Ablation of Ventricular Tachycardia With Percutaneous Hemodynamic Support

Options for Intraprocedural Cardiac Mechanical Support

Mechanical hemodynamic support during VT ablation is designed to maintain cardiac output and mean arterial pressure in the setting of suboptimal contractile properties present during VT, while promoting diuresis, preventing significant increases in pulmonary artery pressures, reducing the incidence of acute heart failure and multisystem organ failure, and perhaps improving safety and permitting more rapid recovery following the procedure. During periods of induced VT, the goals of strategies for hemodynamic support are to maintain acceptable systemic perfusion allowing for prolonged arrhythmia mapping and preventing rapid hemodynamic deterioration. The percutaneous support devices that have been used during VT ablation are intraaortic balloon pump (IABP) counterpulsation, the TandemHeart left atrial-to-iliac artery bypass (CardiacAssist Inc, Pittsburgh, PA), and percutaneous left ventricular assist devices (pLVADs), which are impeller-driven axial flow pumps placed temporarily through the aortic valve to pump blood directly from the left ventricle to the ascending aorta (Impella; Abiomed, Danvers, MA).

Most clinical experience with temporary mechanical support has been with IABPs, which are routinely implanted percutaneously by cardiac interventionalists. As the IABP augments diastolic pressure and diminishes afterload, its greatest benefit is in support of coronary blood flow in patients with active myocardial ischemia. However, IABPs are only able to augment cardiac output by 0.5 L per minute, and the increases in mean arterial pressure and stroke volume afforded by the IABP may not be sufficient to meet the hemodynamic demands of patients in ongoing VT. Furthermore, the IABP is dependent on timing of balloon inflation and deflation to pressure- or electrocardiogram-based triggers, and thus its optimal function requires a stable, regular, and nontachycardic rhythm, and thus is not ideally suited for patients undergoing VT ablation. Although the IABP is likely the most common percutaneous support device used during VT ablation, there is limited published experience demonstrating its safety or effectiveness. The benefits of the IABP are its (1) relatively small arterial sheath size (7 F); (2) ease of insertion; and (3) familiarity to most operators and laboratory personnel.

The TandemHeart device is a percutaneous left atrial to iliac artery bypass system, which uses an external centrifugal pump that provides up to 3.5 to 4 L per minute of forward flow. It received approval from the Food and Drug Administration in the United States for extracorporeal circulatory support of procedures not requiring full cardiopulmonary bypass for up to 6 hours, and longer durations of use have been tolerated in clinical trials. As hemodynamic benefit tracks directly with the degree of continuous flow provided, the TandemHeart appears to achieve greater hemodynamic support than that provided by the smaller Impella 2.5 pLVAD. However, the TandemHeart requires a relatively complicated insertion technique including both arterial and venous access at the level of the femoral vessels. To access the left atrium, venous access is obtained followed by transseptal puncture and dilation to accommodate the 21 F inflow cannula in the left atrium, and a separate arterial access is necessary to place the outflow cannula in the iliac artery. Potential complications of this system are cardiac tamponade, major bleeding, critical limb ischemia, sepsis, arrhythmias, and residual atrial septal defects. Bleeding complications in particular range from 40% to 90%.

The Impella pLVADs use a miniaturized axial flow pump to deliver blood directly from the left ventricle to the atrium. An advantage of this device is its relatively simple implantation technique, which requires a single femoral arterial access and retrograde placement across the aortic valve. Three Impella devices with different pump flow capabilities currently exist: the Impella 2.5, placed through a 13 F introducer sheath in the femoral artery and allowing a maximal flow rate of 2.5 L per minute; the more recently introduced Impella CP, capable of delivering approximately 3.5 L per minute of flow placed through a 14 F access sheath; and the Impella 5.0, which is capable of providing a maximal support of approximately 5.0 L per minute, although this latter device has a larger maximal catheter diameter (21 F) that requires a surgical cutdown for arterial access. Most clinical experience thus far with pLVAD-assisted VT ablation has been with the Impella 2.5 device. Table 37.1 includes selected case series and studies of percutaneous hemodynamic support during Sc-VT ablation.

