How to Utilize ICE for Optimal Safety and Efficacy with Atrial Fibrillation Ablation
Mathew D. Hutchinson, MD
The inherent complexity of catheter-based ablation for AF necessitates multimodality image integration. ICE provides dynamic, real-time images that are uniquely suited to complex ablation procedures. In fact, ICE satisfies all of the requirements of image-guided AF ablation: the delineation of pulmonary venous anatomy and its variants, the facilitation of intracardiac catheter positioning relative to potential ablation targets, the real-time characterization of catheter-tissue contact, the assessment of intracardiac lesion formation, and the early detection of procedural complications.
There are two commercially available types of ICE transducers: radial and phased-array systems. The radial ICE transducer is mounted on a 9-Fr, nonsteerable catheter and emits an imaging beam at a 15° forward angle, perpendicular to the long axis of the catheter. The transducer rotates at 1800 rpm, thus producing a 360° imaging plane. The fixed, 9-MHz transducer frequency provides excellent near-field resolution; however, far-field structures are poorly visualized with radial ICE, necessitating imaging proximate to the structure of interest (Figure 13.1).
The phased-array ICE transducer contains 64 elements with frequencies ranging from 5 to 10 MHz. This provides more flexibility in imaging either adjacent or distant structures with a potential imaging depth of up to 15 cm (Figure 13.1). The transducer is mounted on an 8- or 10-Fr catheter that can be deflected in four directions (anterior, posterior, right, and left) in addition to 360° axial rotation. Its steerability and low profile allow the imaging transducer to be navigated throughout any cardiac chamber of interest. The phased-array transducer is also capable of full spectral and color Doppler measurements, greatly enhancing the physiologic data achievable. Unless otherwise indicated, the remainder of this article will describe imaging with the phased-array ICE system.
PVI procedures are often performed with preacquired tomographic reconstructions of the LA. Image integration has been shown to facilitate ablation in areas with complex 3-dimensional (3D) structure, such as the ridge separating the LAA and LPVs, and is particularly useful in patients who have pulmonary venous anatomical variants.1 Unfortunately, CT imaging incurs the additional risks of radiation and iodinated contrast exposure, which may render it unsuitable in selected patients. Likewise, gated MRI quality is often limited by the rapid and irregular ventricular rates encountered in many preablation patients.
Like with tomographic imaging, complex anatomical relationships can also be clearly defined with phased-array ICE imaging (Figure 13.2). In a similar manner to tomographic reconstructions, 3D ICE contours can also be precisely integrated with electroanatomic mapping systems (Figure 13.3). These reconstructions have been shown to correlate well when coregistered with tomographic LA reconstructions.2
It is our practice to obtain TEEs before AF ablation in patients with either nonparoxysmal forms of AF or inadequate preoperative anticoagulation. Some patients have either a contraindication to TEE or equivocal findings on serial examinations, which are misinterpreted as thrombus. In these cases, a secondary imaging study is required to adjudicate the presence of thrombus. Imaging of the LAA with ICE is accomplished by placing the imaging transducer into the RVOT, CS, or PA (Figure 13.4). The RVOT gives excellent visualization of the LAA ostium and proximal segment; however, additional views from the PA are often required to survey the entire structure. A recent study examined the diagnostic quality of ICE in patients undergoing atrial ablation procedures.3 The authors performed contemporaneous TEE and ICE imaging in 71 patients, and found excellent morphologic correlation of both the RAA and LAA between the two modalities. The also reported a higher sensitivity with ICE compared to TEE to detect small atrial thrombi and spontaneous echo contrast.
There remains a paucity of data supporting the use of ICE in patients undergoing left atrial appendage closure procedures. The potential advantages of using ICE in these cases include obviating the need for TEE with its associated risks and better integration with EP workflows that utilize ICE imaging in cases involving left atrial access. Further investigation is required to determine whether ICE can deliver equivalent TEE imaging planes, and the optimal technique to obtain them.
The phased-array ICE catheter is advanced through a standard vascular access sheath. We routinely upsize the access sheath one French size to allow unencumbered intravenous infusion through the sheath while the ICE catheter is in place. In patients with limited femoral venous access, we often place static diagnostic catheters (e.g., CS) via the internal jugular vein, thus allowing the ICE catheter to be inserted via the femoral vein. This permits the operator to easily manipulate the echo imaging planes throughout the procedure.
