https://orcid.org/0000-0001-8346-1387
Abstract
Life-threatening refractory unstable ventricular arrhythmias in presence of advanced heart failure (HF) may determine haemodynamic impairment. Haemodynamic mechanical support (HMS) in this setting has a relevant role to restore end-organ perfusion. Catheter ablation (CA) of ventricular tachycardia (VT) is effective at achieving rhythm stabilization, allowing patient’s weaning from HMS, or bridging to permanent HF treatments. Acute heart decompensation during CA at anaesthesia induction in presence of advanced heart disease, in selected cases requires a preemptive HMS to prevent periprocedure adverse outcomes. Substrate ablation during sinus rhythm (SR) might be an effective strategy of ablation in presence of unstable VTs; however, in a minority of patients, it might have some limitations and might be unfeasible in some settings, including the case of the mechanical induction of several unstable VTs and the absence of ablation targets. In case of the persistent induction of unstable VTs after a previous failure of a substrate-based ablation in SR, a feasible alternative strategy of ablation might be VT activation/entrainment mapping supported by HMS. Multiple devices are available for HMS in the low-output states related to electrical storm and during CA of VT. The choice of the device is not standardized and it is based on the centres’ expertise. The aim of this article is to provide an up-to-date review on HMS for the management of life-threatening arrhythmias, in the context of catheter ablation and discussing our approach to manage critical VT patients.
Introduction
Refractory unstable ventricular arrhythmias and electrical storm (ES) are often life-threatening. In presence of complex cardiac substrates, heart failure (HF) and comorbidities, they may determine haemodynamic impairment, cardiogenic shock, and cardiac arrest, increasing patients’ mortality rate up to 50%. In this setting, restoration and maintenance of sinus rhythm (SR) is therefore of paramount importance.1 Antiarrhythmic drugs are usually the first line option, however they might be ineffective and may cause negative inotropic effects. Haemodynamic mechanical support (HMS) in this setting improves the mean arterial pressure (MAP) and provides adequate end-organ perfusion, in few cases contributing to SR restoration.1 Catheter ablation (CA) of ventricular tachycardia (VT) is a recognized treatment modality, effective at achieving rhythm stabilization,2–4 allowing patient’s weaning from HMS,5 or bridging to permanent HF treatments.6
Acute heart decompensation during CA of complex substrates may occur in 11% of cases and is associated with an increased risk of mortality.7 Frequently, VTs in presence of structural heart disease are poorly tolerated.3,8 Activation and entrainment mapping of unstable VTs might be unfeasible without arterial pressure support. Furthermore, the occurrence of severe refractory hypotension at anaesthesia induction, even before attempts of inducing VT, is not infrequent and is a recognized cause for failure.7,9,10 Substrate ablation during SR might be an alternative option, but it has limitations, especially in non-ischaemic substrates.11 Even if substrate ablation is performed entirely during SR, prolonged times under general anaesthesia in patients with a very high-risk profile, might be related to adverse outcomes7; preemptive HMS in a selected high-risk population might reduce the risk of acute heart deterioration and death/transplant during follow-up.12–14 The identification of patients at high risk of periprocedural haemodynamic deterioration is a matter of priority, since preemptive HMS in this subpopulation can increase CA safety profile12–14 and, due to the complexity and costs of HMS, those devices should be reserved to this very specific population.7,15
Large randomized clinical trials comparing different devices and different ablation strategies in the setting of cardiogenic shock and ES are not available; prospective multicentre data are missing also about patient risk stratification for prophylactic HMS before CA. The reason is possibly due to ethical and operational issues, in patients with life-threatening conditions.
In this article, we aim to present an up-to-date review on HMS for the management of life-threatening arrhythmias, focusing on its role in the context of CA and discussing our approach to manage critical VT patients.
Mechanical devices for haemodynamic support
Multiple devices are available for HMS in VT patients. The choice of the system depends on different factors: patient’s haemodynamic condition, acquaintance with devices and the goal of the mechanical support. In emergency situations, intra-aortic balloon pump (IABP) is often selected as the quickest way to obtain a minimum degree of haemodynamic stabilization; however, IABP might not be adequate in the worst clinical settings. Other devices such with a centrifugal pump, are more complex but more efficient support systems in case of threatening acute HF.16
Intra-aortic balloon pump
The most widely used device in low-output states is IABP,16 due to its simplicity and availability. It is positioned in the descending aorta, where the balloon is inflated during diastole, increasing diastolic pressure and coronary arteries perfusion. During systole, it is deflated, thus decreasing afterload and left ventricle workload. Despite low incidence of complications (Table 1), cardiac output (CO) can be increased only by 0.5 L/min. The beneficial haemodynamic effect progressively decreases at increasing heart rates, and it is almost lost at rates higher than 120 b/min, due to the impossibility to synchronize with the cardiac cycle17,18; furthermore, it may preclude CA with a retrograde aortic approach. Contraindications to IABP placement include moderate aortic regurgitation, peripheral vascular disease, and aortic aneurysms.
Summary of main studies reporting outcome of an activation approach and a substrate-based ventricular tachycardia ablation procedure
| Study | Design | Patients | Haemodynamic support | Aetiologies | Duration of follow-up | Short-term outcome | Long-term outcome |
|---|---|---|---|---|---|---|---|
| Volkmer 2006 | Retrospective | Stable VTs were mapped during VT (n. 16) and unstable VTs during SR (n. 14). | None | Post-MI cardiomyopathy | Mean of 25 ± 13 months | In the stable VT group, acute success rate was 91% and in the substrate group was 86% after up to two procedures. | There was no difference between the outcome of VT-mapping group and the substrate-mapping group. |
| Bunch 2012 | Retrospective | Patients with haemodynamically unstable VT (13 undergoing HMS-assisted VT ablation and 18 a substrate-based procedure). | TandemHeart | Structural heart disease (LVEF ≤ 40% or arrhythmogenic right ventricular cardiomyopathy) | Mean of 9 ± 3 months | No difference in inducibility between groups (10 of 13 vs. 12 of 18; 77 vs. 67%). There was no difference in acute complications including stroke or death. | Freedom from ICD-therapies for sustained VT were similar (P = 0.96). |
| Di Biase 2012 | Prospective | Patients with ES limited substrate ablation (n. 49) or endo-epicardial ablation of abnormal potentials within the scar (n. 43). | None | Ischaemic cardiomyopathy | Mean of 25 ± 10 months | All patients in both groups were free from inducible or spontaneous VT at the end of the procedure. No periprocedural complication occurred. | VT recurrence rate was 47% in conventional group and 19% in homogenization of the scar group (log-rank P = 0.006). |
