40
How to Map and Ablate Unstable Ventricular Tachycardia: The University of Colorado and University of Pennsylvania Approach
Wendy S. Tzou, MD; Francis E. Marchlinski, MD
Introduction
Ventricular tachycardia most commonly occurs in the setting of structural heart disease, with patients with prior myocardial infarction and chronic coronary artery disease comprising the majority of cases.1 However, individuals with nonischemic heart disease, including nonischemic cardiomyopathy (NICM) and arrhythmogenic right ventricular cardiomyopathy (ARVC),2–4 can have large areas of scar that lead to ventricular tachycardia (VT). In all of these patients, the predominant mechanism for VT is reentry. The presence of surviving myocardium intermixed within the scar provides the appropriate milieu for reentry to initiate and propagate in areas with (1) fixed or functional unidirectional block, and (2) slow conduction that permits recovery of previously depolarized tissue.
Due to limitations in antiarrhythmic drugs or implantable cardioverter-defibrillators (ICDs),5–8 percutaneous catheter mapping and ablation has evolved as an alternative and effective treatment strategy for the management of these often challenging cases.9–11 Critical VT circuit elements can be identified and targeted successfully with ablation as long as the VT is inducible and hemodynamically tolerated. However, the vast majority of clinical VTs are “unmappable.” This chapter will delineate the approach utilized by the Electrophysiology Program at the Hospitals of the University of Colorado and Pennsylvania for ablation of unstable VT.
Preprocedural Planning
Surface ECG
We scrutinize available surface ECGs for clues about underlying disease substrate, scar burden, and, when possible, potential VT site(s) of origin. In sinus rhythm, the presence of Q waves in contiguous leads or ST elevations in absence of acute ischemia suggests the presence of prior myocardial infarction or aneurysm, and helps identify locations of potential VT substrates (described in detail in another chapter). The presence of a widened QRS at baseline, which may be a marker of His-Purkinje disease, suggests a predisposition for bundle-branch reentry or fascicular VT. A pronounced R wave in V1 along with a pronounced S wave in V6 in the absence of marked conduction disturbances suggests presence of basolateral scar in a patient with NICM.12 Discovering epsilon waves or T-wave inversions in the right-sided precordial leads should heighten suspicion for ARVC, particularly in a younger person with multiple VT morphologies. In both of the latter situations, one should be prepared for possible epicardial in addition to endocardial ablation.4,13 Further clues from an ECG in VT in a patient with NICM that may suggest the need for an epicardial approach include Q waves in lead I, along with an absence of Q waves in the inferior leads, a pseudo–delta wave of ≥ 75 ms, and a maximum-deflection index of ≥ 0.59, all of which are highly sensitive and specific for a basal and superior lateral epicardial focus.14 Finally, an ECG acquired during VT not only assists in preprocedural planning but aids in the procedure itself (described below).
Noninvasive Cardiac Imaging
We obtain noninvasive cardiac imaging in all patients planning to undergo VT ablation. This is particularly helpful when no ECGs of VT exist. Information verifying extent of underlying disease, ventricular function and dimensions, scar or aneurysm location, and any significant valvular disease is important in procedural planning. Echocardiography has traditionally been the mainstay for acquiring this information; nuclear scintigraphy is often additionally helpful in verifying substrate information and to assess for active ischemia. We have been increasingly performing cardiac MRI in patients presenting with VT, as it is a more sensitive tool for assessing for nontransmural scar in NICM.15,16 Presence of an ICD or pacemaker has long been a contraindication for performing an MRI, but our center and several other centers have shown that it can be done safely in those who are not pacemaker dependent.17,18 There are, however, limitations in image quality and interpretation due to artifact from the ICD pulse generator and right ventricular lead. The optimal approach, when possible, is to obtain the MRI prior to device implantation.