Passing the imaging catheter from the femoral vein to the heart can be easily performed without fluoroscopy in the majority of patients. The catheter is rotated axially and/or deflected in order to maintain the tip of the imaging transducer within the long axis of the vein (Figure 13.5; Video 13.1). The cardinal rule in safely maneuvering the phased-array catheter without fluoroscopic guidance is to always maintain an echocardiographic clear space between the transducer tip and the wall of the structure being imaged.
Imaging with ICE requires a modest learning curve but is a logical extension for operators with basic catheter manipulation and echocardiography skills. The phased-array catheter has eight degrees of freedom: (1) deflection: anterior, posterior, left, and right; (2) axial rotation: clockwise and counterclockwise; and (3) translational movement: advancement or withdrawal. Given that the imaging planes obtained with ICE vary relative to the position of the catheter, there are infinite potential imaging planes. As a result, any single 2-dimensional (2D) view taken out of anatomic context can be disorienting to the operator. If, however, the catheter is manipulated from a fiducial imaging plane, then the resulting anatomic relationships are intuitive.
Baseline ICE Imaging in AF Ablation
The phased-array catheter is initially placed in the mid-RA. The “home” view in the preablation survey, a long-axis view of the right ventricle and tricuspid valve, is easily obtained without any catheter deflection (Figure 13.6; Video 13.2). The home view allows the assessment of tricuspid valve structure and function, as well as the estimation of pulmonary arterial pressure with continuous-wave Doppler. Most of the subsequent imaging planes can be obtained by careful clockwise rotation of the imaging catheter from the home view. Whenever an unfamiliar imaging plane is encountered at any point during the study, we return the catheter to the home view by removing all catheter deflection and gently rotating the imaging catheter in a clockwise direction until the tricuspid valve is visualized.
With clockwise rotation from the home view, the imaging plane is directed posteriorly and leftward. At 45° CW rotation, a long-axis view of the aortic root is seen; the assessment of aortic valve structure and function, as well as the presence of aortic atheroma, is possible from this view (Figure 13.6). Continued clockwise rotation of the ICE catheter (90° from the home view) brings the mitral annulus and LAA into view (Figure 13.6; Video 13.3). Since the lateral mitral annulus and LAA are obligately present in the echo far-field, it may be necessary to decrease the imaging frequency to improve imaging resolution of these distal structures. As previously mentioned, the LAA is better visualized from more proximate structures such as the LPA or the distal CS.
Further clockwise rotation from the mitral valve plane brings the LPVs into view ( Video 13.4). This view is extremely useful to characterize the size and morphology of the LPVs. The LPVs often share a common ostium; however, the carina between the veins may be quite prominent and require ablation to achieve PVI. The dimensions of the individual LPVs and the common ostium are recorded from this view (Figure 13.7). PV color and pulse-wave Doppler flow velocities are also recorded in the baseline state (Figure 13.7; Video 13.5). The esophagus may lie in closer proximity to the posterior aspect of the LPV or the RPV and can be imaged longitudinally with further clockwise rotation of the ICE catheter (Figure 13.8). At this point, the imaging transducer has been rotated 180° from the original home view.
With slight additional clockwise rotation along the posterior wall, the RIPV is visualized. The RPVs often have separate ostia and the RSPV characteristically originates near the IAS and courses rightward. As a result, the RSPV and RIPV are uncommonly seen in the same 2D imaging plane (Figure 13.9; Video 13.6). The RSPV is usually the most difficult to visualize, since the imaging plane is often directed through the thick, superior limbus of the IAS. This anatomical constraint is overcome either by deflecting the imaging catheter posteriorly toward the tricuspid annulus or by wedging the transducer tip under the superior limbus of the fossa ovalis. When using the latter technique, the imaging catheter will often migrate into the LA via a PFO; this is recognized by a lack of atrial septal tissue interposed between the catheter tip and the LA (Video 13.6). Direct LA imaging provides spectacular visualization of the RPVs. Careful assessment of the RPVs will often identify a separate middle PV ostium, allowing the operator both to incorporate this vein in an antral ablation strategy and to avoid inadvertently ablating inside it (Figure 13.10; Video 13.7).