| DI Biase 2015 | Randomized | Subjects with haemodynamically tolerated VT to clinical ablation (n. 60) vs. extensive substrate-based ablation (n. 58). | None | Post-MI cardiomyopathy | 12 months | Non-inducibility of the clinical VTs was achieved in all patients in both groups. | Substrate-based ablation was associated with significantly lower VT recurrence rates than clinical ablation (15.5% vs. 48.3%; log-rank P < 0.001). Overall mortality 8.6% in the substrate-based ablation group and 15.0%) in the clinical ablation group (log-rank P = 0.21). |
| Makimoto 2015 | Retrospective | Activation mapping (n. 35) vs. substrate-based strategy was adopted (n. 50) because of non-inducibility of VT or haemodynamic instability. | None | Arrhythmogenic right ventricular cardiomyopathy (n. 34), ischaemic heart disease (16), dilated cardiomyopathy (14), sarcoidosis (11), others (10) | Mean of 61 ± 40 months | No significant difference in the success of the procedure (63% in the activation mapping group and 74% in the substrate group, P = 0.27). | No significant differences in sustained VT recurrences (15/50 vs. 15/35, P = 0.22), and cardiac death (2/50 vs. 3/35, P = 0.38). |
| Ventura 2015 | Retrospective | Stable VTs were mapped during VT (n. 16) and unstable VTs during SR (n. 14). | None | Post-MI cardiomyopathy | Mean of 14 ± 6 months | Acute success was comparable being 69% in the stable VT group and 64% in the unstable VT group (P = 0.42). | VT recurrences were comparable being 25% in the stable VT group and 43% in the unstable VT group (P = 0.82). |
| Study | Design | Patients | Haemodynamic support | Aetiologies | Duration of follow-up | Short-term outcome | Long-term outcome |
|---|---|---|---|---|---|---|---|
| Volkmer 2006 | Retrospective | Stable VTs were mapped during VT (n. 16) and unstable VTs during SR (n. 14). | None | Post-MI cardiomyopathy | Mean of 25 ± 13 months | In the stable VT group, acute success rate was 91% and in the substrate group was 86% after up to two procedures. | There was no difference between the outcome of VT-mapping group and the substrate-mapping group. |
| Bunch 2012 | Retrospective | Patients with haemodynamically unstable VT (13 undergoing HMS-assisted VT ablation and 18 a substrate-based procedure). | TandemHeart | Structural heart disease (LVEF ≤ 40% or arrhythmogenic right ventricular cardiomyopathy) | Mean of 9 ± 3 months | No difference in inducibility between groups (10 of 13 vs. 12 of 18; 77 vs. 67%). There was no difference in acute complications including stroke or death. | Freedom from ICD-therapies for sustained VT were similar (P = 0.96). |
| Di Biase 2012 | Prospective | Patients with ES limited substrate ablation (n. 49) or endo-epicardial ablation of abnormal potentials within the scar (n. 43). | None | Ischaemic cardiomyopathy | Mean of 25 ± 10 months | All patients in both groups were free from inducible or spontaneous VT at the end of the procedure. No periprocedural complication occurred. | VT recurrence rate was 47% in conventional group and 19% in homogenization of the scar group (log-rank P = 0.006). |
| DI Biase 2015 | Randomized | Subjects with haemodynamically tolerated VT to clinical ablation (n. 60) vs. extensive substrate-based ablation (n. 58). | None | Post-MI cardiomyopathy | 12 months | Non-inducibility of the clinical VTs was achieved in all patients in both groups. | Substrate-based ablation was associated with significantly lower VT recurrence rates than clinical ablation (15.5% vs. 48.3%; log-rank P < 0.001). Overall mortality 8.6% in the substrate-based ablation group and 15.0%) in the clinical ablation group (log-rank P = 0.21). |
| Makimoto 2015 | Retrospective | Activation mapping (n. 35) vs. substrate-based strategy was adopted (n. 50) because of non-inducibility of VT or haemodynamic instability. | None | Arrhythmogenic right ventricular cardiomyopathy (n. 34), ischaemic heart disease (16), dilated cardiomyopathy (14), sarcoidosis (11), others (10) | Mean of 61 ± 40 months | No significant difference in the success of the procedure (63% in the activation mapping group and 74% in the substrate group, P = 0.27). | No significant differences in sustained VT recurrences (15/50 vs. 15/35, P = 0.22), and cardiac death (2/50 vs. 3/35, P = 0.38). |
| Ventura 2015 | Retrospective | Stable VTs were mapped during VT (n. 16) and unstable VTs during SR (n. 14). | None | Post-MI cardiomyopathy | Mean of 14 ± 6 months | Acute success was comparable being 69% in the stable VT group and 64% in the unstable VT group (P = 0.42). | VT recurrences were comparable being 25% in the stable VT group and 43% in the unstable VT group (P = 0.82). |
ES, electrical storm; HMS, haemodynamic mechanical support; ICD, implantable cardioverter-defibrillator; LVEF, left ventricular ejection fraction; MI, myocardial infarction; SR, sinus rhythm; VT, ventricular tachycardia.
Summary of main studies reporting outcome of an activation approach and a substrate-based ventricular tachycardia ablation procedure
| Study | Design | Patients | Haemodynamic support | Aetiologies | Duration of follow-up | Short-term outcome | Long-term outcome |
|---|---|---|---|---|---|---|---|
| Volkmer 2006 | Retrospective | Stable VTs were mapped during VT (n. 16) and unstable VTs during SR (n. 14). | None | Post-MI cardiomyopathy | Mean of 25 ± 13 months | In the stable VT group, acute success rate was 91% and in the substrate group was 86% after up to two procedures. | There was no difference between the outcome of VT-mapping group and the substrate-mapping group. |
| Bunch 2012 | Retrospective | Patients with haemodynamically unstable VT (13 undergoing HMS-assisted VT ablation and 18 a substrate-based procedure). | TandemHeart | Structural heart disease (LVEF ≤ 40% or arrhythmogenic right ventricular cardiomyopathy) | Mean of 9 ± 3 months | No difference in inducibility between groups (10 of 13 vs. 12 of 18; 77 vs. 67%). There was no difference in acute complications including stroke or death. | Freedom from ICD-therapies for sustained VT were similar (P = 0.96). |
| Di Biase 2012 | Prospective | Patients with ES limited substrate ablation (n. 49) or endo-epicardial ablation of abnormal potentials within the scar (n. 43). | None | Ischaemic cardiomyopathy | Mean of 25 ± 10 months | All patients in both groups were free from inducible or spontaneous VT at the end of the procedure. No periprocedural complication occurred. | VT recurrence rate was 47% in conventional group and 19% in homogenization of the scar group (log-rank P = 0.006). |
| DI Biase 2015 | Randomized | Subjects with haemodynamically tolerated VT to clinical ablation (n. 60) vs. extensive substrate-based ablation (n. 58). | None | Post-MI cardiomyopathy | 12 months | Non-inducibility of the clinical VTs was achieved in all patients in both groups. | Substrate-based ablation was associated with significantly lower VT recurrence rates than clinical ablation (15.5% vs. 48.3%; log-rank P < 0.001). Overall mortality 8.6% in the substrate-based ablation group and 15.0%) in the clinical ablation group (log-rank P = 0.21). |
| Makimoto 2015 | Retrospective | Activation mapping (n. 35) vs. substrate-based strategy was adopted (n. 50) because of non-inducibility of VT or haemodynamic instability. | None | Arrhythmogenic right ventricular cardiomyopathy (n. 34), ischaemic heart disease (16), dilated cardiomyopathy (14), sarcoidosis (11), others (10) | Mean of 61 ± 40 months | No significant difference in the success of the procedure (63% in the activation mapping group and 74% in the substrate group, P = 0.27). | No significant differences in sustained VT recurrences (15/50 vs. 15/35, P = 0.22), and cardiac death (2/50 vs. 3/35, P = 0.38). |