ICD Electrogram
In patients with ICDs, stored information during the VT episodes should be carefully reviewed. The number of events, both nonsustained and those resulting in therapy, can guide urgency with which subsequent medical care is executed. Additionally, device-recorded intracardiac electrograms (EGMs) may be especially helpful for determining (1) if there is potential utility of catheter ablative therapy, and (2) the target(s) for ablation if a surface 12-lead ECG of the clinical VT is not available. Events predominated by polymorphic VT, for instance, may be less amenable to ablation unless a reproducible premature ventricular contraction (PVC) trigger is evident that may be targeted. Depending on patient stability, one could consider noninvasive programmed stimulation (NIPS) testing through the device to see if any VT is induced whose intracardiac EGM resembles those stored from spontaneous clinical events in order to identify on 12-lead ECG the VT that has occurred clinically.
Additional Preparation
Before pursuing more invasive evaluation and treatment, a recent evaluation of the presence and status of coronary artery disease and determination of left ventricle (LV) function and reserve, either noninvasively, as described above, or via cardiac catheterization, is important. Deaths related to the procedure often occur due to underestimation of the importance of ischemic burden prior to and volume overload during the procedure. Documentation of anticoagulation status and exclusion of an unstable LV thrombus with echocardiography also are important. Identification of vascular access problems prior to the procedure facilitates its success and safety. Presence of severe arterial or aortic valvular disease, including mechanical prosthesis, usually mandates a transseptal approach to the endocardial aspect of the LV. Prior cardiac surgery and prior pericarditis may limit or preclude epicardial access and mapping/ablation. Details of present and past antiarrhythmic therapy are also important. Ideally, discontinuation of antiarrhythmic therapy for at least 5 half-lives should occur before the procedure, although this is often not possible because of the unstable nature of our patients. Furthermore, the electrophysiological effects of amiodarone often cannot be reversed due to its long half-life, even if the drug were stopped several days before the procedure. The results of programmed stimulation and subsequent mapping maneuvers should be interpreted with this knowledge.
Procedure
Patient Preparation
We initiate most VT ablations using conscious sedation or monitored anesthesia care (MAC) due to occasional difficulty in arrhythmia induction and the hypotension that can result from deeper levels of sedation. Occasionally, intubation and general anesthesia are necessary at the outset due to patient discomfort or ongoing or anticipated instability. In the setting of marginal hemodynamic stability and/or severe disease at baseline, additional mechanical support may also be initiated at the outset if not already present. This can be accomplished by intra-aortic balloon pump (IABP) insertion, as long as there is no severe peripheral arterial disease or aortic insufficiency. At our institution, this procedure has been increasingly performed in our electrophysiology laboratory. VT ablation can be performed safely in patients with surgically placed ventricular assist devices. Placement of a percutaneous ventricular assist device (e.g., Impella, Abiomed Inc., Danvers, MA, or TandemHeart PTVA, CardiacAssist, Inc., Pittsburgh, PA) to provide temporary support during ablation is also an option. We have also found utility in use of peripheral veno-arterial extracorporeal membrane oxygenation (ECMO), placed either at the time of or before VT ablation. Advantages of ECMO include ability to provide full cardiac output support without limiting access options or mapping within the LV, as well as possibly reducing the risk of bleeding. The major limitations of ECMO include the requirement for cardiothoracic surgical assistance for placement and removal, availability of a perfusionist to maintain the circuit, and having a contingency plan in place in case the ECMO cannot be easily weaned following ablation.
Foley urinary catheters are placed in all patients, given the long procedure lengths and the significant volume administered with irrigated-tip ablation, necessitating an ability to accurately track urine output. The latter is particularly important in management of patients with severe ventricular dysfunction. Two sets of defibrillation pads, anterior-posterior and lateral, are often placed to prepare for shock-refractory VT/ VF. The chest and upper abdomen are sterilely prepped if epicardial mapping and ablation are anticipated. If present, the patient’s ICD is reprogrammed to disable tachycardia detection and therapies, after placement of ECG electrodes and external defibrillation pads. Baseline pacing mode is usually changed to maximize intrinsic ventricular conduction. However, presence of significant bradycardia, high-grade AV block, or ventricular ectopy may make continuous ventricular pacing preferable, including biventricular pacing when possible. Importantly, the ICD programmer remains in communication with the patient throughout the case, so that instantaneous intracardiac EGMs can be recorded during induced or spontaneously occurring VTs.