Compared with the TandemHeart, the percutaneously implanted pLVADs have smaller catheter diameters and avoid additional venous access and transseptal puncture, circumventing related complications and permitting shorter implantation times. As the Impella device directly unloads the left ventricle, it efficiently reduces myocardial oxygen demand and consumption, augments mean arterial pressure, and decreases left ventricular end-diastolic pressure, likely to a greater degree than the TandemHeart at comparable flow rates, particularly in conditions of low cardiac output. Most clinical experience with the Impella thus far has been in patients undergoing high-risk percutaneous coronary intervention with or without cardiogenic shock, in which cases the Impella 2.5 has provided greater augmentation of cardiac index and mean arterial pressure, and greater improvements in left ventricular ejection fraction (LVEF), than in procedures supported by IABP. Contraindications to Impella placement are any mechanical aortic valve, aortic valve stenosis/calcification (with aortic valve area ≤1.5 cm ² ), moderate or greater degrees of aortic insufficiency, and significant vascular disease precluding percutaneous implantation, such as aneurysms and extreme tortuosity or calcifications. Vascular access complications are less common with the smaller Impella 2.5 device than with the larger Impella 5.0. For interventional procedures in which patients are typically hemodynamically stable at the start of the procedure, the Impella 2.5 may be particularly attractive given its ease of use, small profile, and minimal access site trauma, coupled with its superior ability to improve systemic perfusion and relieve elevated venous pressure. By contrast, in patients with severe heart failure and for those in cardiogenic shock, the greater degree of hemodynamic support provided by the Impella 5.0 or the TandemHeart may be required despite associated higher complication rates and longer times to implantation. Table 37.2 is a comparison of commonly used percutaneous hemodynamic support devices.

Hemodynamic Support During Ventricular Tachycardia Ablation

For catheter ablation of Sc-VT, preferred ablation strategies generally combine substrate modification during stable sinus rhythm with complete mapping of the VT circuit. Activation and entrainment mapping require VT induction and maintenance to identify and target critical sites for ablation accurately. However, most patients with Sc-VT have significant left ventricular dysfunction and other associated comorbidities that contribute to the development of hemodynamic instability during prolonged episodes of VT. Therefore in the absence of adequate intraprocedural hemodynamic support, extended periods of activation and entrainment mapping are often difficult, if not impossible, to achieve.

Although IABP involves relatively simple percutaneous implantation, its ability to provide hemodynamic support during VT is limited by the associated high heart rates and need for accurate synchronization with the cardiac cycle. Therefore enthusiasm has grown for the use of pLVADs for temporary cardiac support during ablation of unstable VT, but experience with this technique has been limited up until recently. In a small single-center retrospective analysis of VT ablations performed in patients with structural heart disease and hemodynamically unstable VT, ablations supported by a pLVAD (Impella 2.5) allowed a nearly 2.5-fold increase in the total duration of VT maintained per procedure (66.7 vs. 27.5 min, P = .03) with fewer early arrhythmia terminations for hemodynamic instability during entrainment/activation mapping (0.9 vs. 3.9 terminations/procedure, P ≤ .001) compared with ablations using IABP or no mechanical support. By increasing the time available for extensive mapping in VT, 9 of 10 ablations supported by pLVAD had at least one VT termination by energy delivery during ongoing VT, whereas this acute outcome was possible in only 5 of 13 (39%) of patients ablated without pLVAD support. Regarding relative safety, this study found no significant differences in evidence of tissue hypoperfusion or end-organ damage despite the greater total duration of VT in the pLVAD group.

In a multicenter, observational study including 66 patients undergoing VT ablation, IABP ( n = 22) was compared with pLVAD (Impella [ n = 25]/TandemHeart [ n = 19]). There was no significant difference in baseline comorbid conditions, including number of prior VT ablations, failed antiarrhythmic medications, and implantable cardioverter-defibrillator (ICD) therapies between both groups. Activation and entrainment mapping was attempted in all patients in whom VT was hemodynamically tolerated. VT was prematurely terminated if there was a decrease in mean arterial blood pressure to lower than 45 mm Hg. Pace mapping and substrate ablation was performed in patients who could not tolerate VT despite maximal circulatory support. In the pLVAD group, higher percentage of patients underwent activation and entrainment mapping of unstable VTs (82% vs. 59%, P = .04), more unstable VTs were mapped and ablated per patient (1.05 ± 0.78 vs. 0.32 ± 0.48; P < .001), more VTs were terminated during ablation (1.59 ± 1.0 vs. 0.91 ± 0.81; P = .007) with fewer rescue shocks (1.9 ± 2.2 vs. 3.0 ± 1.5; P = .049) when compared with patients who received circulatory support with IABP. There was also no significant difference in total fluoroscopy and procedure time between both groups. Despite having numerically higher complications in the pLVAD group that did not reach statistical significance (32% vs. 14%, P = .14), there was no significant difference in acute procedural success between both groups. Fig. 37.2 is an example of a patient with NICM and limited substrate-based ablation targets who benefited from the extended duration of entrainment mapping afforded by a pLVAD.

A prospective clinical study in 20 consecutive patients undergoing ablation for unstable VT using pLVAD (Impella 2.5) support (PERMIT1) performed to assess the safety and short-term efficacy of this strategy systematically. In this study, entrainment mapping/ablation was used as the first-line approach in all patients inducible for sustained monomorphic VT, with additional substrate modification performed at the operator’s discretion. Sustained VT was tolerated for nearly 1 hour of mapping and ablation per procedure in this study, permitting VT termination during ablation in two-thirds of patients; at 1 month follow-up, 80% of patients were free of clinical recurrence (including appropriate ICD therapy). Importantly, despite extended periods of mapping during VT, there were no instances of advanced kidney injury or cognitive dysfunction/deficit. This study was also the first to evaluate the use of a noninvasive neuromonitoring modality to assess the hemodynamic effects of VT in this setting.