| Ventura 2015 | Retrospective | Stable VTs were mapped during VT (n. 16) and unstable VTs during SR (n. 14). | None | Post-MI cardiomyopathy | Mean of 14 ± 6 months | Acute success was comparable being 69% in the stable VT group and 64% in the unstable VT group (P = 0.42). | VT recurrences were comparable being 25% in the stable VT group and 43% in the unstable VT group (P = 0.82). |
| Study | Design | Patients | Haemodynamic support | Aetiologies | Duration of follow-up | Short-term outcome | Long-term outcome |
|---|---|---|---|---|---|---|---|
| Volkmer 2006 | Retrospective | Stable VTs were mapped during VT (n. 16) and unstable VTs during SR (n. 14). | None | Post-MI cardiomyopathy | Mean of 25 ± 13 months | In the stable VT group, acute success rate was 91% and in the substrate group was 86% after up to two procedures. | There was no difference between the outcome of VT-mapping group and the substrate-mapping group. |
| Bunch 2012 | Retrospective | Patients with haemodynamically unstable VT (13 undergoing HMS-assisted VT ablation and 18 a substrate-based procedure). | TandemHeart | Structural heart disease (LVEF ≤ 40% or arrhythmogenic right ventricular cardiomyopathy) | Mean of 9 ± 3 months | No difference in inducibility between groups (10 of 13 vs. 12 of 18; 77 vs. 67%). There was no difference in acute complications including stroke or death. | Freedom from ICD-therapies for sustained VT were similar (P = 0.96). |
| Di Biase 2012 | Prospective | Patients with ES limited substrate ablation (n. 49) or endo-epicardial ablation of abnormal potentials within the scar (n. 43). | None | Ischaemic cardiomyopathy | Mean of 25 ± 10 months | All patients in both groups were free from inducible or spontaneous VT at the end of the procedure. No periprocedural complication occurred. | VT recurrence rate was 47% in conventional group and 19% in homogenization of the scar group (log-rank P = 0.006). |
| DI Biase 2015 | Randomized | Subjects with haemodynamically tolerated VT to clinical ablation (n. 60) vs. extensive substrate-based ablation (n. 58). | None | Post-MI cardiomyopathy | 12 months | Non-inducibility of the clinical VTs was achieved in all patients in both groups. | Substrate-based ablation was associated with significantly lower VT recurrence rates than clinical ablation (15.5% vs. 48.3%; log-rank P < 0.001). Overall mortality 8.6% in the substrate-based ablation group and 15.0%) in the clinical ablation group (log-rank P = 0.21). |
| Makimoto 2015 | Retrospective | Activation mapping (n. 35) vs. substrate-based strategy was adopted (n. 50) because of non-inducibility of VT or haemodynamic instability. | None | Arrhythmogenic right ventricular cardiomyopathy (n. 34), ischaemic heart disease (16), dilated cardiomyopathy (14), sarcoidosis (11), others (10) | Mean of 61 ± 40 months | No significant difference in the success of the procedure (63% in the activation mapping group and 74% in the substrate group, P = 0.27). | No significant differences in sustained VT recurrences (15/50 vs. 15/35, P = 0.22), and cardiac death (2/50 vs. 3/35, P = 0.38). |
| Ventura 2015 | Retrospective | Stable VTs were mapped during VT (n. 16) and unstable VTs during SR (n. 14). | None | Post-MI cardiomyopathy | Mean of 14 ± 6 months | Acute success was comparable being 69% in the stable VT group and 64% in the unstable VT group (P = 0.42). | VT recurrences were comparable being 25% in the stable VT group and 43% in the unstable VT group (P = 0.82). |
ES, electrical storm; HMS, haemodynamic mechanical support; ICD, implantable cardioverter-defibrillator; LVEF, left ventricular ejection fraction; MI, myocardial infarction; SR, sinus rhythm; VT, ventricular tachycardia.
Tandem heart
Tandem heart is an arterial bypass system from the left atrium to the femoral artery. It is percutaneously inserted through a 21-French venous cannula, placed transeptally with either fluoroscopic or intracardiac echography guidance. A 17-French arterial cannula is placed femorally for return blood from the extracorporeal pump. Anticoagulation with unfractioned heparin is administered to maintain an activated clotting time (ACT) of 300 s. The left atrial and left femoral ports are connected to a centrifugal pump which can maintain a CO up to 4–5 L/min. The use of Tandem heart precludes CA with a transeptal approach and its output may need to be reduced to introduce the ablation catheters across the aortic valve.19 Due to the large sheaths’ size, vascular complications are often secondary to the use of this device; bleeding due to anticoagulation (including cardiac tamponade during left cannula placement), limb ischaemia, sepsis, and residual septal defects are also described.17,18 Contraindication to tandem heart might be severe vascular disease.
Impella
The Impella is an axial blood flow pump, percutaneously inserted through a 13-Fr arterial access and retrogradely positioned in the left ventricle across the aortic valve under the guidance of pressure sensors, built over its distal portion. Its position has to be confirmed with fluoroscopy and allows to pump blood from the left ventricle to the ascending aorta.18 An ACT between 250 and 300 s is required. The blood inlet and outlet are on the same cannula and specifically located in the left ventricle and thoracic aorta, respectively.19 Impella 2.5 can produce a CO up to 2.5 L/min, which can be improved up to 3.5 L/min with Impella CP and up to 5 L/min with Impella 5.0 at the expense of a larger vascular accesses (14-French sheath and a surgical cut down, respectively). Disadvantages of the Impella device are the interferences with electromagnetic mapping systems, in particular when mapping the outflow tract. This can be negotiated by a transeptal access, or, alternatively, reducing its performance level.20 Given comparable flow rates, left ventricular end-diastolic pressure and myocardial oxygen demand are more effectively reduced by the Impella device, than other HMS devices, due its capability to directly unload the left ventricle.17,18 Complications include vascular injuries, retroperitoneal bleeding, aortic valve injury, stroke, systemic embolism, and thrombus formation. Contraindication include: mechanical aortic prosthesis, severe aortic stenosis, significant aortic insufficiency, left ventricular (LV) thrombus, and severe vascular disease.16,17
Extracorporeal membrane oxygenation
Extracorporeal membrane oxygenation (ECMO) is a portable modification of a cardiopulmonary bypass, capable of supporting patients with refractory cardiopulmonary impairment without limitation of time.21 Use of venous-arterial (VA) ECMO stemmed from the evidence of a neurologically intact survival benefit over conventional cardiopulmonary resuscitation.22,23 The VA-ECMO is the only HMS which provides complete biventricular circulatory support. This device collects deoxygenated blood from the venous system (19- to 25-French cannula in the right atrium via femoral vein) to an oxygenator, where gas exchange occurs; therefore oxygenated blood is actively pumped into the arterial circulation (17- to 21-Fr cannula in the aorta via femoral artery). Both venous and arterial accesses are percutaneoulsly gained or, in specific instances, surgical techniques are used to obtain cannulation of the right atrium or the aorta. The circuit takes advantage of a centrifugal pump and an oxygenator (composed by a membrane and a heat exchanger); as blood flows into the oxygenator, a normal haemoglobin saturation is supplied, depending on the flow rate calibration; CO2 is removed through diffusion.23 The target ACT is >250 s.