Vascular Access and Catheter Placement
The standard catheters used for ablation of LV VT include a quadripolar catheter for placement in the apical RV, an intracardiac echocardiography (ICE) catheter, and an ablation catheter. Occasionally, simultaneous recordings of His, right atrium (RA), and/or coronary sinus signals are desired, in which case vascular access to accommodate additional quadripolar or decapolar catheters is obtained. We generally place sheaths that are at least 1-Fr caliber larger than the catheter that will be inserted within them to allow for continuous fluid and/or drug infusions via the sheath side ports. In general, each femoral vein can accommodate in the range of up to 2 11.5-Fr or 2 7-Fr and 1 11-Fr sheaths.
Depending on whether retrograde aortic or transseptal approach to the LV is planned, vascular access is obtained either via the right femoral artery or the right femoral vein, respectively. A short 8- or 9-Fr sheath is initially placed and exchanged later for a longer (transseptal or braided) sheath, if indicated. In the standard LV VT ablation, we also insert 6- or 7-Fr and 9-Fr sheaths in the left femoral vein for the RV quadripolar and standard ICE catheters, and up to 1 additional (e.g., a 7-Fr catheter for a dynamic decapolar catheter) if necessary. Another 4- or 5-Fr sheath is often placed in the left femoral artery in cases where IABP insertion is not performed at the outset, to facilitate urgent insertion during the case if needed. In the case of LV transseptal access, this also functions as a means to continuously monitor arterial blood pressure during the case. If RV mapping or ablation is anticipated, an 8-Fr sheath is placed in the right femoral vein, which can later be exchanged for a longer sheath if necessary. Occasionally, either initial or continuous pulmonary artery pressure recordings are obtained using a balloon-tipped catheter, to assist in hemodynamic monitoring.
Once sheaths are in place, heparin is initiated (bolus of 100 to 120 U/kg and infusion of 12–15 U/kg/min). The ACT is checked every 15 to 30 minutes during the procedure, with additional heparin bolus and drip titrated as needed to maintain an ACT of 250 to 300 seconds (350–400 seconds for transseptal cases).
ICE imaging is performed at baseline to assess for anatomical features (e.g., valvular disease, wall motion abnormalities, LV size and overall function, and presence of pericardial effusion) and throughout the case to monitor for acute complications, as well as to assist in mapping and ablation. ICE is particularly helpful in confirming adequate catheter contact when mapping or ablating in the proximity of the papillary muscles (Figure 40.1). Finally, ICE is also a valuable tool in assisting with transseptal puncture when required and can be used to actually track lesion formation.
Retrograde Aortic Versus Transseptal LV Access
The retrograde aortic approach is the preferred method for LV access, even when an IABP or peripheral ECMO circuit is present. Exceptions to this include presence of severe peripheral arterial, aortic, and/or severe aortic valvular disease. Also, in some cases, ablation of VT suspected to originate from the LV septum might be performed with greater contact and catheter stability using the transseptal method. In patients with coronary artery disease, known mild to moderate peripheral arterial disease, or elderly patients with tortuous vessels, we frequently exchange the short sheath placed in the right femoral artery for a 45- to 65-cm-long, 8- or 9-Fr braided or nonbraided sheath (Arrow, Teleflex Medical, Research Triangle Park, NC), over a long, stiff, J-tipped wire (e.g., Amplatz, Cook Medical, Bloomington, IN) under fluoroscopic guidance. Care is made to prevent the sheath or wire from crossing the aortic valve, to minimize the amount of hardware that can cross and potentially damage valve leaflets. This is usually only a consideration when using the 65-cm-long sheath. In the absence of these characteristics, mapping and ablation through the initially placed short sheath is preferable.
