 3D space with cardiomyopathy (Figures 4 and 5) [1, 14, 15]. the X-ray system, using objects that are present The learning curve of applying this novel in both images (e.g. vertebral column, cardiac technology has been steep and has involved not borders and other internal landmarks). only gaining familiarity with the equipment, but Subsequently, when the X-ray system is rotated, also standardizing views, achieving effective the registered CTA or MRA image rotates to communication between the echocardiographer maintain the alignment of the two images. and interventionalist and dissecting every Thus, using this system, the operator has the intervention into a series of tasks, each requiring unique opportunity to use live-fluoroscopy with a unique visual guidance solution. The technical a background 3D image containing the soft benefits offered by RT 3D TEE guidance in tissue cardiac structures. This type of 3D SHD interventions, however, far outweigh the image-guided approach has been utilized in present logistical challenges faced in procedures such as aortic coarctation stenting, incorporating its use. The surface renderings VSD closure and ASD closure (Figure 3) [7, 14]. generated from RT 3D TEE image data truly represent a landmark in interventional cardiology The second major approach to 3D image as they offer the ability to navigate in the heart guidance is the recently available real-time (RT) using live 3D imaging. For example, when used 3D ultrasound imaging. In 2007 Philips to navigate a catheter across the mitral valve for Healthcare released the Live 3D TEE iE33 Echo either delivery of an Evalve clip or during System, which enables viewing of 3D image percutaneous balloon valvuloplasty, the catheter, MEDICAMUNDI 52/2 2008 47 E 4a 4b Figure 4. This four-panel figure shows images from the three dimensional transesophageal echocardiography (3D TEE)-guided placement of a closure device in a patient’s patent foramen ovale (PFO). Panel A shows the deployed left atrial disc and the delivery catheter. Panel B shows the position of the left atrial disc after the interventionalist has pulled back the delivery catheter and partially deployed the device to engage the 4c 4d left atrial side of the PFO. The bottom panels show the successfully deployed device by 3D TEE (Panel C) and X-ray (Panel D). Thin moving structures, such as the interatrial septum, are sometimes difficult to visualize completely with 3D TEE, so that it is common to switch from live 3D imaging to the simultaneous display of two planes of 2D ultrasound imaging, which is also provided for in the iE33 Echo System. Images courtesy of Ernesto Salcedo, MD.

E 5 Figure 5. One of the more challenging investigative SHD interventions is the placement of a clip that fastens the mid-portion of the anterior and posterior leaflets of the mitral valve together, in order to reduce the severity of mitral regurgitation. The clinical trial is comparing this form of repair of the valve to traditional open surgical repair. 3D TEE, as shown here, is making a major difference in facilitating the procedure by simultaneously showing the entire guiding catheter, clip delivery system, along with the mitral valve and surrounding structures in 3D during manipulation of the equipment. This allows easier alignment with the mitral valve orifice, which is challenging in 2D slice images [13]. The insert image is the X-ray image showing a deployed clip and a second clip attached to the delivery system. The white arrows point to the delivery catheter in each image. Image courtesy of Robert Quaife, MD.

48 MEDICAMUNDI 52/2 2008 6 F Figure 6. Advances in imaging are opening new vistas in patient- specific anatomical characterization. In this 3D TEE image of the left atrial appendage (LAA) the determination of the size and shape of the orifice is shown by the dotted green lines and the white arrows. This measurement is important in choosing the correct size of a new group of investigative devices that are inserted in order to occlude the LAA and thus prevent the thrombus formation that is common in patients with atrial fibrillation. cardiac tissue and mitral valve are displayed as visual-feedback loop where images are processed solid structures with no transparency thereby by the operator before additional equipment mimicking the process actually taking place manipulation is made. Thus, robotic surgery inside the patient’s heart (Figure 5). replaces direct visualization with image data obtained through fiber optic technology. For With these novel 3D imaging techniques SHD interventions, visualization will not be comes the need to address several outstanding accomplished via fiber optics, but instead with issues. First, the presentation and orientation of the use of medical images (such as X-ray and 3D image data that are most appropriate and ultrasound) for navigation and manipulation of useful to the interventionalist need to be devices within cardiac chambers. determined. Second, there is the need to develop a means to interactively correlate the delivery Finally, the advances in 3D imaging and SHD catheter and device deployment controls to the interventions allow a more in-depth approach 3D image being presented to the to anatomical characterization (Figure 6). New interventionalist. Many SHD delivery systems devices are also being designed in response to have controls that allow the system to change advances in the anatomical understanding of shape, turn and advance, but it is up to the SHD targets [16]. interventionalist to register the results of these manipulations in the 3D space of the patients’ Conclusions hearts. While RT 3D imaging offers the interventionalist realistic visual images, improving SHD interventions are growing in number and the process of orienting and registering the form the third major arena of interventional E CTA, MRA and RT 3D interventional equipment to the presented image cardiology after coronary and non-coronary TEE are increasing the data remains unchartered terrain. Forging a vascular interventions. Imaging modalities, both efficient performance marriage of imaging equipment and interventional primary and assisted, are evolving as rapidly as of complex SHD devices, where precision and predictability are the interventions being performed. Whereas interventions. possible, and thereby moving beyond the ICE has already significantly influenced the practice of “turning the catheter and seeing performance of simple SHD interventions, where it goes” could further revolutionize 3D integration of novel 3D imaging modalities, such image guidance in SHD interventions. as CTA, MRA and RT 3D TEE is drastically impacting the efficient performance of complex While the current manipulation of SHD SHD interventions. Several logistical equipment is completely manual, assisted or challenges, including the need to optimize and robotic navigation emerges as a logical standardize 3D views and registering device development, which could potentially overcome manipulation with presented image data, still the issues of registration and enhance the require investigation. Solutions to these issues precision and efficiency of catheter manipulation. in the form of advanced image processing, Robotic navigation in medicine has generally anatomy-based comprehensive analysis, involved the placement of mechanical controls multidimensional fusion and integrated between the operator and the image. Operator navigation systems could further revolutionize control is preserved through a constant SHD interventions K MEDICAMUNDI 52/2 2008 49 References