As a secondary effect, ECMO increases afterload thus increasing wall stress and oxygen demand.24 To overcome this effect (LV distension) treatment options include: inotropes or Impella use to decrease wall stress, IABP to reduce LV afterload, mechanical decompression by shunting the blood to the right side, through septostomy, transeptal cannulation of the left ventricle or left atrium, or direct surgical apical drainage (LV venting). Further strategies include the removal of LV volume through the administration of diuretic agents, eventually through ultrafiltration or haemodialysis, and the reduction of ECMO flow.
V-A ECMO is an effective haemodynamic support to patients with refractory VT, ES, or recurrent ventricular fibrillation. It allows continues blood flow, supplying end-organ perfusion despite recurrent arrhythmias and during CA, bridging patients either to stabilization and ECMO withdrawal, heart transplantation, or permanent left ventricular assist device (VAD) implantation.1,18,24
Its main limitation is complexity and the need for a perfusionist16; described complications are major vascular events, peripheral arterial ischaemia, bleeding at site of cannulation, thrombosis of the ECMO cannula, embolism of LV thrombus. Caution might be taken treating patients with aortic regurgitation and impaired systolic function; in this subpopulation VA-ECMO is poorly tolerated due to volume overload, requiring higher inotropic support.25
Haemodynamic support during cardiogenic shock secondary to refractory VTs/electrical storm
Cardiogenic shock is defined as systemic tissue hypo-perfusion secondary to impaired CO despite adequate circulatory volume and LV filling pressure. HMS, in this setting, should be initiated as soon as possible, particularly when patients are unresponsive to pharmacological treatment. Early initiation of HMS can reduce the consequences of systemic hypoperfusion, as worsening ischaemia and declining cardiac function.16 Few studies have presented data from patients with cardiogenic shock secondary to ventricular arrhythmias, and many of them do not provide a detailed description of arrhythmia management beyond haemodynamic stabilization with HMS.1,26,27 These studies suggest that arrhythmias are a reversible cause of cardiac arrest with patients’ prognosis strictly related to rhythm stabilization. The achievement of haemodynamic stability in some cases might be followed by immediate SR restoration.1 In other cases, whenever clinical stabilization and weaning from HMS are precluded by refractory arrhythmias, recovery of SR might be achieved performing CA.27 Ablation of ventricular arrhythmias showed a synergic role with HMS, as together they might represent a feasible and effective strategy, aiming to achieve electric and haemodynamic stabilization.
Patients with unstable VTs: what’s the best ablation strategy?
Substrate modification approach in SR is a safe strategy of VT ablation; however, it might have some limitations in selected cases related to the choice of ablation target sites; LAVAs28 and late potentials29 are actually downstream consequences of the conduction slowing that, in turn, is the site where most frequently the VT isthmus is blocked.30 In 22% of a previous series of patients who underwent VT ablation, HMS was required because of the failure of a previous unsuccessful substrate-based ablation and consequent persistent induction of unmappable non-tolerated VTs;12 these data underline the need for an alternative strategy in selected few cases in whom SR substrate modification is unfeasible or failed.
Some data recently demonstrated the effectiveness of CA long-term results, when ablation is performed during VT, allowing the complete tracking of the diastolic pathway.31 These findings, if supported by similar studies in the future, may provide the rational of ‘extending’ the concept of mapping guided VT ablation, at least in specific settings, as a prior failed substrate ablation modification procedure, VTs originating in the context of extensive scars, VTs secondary to non-ischaemic dilated cardiomyopathies and when ablation targets are not easily identifiable during SR. In all those cases, an alternative strategy to SR substrate modification might be required.
Multiple non-randomized trials have compared the two strategies in patients with stable arrhythmias.9
One study by Bunch et al.32 provides evidences that HMS might allow activation mapping of VT, with comparable outcomes and complications to an exclusive substrate mapping in SR. Haemodynamic support allowed the termination of all the inducible VTs.
A meta-analysis including six retrospective observational studies9,32–36 showed that, compared to a predominantly activation/entrainment-guided ablation strategy, a substrate-based ablation strategy had similar acute results, long-term outcome and complications rates37 (Table 1). No significant differences in procedural duration were found. Activation/entrainment mapping of VT with HMS might be a reasonable alternative strategy in patients with unstable VTs, who underwent a previous failing substrate-based ablation of VT during SR, followed by the persistent induction of VT, despite the accomplishment of substrate modification, or in case of VT recurrences after ablation.
Haemodynamic support during catheter ablation of VT
Among the possible different devices, Counterpulsation was proven to be the least effective for the haemodynamic assistance during CA of VT.10
Several studies have tested the feasibility of CA of unstable VTs with the use of HMS, aiming to avoid a drop of the MAP during fast VTs, maintain haemodynamic balance and prevent end-organ hypoperfusion, allowing to safely perform activation and entrainment mapping of all the index VTs and achieve successful results of CA12,13,19,20,32,38–42 (Tables 2 and 3). In the comparative studies among Impella and IABP, the use of Impella allowed the induction and mapping of unstable VTs for a longer time and requiring fewer terminations of VTs due to haemodynamic instability, as compared to the non-Impella group (IABP or no device at all). No differences between the two groups were demonstrated in surrogates of end-organ perfusion, nor in CA acute and long-term outcome.20,40 Unstable VTs were sustained longer, allowing a greater number of VT terminations, during faster procedures in the Impella group, as compared to the patients supported only by inotropic drugs and to the non-assisted group.40
Summary of main studies reporting complications related to VT ablation procedure assisted by haemodynamic support
| Author year | n. patients | Major bleeding | Major vascular | Limb ischaemia | AKI | CVA | FU | Outcomes |
|---|---|---|---|---|---|---|---|---|
| VA-ECMO | ||||||||
| Baratto 2016 | 64 | 0 | 0 | 2 (3) | 4 (6) | 0 | 23 months | VT recurred in 33%, overall mortality was 12%, and rate of transplantation was 9%. |
| Enriquez 2018 | 18 | 3 (14) | 2 (10) | 0 | – | – | Median 10 days (range 1 days–27 months) | Seven patients survived >6 months. Five of these remained free of VT/VF and three ultimately received LVAD or heart transplant. |
| Di Monaco 2019 | 19 | 5 (26) | 3 (16) | 0 | 0 | 0 | 10 months | The procedural success rate was 68% and the Kaplan–Meier mortality rate was 21%. |
| Tandem heart | ||||||||
| Bunch 2012 | 13 | 1 (7) | 1 (7) | 0 | 0 | 0 | ||
| Impella | ||||||||
| Kusa 2017 | 109 | 7 (6) | 7 (6) | 0 | 0 | 0 | 7 months | VT, heart transplantation, or death occurred in 36% of the pLVAD vs. 26% of the non-pLVAD groups. |
| IABP | ||||||||
| Reddy 2014 | 22 | 3 (15) | 0 | – | – | – | 12 months | Mortality and VT recurrence were similar between IABP group and Impella and TandemHeart cohort. |
| Aryana 2017 | 115 | – | – | – | 26 (22) | – | 12 months | Catheter ablation of VT associated with mechanical support using PVAD vs. IABP was associated with reduced in-hospital cardiogenic shock, renal failure, length of stay, hospital readmissions, and mortality, but no difference in redo-VT ablation at 1 year. |
| Author year | n. patients | Major bleeding | Major vascular | Limb ischaemia | AKI | CVA | FU | Outcomes |
|---|---|---|---|---|---|---|---|---|
| VA-ECMO | ||||||||
| Baratto 2016 | 64 | 0 | 0 | 2 (3) | 4 (6) | 0 | 23 months | VT recurred in 33%, overall mortality was 12%, and rate of transplantation was 9%. |
| Enriquez 2018 | 18 | 3 (14) | 2 (10) | 0 | – | – | Median 10 days (range 1 days–27 months) | Seven patients survived >6 months. Five of these remained free of VT/VF and three ultimately received LVAD or heart transplant. |
| Di Monaco 2019 | 19 | 5 (26) | 3 (16) | 0 | 0 | 0 | 10 months | The procedural success rate was 68% and the Kaplan–Meier mortality rate was 21%. |
| Tandem heart | ||||||||
| Bunch 2012 | 13 | 1 (7) | 1 (7) | 0 | 0 | 0 | ||
| Impella | ||||||||
| Kusa 2017 | 109 | 7 (6) | 7 (6) | 0 | 0 | 0 | 7 months | VT, heart transplantation, or death occurred in 36% of the pLVAD vs. 26% of the non-pLVAD groups. |
| IABP | ||||||||
| Reddy 2014 | 22 | 3 (15) | 0 | – | – | – | 12 months | Mortality and VT recurrence were similar between IABP group and Impella and TandemHeart cohort. |
| Aryana 2017 | 115 | – | – | – | 26 (22) | – | 12 months | Catheter ablation of VT associated with mechanical support using PVAD vs. IABP was associated with reduced in-hospital cardiogenic shock, renal failure, length of stay, hospital readmissions, and mortality, but no difference in redo-VT ablation at 1 year. |
Data are presented as number (%).