Transseptal puncture is performed under ICE and fluoroscopic guidance once heparin has been initiated, as above. For VT ablation, we often use a large-curve Agilis or LAMP90 Transseptal sheath (St. Jude Medical, St. Paul, MN) with a transseptal needle (BRK Transseptal, St. Jude Medical or NRG RF Transseptal Needle, Baylis Medical, Burlington, MA). LA access is obtained with manual pressure once the sheath and needle assembly are seen to tent the interatrial septum with the left pulmonary veins and LA appendage in view. A somewhat more anterior approach than is used for pulmonary vein isolation is preferred, to better direct the sheath toward the LV. Pressure transduced through the needle tip is additionally observed throughout this maneuver to confirm successful LA access.
We typically use a THERMOCOOL (standard, SMARTTOUCH, or SF) 3.5-mm-tip catheter (Biosense Webster, Diamond Bar, CA) for VT ablation. A standard F-curve radius is usually sufficient. However, if an LV outflow tract or cusp VT is suspected, a D curve may be more effective and allow for greater maneuverability in more confined spaces. Conversely, for maneuvering within a severely dilated ventricle, a J curve may provide greater reach. Bidirectional catheters tend to be used more often, as they often allow for greater maneuverability, especially when choosing a catheter with differing radius per direction (e.g., F-J curve).
VT Induction
Particularly when no 12-lead ECGs of the clinical VT exist, we routinely attempt to induce VT. Depending on the clinical situation, this may be performed before, during, or after electroanatomic mapping is completed. If epicardial mapping is anticipated, we typically perform programmed stimulation after acquisition of the endocardial map. Similar to the goals of NIPS testing described above, we aim to see if there is a VT induced that matches the ICD EGM VT morphology. This may help to define our “clinical” or principally targeted VTs and guide our mapping and ablation strategy. In addition, limited activation and entrainment mapping can sometimes be performed by positioning the mapping catheter at a suspected site of importance based on analysis of the 12-lead ECG of VT and regions of abnormal bipolar voltage just before VT induction. During VT and in the short interval before hemodynamic instability necessitates arrhythmia termination, mid-diastolic potentials at that or immediately adjacent sites often can be quickly characterized.
Methods for programmed electrical stimulation (PES) vary significantly. Pacing for 8 beats with basic drive cycle lengths of 600 and 400 ms followed by the introduction of 1 to 3 extrastimuli is generally the most accepted protocol. On occasion a short-long-short stimulation sequence may be advantageous. This stimulation has been described for the induction of bundle branch reentry but can be used for the induction of other VT associated with structural heart disease that is not induced with a more standard stimulation protocol. Burst pacing or sympathetic stimulation, for instance with isoproterenol infusion or stress testing, is often more useful for inducing arrhythmias based on triggered activity. Specific VT morphologies may require unique VT sites of stimulation. We have found that VTs with a right bundle branch block morphology may frequently require stimulation from the lateral LV.
Endocardial Mapping
A detailed, endocardial, electroanatomic voltage map is created, typically using CARTO (Biosense Webster), with efforts initially focused on areas of known infarction or scar based on prior imaging. Using well-characterized bipolar voltage criteria of > 1.5 mV for identifying normal signal amplitude recorded from the LV endocardium,19 we identify regions of bipolar voltage consistent with densely scarred myocardium or aneurysm (< 0.5 mV), border zone of scar (0.5 to ≤ 1.5 mV), and healthy myocardium (> 1.5 mV). Similar criteria are utilized for bipolar RV mapping. The epicardial bipolar voltage cutoff for defining abnormal is kept at 1.0 mV.13 The bipolar voltage criteria need to be adjusted upward if patchy scar is anticipated (reperfused infarct, NICM) and more attention is paid to EGM characteristics and abnormalities (see below). Importantly, we found that unipolar voltage can also be helpful in revealing areas of deeper (midmyocardial or epicardial) substrate abnormalities (< 8.3 mV in the LV and < 5.5 mV in the RV).20,21 After the bipolar voltage map is constructed, we then also examine the unipolar voltage, particularly when there is a disparately normal amount of endocardial bipolar voltage compared to the LV ejection fraction, or when the suspected VT exit or earliest site of activation occurs in an area of preserved bipolar voltage (Figure 40.2).