[1] Kim M, Klein A, Carroll JD. Transcatheter Closure of [10] Pediatric and Adult Interventional Cardiac Symposium Intracardiac Defects in Adults. J Interv Cardio. 2007; 20(6): PICS-AICS 08 524-545. http://www.picsymposium.com

[2] Garg P, Walton AS. The New World of Cardiac Interventions: [11] EuroPCR 2008 A Brief Review of the Recent Advances in Non-Coronary http://www.europcronline.com/ Percutaneous Interventions. Heart Lung Circ. 2008; 17(3): 186-199. [12] The Society for Cardiovascular Angiography and Interventions (SCAI) Annual Scientific Sessions in [3] Bonow RO, Carabello BA, Kanu C, de Leon AC Jr, Faxon Partnership with the American College of Cardiology DP, Freed MD et al. ACC/AHA 2006 Guidelines for the (ACC) i2 Summit SCAI-ACCi2 2008 Management of Patients with Valvular Heart Disease: A Report http://www.scai-acci2.org of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. [13] Silvestry FE, Rodriguez LL, Herrmann HC, Rohatgi S, 2006; 114(5): e84-231. Weiss SJ, Stewart WJ et al. Echocardiographic Guidance and Assessment of Percutaneous Repair for Mitral Regurgitation [4] Carroll JD. PFO Closure in 2008: See Through a Glass with the Evalve MitraClip: Lessons Learned from EVEREST Darkly. Catheter Cardiovasc Interv. 2008; 71(3): 388-389. I. J Am Soc Echocardiogr. 2007; 20(10): 1131-1140.

[5] Carroll JD. Migraine Intervention with STARFlex Technology [14] Garcia JA, Eng MH, Wink O, Chen SYJ, Hansgen A, Trial: A Controversial Trail of Migraine and Patent Foramen Carroll JD. Image Guidance of Percutaneous Coronary and Ovale Closure. Circulation. 2008; 117(11): 1358-1360. Structural Heart Disease Interventions Using a Computed Tomography and Fluoroscopic Integration. Vascular Disease [6] Carroll JD, Messenger JC. Medical Simulation: The New Management. 2007; 4(3): 89-97. Tool for Training and Skill Assessment. Perspect Biol Med. 2008; 51(1): 47-60. [15] Eng M, Salcedo E, Quaife R, Carroll JD. Real-Time 3-Dimensional Transesophageal Echocardiography for [7] Carroll JD. The Future of Image Guidance of Cardiac Percutaneous Structural Heart Disease Interventions. J Am Interventions. Catheter Cardiovasc Interv. 2007; 70(6): 783. Coll Cardiol. 2008; Submitted.

[8] Kim M, Hansgen AR, Wink O, Quaife RA, Carroll JD. [16] Carroll JD. PFO Anatomy: 3D Characterization and Device Rapid Prototyping: A New Tool in Understanding and Treating Performance. Catheter Cardiovasc Interv. 2008; 71(2): Structural Heart Disease. Circulation. 2008; 117(18): 229-230. 2388-2394.

[9] Transcatheter Cardiovascular Therapeutics TCT 2008 http://www.tctconference.com

50 MEDICAMUNDI 52/2 2008