AKI, acute kidney injury; CVA, cerebrovascular accident; FU, follow-up; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device; pLVAD, permanent LVAD; VA-ECMO, veno-arterial extracorporeal membrane oxygenator; VT, ventricular tachycardia.
Summary of main studies reporting complications related to VT ablation procedure assisted by haemodynamic support
| Author year | n. patients | Major bleeding | Major vascular | Limb ischaemia | AKI | CVA | FU | Outcomes |
|---|---|---|---|---|---|---|---|---|
| VA-ECMO | ||||||||
| Baratto 2016 | 64 | 0 | 0 | 2 (3) | 4 (6) | 0 | 23 months | VT recurred in 33%, overall mortality was 12%, and rate of transplantation was 9%. |
| Enriquez 2018 | 18 | 3 (14) | 2 (10) | 0 | – | – | Median 10 days (range 1 days–27 months) | Seven patients survived >6 months. Five of these remained free of VT/VF and three ultimately received LVAD or heart transplant. |
| Di Monaco 2019 | 19 | 5 (26) | 3 (16) | 0 | 0 | 0 | 10 months | The procedural success rate was 68% and the Kaplan–Meier mortality rate was 21%. |
| Tandem heart | ||||||||
| Bunch 2012 | 13 | 1 (7) | 1 (7) | 0 | 0 | 0 | ||
| Impella | ||||||||
| Kusa 2017 | 109 | 7 (6) | 7 (6) | 0 | 0 | 0 | 7 months | VT, heart transplantation, or death occurred in 36% of the pLVAD vs. 26% of the non-pLVAD groups. |
| IABP | ||||||||
| Reddy 2014 | 22 | 3 (15) | 0 | – | – | – | 12 months | Mortality and VT recurrence were similar between IABP group and Impella and TandemHeart cohort. |
| Aryana 2017 | 115 | – | – | – | 26 (22) | – | 12 months | Catheter ablation of VT associated with mechanical support using PVAD vs. IABP was associated with reduced in-hospital cardiogenic shock, renal failure, length of stay, hospital readmissions, and mortality, but no difference in redo-VT ablation at 1 year. |
| Author year | n. patients | Major bleeding | Major vascular | Limb ischaemia | AKI | CVA | FU | Outcomes |
|---|---|---|---|---|---|---|---|---|
| VA-ECMO | ||||||||
| Baratto 2016 | 64 | 0 | 0 | 2 (3) | 4 (6) | 0 | 23 months | VT recurred in 33%, overall mortality was 12%, and rate of transplantation was 9%. |
| Enriquez 2018 | 18 | 3 (14) | 2 (10) | 0 | – | – | Median 10 days (range 1 days–27 months) | Seven patients survived >6 months. Five of these remained free of VT/VF and three ultimately received LVAD or heart transplant. |
| Di Monaco 2019 | 19 | 5 (26) | 3 (16) | 0 | 0 | 0 | 10 months | The procedural success rate was 68% and the Kaplan–Meier mortality rate was 21%. |
| Tandem heart | ||||||||
| Bunch 2012 | 13 | 1 (7) | 1 (7) | 0 | 0 | 0 | ||
| Impella | ||||||||
| Kusa 2017 | 109 | 7 (6) | 7 (6) | 0 | 0 | 0 | 7 months | VT, heart transplantation, or death occurred in 36% of the pLVAD vs. 26% of the non-pLVAD groups. |
| IABP | ||||||||
| Reddy 2014 | 22 | 3 (15) | 0 | – | – | – | 12 months | Mortality and VT recurrence were similar between IABP group and Impella and TandemHeart cohort. |
| Aryana 2017 | 115 | – | – | – | 26 (22) | – | 12 months | Catheter ablation of VT associated with mechanical support using PVAD vs. IABP was associated with reduced in-hospital cardiogenic shock, renal failure, length of stay, hospital readmissions, and mortality, but no difference in redo-VT ablation at 1 year. |
Data are presented as number (%).
AKI, acute kidney injury; CVA, cerebrovascular accident; FU, follow-up; IABP, intra-aortic balloon pump; LVAD, left ventricular assist device; pLVAD, permanent LVAD; VA-ECMO, veno-arterial extracorporeal membrane oxygenator; VT, ventricular tachycardia.