Subsequent retrospective comparisons from other centers have had similar findings. One such study compared 13 consecutive pLVAD-assisted VT ablations with purely substrate-based mapping and ablation in 18 matched patients with unstable VT; this study, which used mean arterial blood pressure of 60 mm Hg or higher to determine the adequacy of hemodynamic support provided by the pLVAD, identified longer procedural times with more VTs ablated in the pLVAD group, but no associated differences in acute procedural success or event-free survival on follow-up.

In a retrospective, nonrandomized study including 68 patients who underwent Sc-VT ablation with hemodynamic support, pLVAD was used in 34 patients. Hemodynamically unstable VT was defined as a decrease in mean arterial blood pressure of lower than 50 mm Hg. There was no significant difference in baseline characteristics including age, type of cardiomyopathy, LVEF, New York Heart Association (NYHA) class, antiarrhythmic medications, and electrical storm between both groups. Detailed entrainment and partial activation mapping was performed in the pLVAD group but not in the control group because of hemodynamic instability. Pace mapping and substrate modification was performed in both groups. Patients in the pLVAD group had longer times in sustained VT (27.4 ± 18.7 vs. 5.3 ± 3.6, P < .01), and duration of VTs of 20 seconds or more was maintained in 44% versus 3% in the control group. In addition, a greater number of VTs were terminated with ablation (1.2 ± 0.9 vs. 0.4 ± 0.6, P < .01) in the pLVAD patients compared with control. Despite the greater number of VTs terminated in the pLVAD group, there was no significant difference in acute procedural success between the groups. Procedural success was significantly higher in patients with ischemic cardiomyopathy than those with NICM (89% vs. 53%, P = .001) likely because of difficulty in adequately dechanneling the VT circuit with substrate-based ablation. Complications were minimal and occurred in two patients receiving pLVAD and none in the control group.

We recently reported our experience in 194 consecutive ablation procedures for unstable VT of which 109 received hemodynamic support with pLVAD and 85 no-pLVAD. Patients who received pLVAD were sicker with greater proportion of dilated cardiomyopathy (33% vs. 13%, P = .001), NYHA class III or higher (51% vs. 25%, P < .001), electrical storm (49% vs. 34%, P = .04), amiodarone use (72% vs. 53%, P = .01), and lower LVEF (26 ± 10% vs. 39 ± 16%; P < .001). Entrainment and activation mapping-guided ablation was the first-line approach and was attempted in all patients if VT was hemodynamically tolerated. In addition, pace mapping with substrate-based ablation was performed at the discretion of the operator when VT could not be tolerated. Unstable VT was defined as a decrease in mean arterial pressure of 50 mm Hg or less on maximum circulatory support including vasopressors and inotropes. pLVAD group had a larger number of VTs induced (3.3 ± 2.1 vs. 2.4 ± 2.1, P = .004) and had higher proportion of patients with at least one hemodynamically mappable VT (80% vs. 68%; P = .06), but there was no significant difference in VTs terminated with ablation (1.4 ± 0.8 vs. 1.4 ± 0.7, P = .96) between groups. Ablation time was similar between groups, but total procedure time was longer in the pLVAD group with no significant difference in postablation VT inducibility or postprocedure length of hospitalization between groups. There was also no significant difference in complications between groups (17% vs. 9%, P = .15). In addition, we performed a propensity score-match analysis to account for underlying bias created by any confounding factors. The analysis demonstrated no difference in acute procedural success, complications, or postprocedural length of hospitalization between groups.

In the large multicenter study from the International Ventricular Tachycardia Collaborative Consortium (IVTCC) including 1655 consecutive patients who underwent VT ablation, 105 patients received hemodynamic support with a pLVAD (Impella 2.5, extracorporeal membrane oxygenation and TandemHeart). Similar to the prior published experience, patients who received pLVAD were sicker with multiple comorbid conditions including lower LVEF, higher VT storm, ICD shocks, higher NYHA class, and greater use of amiodarone than compared with those with no-pLVAD. Patients who received pLVAD had a greater number of VTs induced (2.4 ± 1.4 vs. 1.9 ± 1.6, P = .001), fewer percentage of patients with unmappable VT (16% vs. 34%, P < .001), and received more ablation (2730 ± 2214 vs. 2104 ± 1627, P = .01) when compared with no-pLVAD patients. However, acute procedural success was lower (71.8 % vs. 73.7%, P = .04), and complications (12.5% vs. 6.5%, P = .03) and in-hospital mortality (21% vs. 2.3%, P < .001) were significantly higher in patients who received pLVAD compared with controls. In a subgroup analysis in patients with LVEF 20% or less and NYHA class III-IV, unmappable VTs were less prevalent with no difference in number of VTs induced, acute procedural success, and complications. The aforementioned trials demonstrated safety and feasibility of pLVAD support in patients with unstable VT undergoing ablation.