Summary of main studies reporting outcomes following use of mechanical circulatory support devices during ventricular tachycardia ablation
| Study | Design | Patients | Haemodynamic support | Aetiologies | Duration of follow-up | Short-term outcome | Long-term outcome |
|---|---|---|---|---|---|---|---|
| Reddy 2014 | Prospective, multi-centre registry | 66 | IABP (n. 22), Impella 2.5 (n. 25), and Tandem Heart (n. 19) | 70% IHD, 30% NIDC | 12 ± 5 months | Compared to IABP, although with the use of Impella or Tandem Heart, more unstable VTs were mapped (1.05 vs. 0.32; P < 0.001) and a greater number of VTs was terminated (1.59 vs. 0.91; P = 0.007), periprocedural success and complications were similar. | Compared to IABP, Impella, and TandemHeart cohort had similar VT recurrence (42% vs. 50%) and mortality (36% in both). |
| Baratto 2016 | Retrospective, single-centre | 64 | ECMO (n. 64) | 45% IHD, 55% NIDC | 23 ± 13 months | No inducible VT achieved in 69%. In-hospital mortality was 1.5%. Major complications included acute kidney injury (6%), vascular injury (3%), and acute heart failure (8%). | At follow-up, VT recurred in 33%, overall mortality was 12%, and rate of transplantation was 9%. Transplant and LVAD-free survival was 69%. |
| Turagam 2017 | Retrospective, multicentre study (International VT Ablation Center Collaborative group) | 105 | Impella 2.5 ECMO Tandem Heart | 60% IHD, 40% NIDC | 527 days (208–1048) | Compared to 1550 undergoing VT ablation without HS, patients in the HS group were sicker, acute procedural success was lower and complications higher. | One-year mortality was significantly higher in the HS group. The use of HS was an independent predictor of mortality. However, in patients with LVEF <20% and NYHA functional Class III to IV, there was also no significant difference in clinical outcomes when compared with no HS. |
| Aryana 2017 | Retrospective analysis of US Medicare database | 345 | PVAD (not reported if Impella, ECMO or Tandem Heart) (n. 230), IABP (n. 115) | 60% IHD, 40% NIDCM | >12 months | Compared to IABP, PVAD was associated with reduced mortality (6.5% vs. 19.1%), cardiogenic shock (9.1% vs. 23.5%), acute kidney injury (11.7% vs. 21.7%), 30-day re-hospitalization (27.0% vs. 37.8%), and shorter hospital LOS (8.4 vs. 10.6 days). | PVAD group had similar re- do VT ablation rates at 1-year, compared with IABP (10.2% vs. 14.0%; P = 0.34). |
| Kusa 2017 | Retrospective, single-centre, PM | 194 | Impella 2.5 (n. 80), Impella CP (n. 29), control (n. 85)a | 57% IHD; 4% NIDCM | 7.17 months | In PM analysis, procedure duration was longer in Impella group but no significant difference in VT inducibility at procedure end (14% vs. 10%; P = 0.43) and complications (11% vs. 3%; P = 0.18) or hospital LOS. | No significant difference between Impella and control group in PM analysis for mortality (5% vs. 8%; P = 0.50), transplantation (5% vs. 0%; P = 0.25) or recurrent VT (26% vs. 21%; P = 0.29). |
| Study | Design | Patients | Haemodynamic support | Aetiologies | Duration of follow-up | Short-term outcome | Long-term outcome |
|---|---|---|---|---|---|---|---|
| Reddy 2014 | Prospective, multi-centre registry | 66 | IABP (n. 22), Impella 2.5 (n. 25), and Tandem Heart (n. 19) | 70% IHD, 30% NIDC | 12 ± 5 months | Compared to IABP, although with the use of Impella or Tandem Heart, more unstable VTs were mapped (1.05 vs. 0.32; P < 0.001) and a greater number of VTs was terminated (1.59 vs. 0.91; P = 0.007), periprocedural success and complications were similar. | Compared to IABP, Impella, and TandemHeart cohort had similar VT recurrence (42% vs. 50%) and mortality (36% in both). |
| Baratto 2016 | Retrospective, single-centre | 64 | ECMO (n. 64) | 45% IHD, 55% NIDC | 23 ± 13 months | No inducible VT achieved in 69%. In-hospital mortality was 1.5%. Major complications included acute kidney injury (6%), vascular injury (3%), and acute heart failure (8%). | At follow-up, VT recurred in 33%, overall mortality was 12%, and rate of transplantation was 9%. Transplant and LVAD-free survival was 69%. |
| Turagam 2017 | Retrospective, multicentre study (International VT Ablation Center Collaborative group) | 105 | Impella 2.5 ECMO Tandem Heart | 60% IHD, 40% NIDC | 527 days (208–1048) | Compared to 1550 undergoing VT ablation without HS, patients in the HS group were sicker, acute procedural success was lower and complications higher. | One-year mortality was significantly higher in the HS group. The use of HS was an independent predictor of mortality. However, in patients with LVEF <20% and NYHA functional Class III to IV, there was also no significant difference in clinical outcomes when compared with no HS. |
| Aryana 2017 | Retrospective analysis of US Medicare database | 345 | PVAD (not reported if Impella, ECMO or Tandem Heart) (n. 230), IABP (n. 115) | 60% IHD, 40% NIDCM | >12 months | Compared to IABP, PVAD was associated with reduced mortality (6.5% vs. 19.1%), cardiogenic shock (9.1% vs. 23.5%), acute kidney injury (11.7% vs. 21.7%), 30-day re-hospitalization (27.0% vs. 37.8%), and shorter hospital LOS (8.4 vs. 10.6 days). | PVAD group had similar re- do VT ablation rates at 1-year, compared with IABP (10.2% vs. 14.0%; P = 0.34). |
| Kusa 2017 | Retrospective, single-centre, PM | 194 | Impella 2.5 (n. 80), Impella CP (n. 29), control (n. 85)a | 57% IHD; 4% NIDCM | 7.17 months | In PM analysis, procedure duration was longer in Impella group but no significant difference in VT inducibility at procedure end (14% vs. 10%; P = 0.43) and complications (11% vs. 3%; P = 0.18) or hospital LOS. | No significant difference between Impella and control group in PM analysis for mortality (5% vs. 8%; P = 0.50), transplantation (5% vs. 0%; P = 0.25) or recurrent VT (26% vs. 21%; P = 0.29). |
ECMO, extracorporeal membrane oxygenation; HS, haemodynamic support; IABP, intra-aortic balloon pump; IHD, ischaemic heart disease; LOS, length of stay; LVAD, left ventricular assist device; LVEF, left ventricular ejection fraction; NIDC, •••; NIDCM, non-ischaemic dilated cardiomyopathy; NYHA, New York Heart Association; PM, propensity matched; PVAD, percutaneous ventricular assist device; VT, ventricular tachycardia.
Seventy-eight patients included in 1:1 propensity-matched analysis.
Summary of main studies reporting outcomes following use of mechanical circulatory support devices during ventricular tachycardia ablation
| Study | Design | Patients | Haemodynamic support | Aetiologies | Duration of follow-up | Short-term outcome | Long-term outcome |
|---|---|---|---|---|---|---|---|
| Reddy 2014 | Prospective, multi-centre registry | 66 | IABP (n. 22), Impella 2.5 (n. 25), and Tandem Heart (n. 19) | 70% IHD, 30% NIDC | 12 ± 5 months | Compared to IABP, although with the use of Impella or Tandem Heart, more unstable VTs were mapped (1.05 vs. 0.32; P < 0.001) and a greater number of VTs was terminated (1.59 vs. 0.91; P = 0.007), periprocedural success and complications were similar. | Compared to IABP, Impella, and TandemHeart cohort had similar VT recurrence (42% vs. 50%) and mortality (36% in both). |
| Baratto 2016 | Retrospective, single-centre | 64 | ECMO (n. 64) | 45% IHD, 55% NIDC | 23 ± 13 months | No inducible VT achieved in 69%. In-hospital mortality was 1.5%. Major complications included acute kidney injury (6%), vascular injury (3%), and acute heart failure (8%). | At follow-up, VT recurred in 33%, overall mortality was 12%, and rate of transplantation was 9%. Transplant and LVAD-free survival was 69%. |
| Turagam 2017 | Retrospective, multicentre study (International VT Ablation Center Collaborative group) | 105 | Impella 2.5 ECMO Tandem Heart | 60% IHD, 40% NIDC | 527 days (208–1048) | Compared to 1550 undergoing VT ablation without HS, patients in the HS group were sicker, acute procedural success was lower and complications higher. | One-year mortality was significantly higher in the HS group. The use of HS was an independent predictor of mortality. However, in patients with LVEF <20% and NYHA functional Class III to IV, there was also no significant difference in clinical outcomes when compared with no HS. |
| Aryana 2017 | Retrospective analysis of US Medicare database | 345 | PVAD (not reported if Impella, ECMO or Tandem Heart) (n. 230), IABP (n. 115) | 60% IHD, 40% NIDCM | >12 months | Compared to IABP, PVAD was associated with reduced mortality (6.5% vs. 19.1%), cardiogenic shock (9.1% vs. 23.5%), acute kidney injury (11.7% vs. 21.7%), 30-day re-hospitalization (27.0% vs. 37.8%), and shorter hospital LOS (8.4 vs. 10.6 days). | PVAD group had similar re- do VT ablation rates at 1-year, compared with IABP (10.2% vs. 14.0%; P = 0.34). |
| Kusa 2017 | Retrospective, single-centre, PM | 194 | Impella 2.5 (n. 80), Impella CP (n. 29), control (n. 85)a | 57% IHD; 4% NIDCM | 7.17 months | In PM analysis, procedure duration was longer in Impella group but no significant difference in VT inducibility at procedure end (14% vs. 10%; P = 0.43) and complications (11% vs. 3%; P = 0.18) or hospital LOS. | No significant difference between Impella and control group in PM analysis for mortality (5% vs. 8%; P = 0.50), transplantation (5% vs. 0%; P = 0.25) or recurrent VT (26% vs. 21%; P = 0.29). |
| Study | Design | Patients | Haemodynamic support | Aetiologies | Duration of follow-up | Short-term outcome | Long-term outcome |
|---|---|---|---|---|---|---|---|
| Reddy 2014 | Prospective, multi-centre registry | 66 | IABP (n. 22), Impella 2.5 (n. 25), and Tandem Heart (n. 19) | 70% IHD, 30% NIDC | 12 ± 5 months | Compared to IABP, although with the use of Impella or Tandem Heart, more unstable VTs were mapped (1.05 vs. 0.32; P < 0.001) and a greater number of VTs was terminated (1.59 vs. 0.91; P = 0.007), periprocedural success and complications were similar. | Compared to IABP, Impella, and TandemHeart cohort had similar VT recurrence (42% vs. 50%) and mortality (36% in both). |
| Baratto 2016 | Retrospective, single-centre | 64 | ECMO (n. 64) | 45% IHD, 55% NIDC | 23 ± 13 months | No inducible VT achieved in 69%. In-hospital mortality was 1.5%. Major complications included acute kidney injury (6%), vascular injury (3%), and acute heart failure (8%). | At follow-up, VT recurred in 33%, overall mortality was 12%, and rate of transplantation was 9%. Transplant and LVAD-free survival was 69%. |
| Turagam 2017 | Retrospective, multicentre study (International VT Ablation Center Collaborative group) | 105 | Impella 2.5 ECMO Tandem Heart | 60% IHD, 40% NIDC | 527 days (208–1048) | Compared to 1550 undergoing VT ablation without HS, patients in the HS group were sicker, acute procedural success was lower and complications higher. | One-year mortality was significantly higher in the HS group. The use of HS was an independent predictor of mortality. However, in patients with LVEF <20% and NYHA functional Class III to IV, there was also no significant difference in clinical outcomes when compared with no HS. |
| Aryana 2017 | Retrospective analysis of US Medicare database | 345 | PVAD (not reported if Impella, ECMO or Tandem Heart) (n. 230), IABP (n. 115) | 60% IHD, 40% NIDCM | >12 months | Compared to IABP, PVAD was associated with reduced mortality (6.5% vs. 19.1%), cardiogenic shock (9.1% vs. 23.5%), acute kidney injury (11.7% vs. 21.7%), 30-day re-hospitalization (27.0% vs. 37.8%), and shorter hospital LOS (8.4 vs. 10.6 days). | PVAD group had similar re- do VT ablation rates at 1-year, compared with IABP (10.2% vs. 14.0%; P = 0.34). |
| Kusa 2017 | Retrospective, single-centre, PM | 194 | Impella 2.5 (n. 80), Impella CP (n. 29), control (n. 85)a | 57% IHD; 4% NIDCM | 7.17 months | In PM analysis, procedure duration was longer in Impella group but no significant difference in VT inducibility at procedure end (14% vs. 10%; P = 0.43) and complications (11% vs. 3%; P = 0.18) or hospital LOS. | No significant difference between Impella and control group in PM analysis for mortality (5% vs. 8%; P = 0.50), transplantation (5% vs. 0%; P = 0.25) or recurrent VT (26% vs. 21%; P = 0.29). |
ECMO, extracorporeal membrane oxygenation; HS, haemodynamic support; IABP, intra-aortic balloon pump; IHD, ischaemic heart disease; LOS, length of stay; LVAD, left ventricular assist device; LVEF, left ventricular ejection fraction; NIDC, •••; NIDCM, non-ischaemic dilated cardiomyopathy; NYHA, New York Heart Association; PM, propensity matched; PVAD, percutaneous ventricular assist device; VT, ventricular tachycardia.
Seventy-eight patients included in 1:1 propensity-matched analysis.
Some other studies were able to demonstrate an impact of HMS on patient’s outcome as compared to the IABP group, despite similar redo ablation rates39 (Table 2).
In the one prospective comparative multicentre study of patients who underwent IABP vs. p-VAD (Impella, or Tandem Heart) supported CA procedures. An impact of p-VAD on patients’ outcome was still not demonstrated.43
Despite being a group of patients at higher risk, those who received p-VAD, experienced a similar combined rate of recurrent VTs, heart transplantation and death, as compared to the relatively healthier non-pVAD group.42 Taking into account that patients in the HMS group have a higher incidence of ES and comorbidities with a more advanced functional class, they consequently have a significantly higher incidence of complications, in-hospital death and 1-year mortality, as compared to the non-HMS group. Thus, despite no significant difference in VT recurrence between HMS and no- HMS at 12 months, patients who required circulatory support seemed to be at higher risk of death, requiring further treatment for HF to improve their outcome.41 This confirms that HMS is not able itself to provide an adjunctive anti-arrhythmic effect to the ablation procedure, but its effect is of protection of the population at higher risk of death, against periprocedure adverse outcomes, allowing comparable outcomes to the no-p-VAD patients. The use of HMS in a sicker population with advanced heart disease and ES, which was expected to show a higher mortality rate, resulted in a better than expected outcome; these data do not support the wide use of these devices in all patients undergoing scar-related VT ablation, but in an accurately selected population most likely to benefit of it: those with dilated cardiomyopathy, ES, advanced HF, and severe LV dysfunction.42
Five of the aforementioned comparative studies were included in a meta-analysis, reporting procedural and clinical outcomes of patients who underwent ablation with HMS, vs. patients who did not receive intra-procedure haemodynamic support.13,32,40–42 Data of outcome, after at least 1 month of follow-up were analysed. The majority of patients were supported with Impella,13,41–43 followed by tandem heart.13,32,40–42 Ablation strategy was a combined approach of substrate ablation in SR and activation/entrainment mapping during VT in four studies,13,41–43 and only in one of them patients not receiving p-VAD were exclusively ablated on the base of substrate map during SR.32 There was no difference in terms of acute ablation success, CA-related complications, VT recurrence, nor mortality during the long-term follow-up in the two groups. Compared to the non-p-VAD, CA duration time was longer in the p-VAD group. Haemodynamic support was proven to be effective in allowing the induction and mapping of VTs during CA, with comparable outcomes of the non-supported patients.
Haemodynamic support with ECMO during CA of VT was previously reported in 64 consecutive patients with unstable, drug-refractory VTs12 (Tables 2 and 3). This study focused on ECMO support, not only as an HMS for fast VTs during ablation, but, adjunctively, as a comprehensive circulatory support of end-stage HF. In high-risk and decompensated patients, it might be a supportive strategy from the admission, to ablation and during the very first hours after the procedure, to overcome end-organ deterioration and treat low-output states. In patients with end-stage biventricular HF the combined strategy of ECMO and ablation of VT can be proposed as a bridge to destination therapies of HF. A clinical case at our centre is described in Figure 1.
A case of VT catheter ablation in the setting of electrical storm and ischaemic heart disease with severely impaired left ventricular ejection fraction (23%). The first line ablation strategy was substrate modification due to haemodynamic instability during VT and small caliber femoral vessels (ECMO was at high risk of peripheral ischaemia). Following catheter insertion in the left ventricle, incessant ventricular tachycardia was mechanically induced, determining electrical instability and consequently a low-output state, precluding a substrate modification approach (A and B). Rescue ECMO (a reperfusion cannula prevented peripheral ischaemia) allowed eventual rhythm stabilization and substrate mapping. A wide area of low voltage was recorded, involving the entire inferior wall (C), in absence of late potentials at the LAT map (D) and abnormal electrograms recorded during sinus rhythm at the site of ablation (E). During ECMO support, VT induction was possible, achieving a mean arterial pressure of ≥90 mmHg during VT (F). VT activation mapping was performed (G) and the diastolic pathway of the re-entrant circuit was located at the mid-basal inferior wall (H). RF ablation at this site determined interruption of the tachycardia and non-inducibility of any VT at the end of the procedure (I). At 18 months of follow-up, no VT recurrences were recorded. ECMO, extracorporeal membrane oxygenation; LAT, local activation time; RF, radiofrequency; VT, ventricular tachycardia.
Principal studies on CA supported by HMS are summarized in Tables 2 and 3.
There is no evidence from prospective randomized studies yet of a clear benefit on outcome of one device, compared to the others, however every device has its pros and its cons and each operator might be confident with one device, over the others, due to personal and team experiences and the availabilities of the centre.
The Impella 2.5 and Tandem Heart devices are superior to IABP during ablation of unstable VTs,20,43 however, they have some contraindications (mechanical aortic valves, LV thrombosis, and ventricular septal defects), limiting their application in this setting. Additional limitations are electromagnetic interferences (Impella 2.5) and the requirement of trans-septal approach (Tandem Heart). Furthermore, when the right ventricular assistance or respiratory support are required, ECMO is the only possible device of choice. ECMO allows multiple accesses to the left ventricle, effectively supporting extensive endo-epicardial mapping, in absence of electromagnetic interferences and without time limitation of use12. ECMO setting includes the presence of a perfusionist, and its availability is not granted in every electrophysiology laboratory; however, such device is superior to the others in the setting of advanced heart disease, because its biventricular support provides an adequate MAP during VT induction, favouring end-organ perfusion, thus improving haemodynamic balance and reducing HF episodes. It might also be the bridge to end-stage HF therapies in case of the persistence of low-output states despite the achievement of stable SR after VT ablation and optimal medical therapy. In our centre, we follow an algorithm for HMS management as shown in Figure 2.
Decisional algorithm at San Raffaele Center Milan. In case of paroxysmal VTs, the choice of preemptive haemodynamic support is performed on the base of risk stratification upon admission. In case of high risk, a preemptive ECMO is provided periprocedurally to prevent adverse outcomes. In case of cardiogenic shock secondary to refractory VTs, ECMO is inserted to achieve rhythm stabilization; CA is then performed aiming to rhythm stabilization and consequent ECMO weaning. CA, catheter ablation; ECMO, extracorporeal membrane oxygenation; HMS, haemodynamic mechanical support; LAT, local activation time; LVAD, left ventricular assist device; VT, ventricular tachycardia.
Preemptive vs. rescue haemodynamic support
Timing of HMS is of pivotal importance in patients undergoing CA of VT with complex substrates, multiple co-morbidities and a recent history of multiple VT-related implantable cardioverter-defibrillator (ICD) shocks. These patients have an intrinsic risk of periprocedural death, which becomes even higher in relation to the arrhythmia burden.44 The late HMS initiation after procedure-related acute haemodynamic decompensation have a prognostic negative impact with an up to 50% of mortality at 21 ± 7 months of follow-up.7
The same results were confirmed within the Milan series of 74 ECMO-supported CA procedures. In fact, in the five patients in whom ECMO was implemented as an intra-procedure bailout strategy, faster VTs were observed, reducing the efficacy of ECMO in preventing end-organ hypoperfusion.12
The university of Pennsylvania reported a series of 21 patients who underwent CA of VT, in whom acute heart deterioration required ECMO support. Even achieving a procedural success rate of 83%, after a median follow-up of 10 days, 88% of patients died, in 2% of them due to refractory VT, thus confirming that ECMO support as a rescue strategy, allows the performance of CA but is not able to prevent acute mortality related to heart decompensation.24
Nevertheless, the results of CA supported preemptively with HMS (Impella or Tandem Heart) are promising. The direct comparison of rescue vs. preemptive p-VAD vs., the absence of HMS during CA of VT in a population of 21 patients showed a higher 30-day mortality rate in the rescue group as compared to the others, with higher procedure time, despite similar ablation results.13 Accordingly, the identification of patients who should undergo HMS preemptively is crucial and should be based on a clinical evaluation upon patients’ admission at the hospital. Experience of high volume centres has favoured the implementation of algorithms that, on the base of clinical arrhythmia pattern, haemodynamic tolerance, clinical status and co-morbidities, are effective in identifying the patients at a higher risk of adverse periprocedural outcome.45
An effective risk stratification of patients upon admission can be performed based on the clinical variables at admission.7 In a large multicentre series of 1251 patients, six demographic clinical and procedure-related variables were evaluated as potential prognostic factors using the Survival Tree analysis method. LV ejection fraction (LVEF), ICD/cardiac resynchronization device, previous ablation were, in hierarchical order, identified as best predictors of VT recurrence; while LVEF, previous ablation, ES were identified as best predictors of mortality. Three groups with significantly different survival rates were identified. The highest mortality and VT recurrence rates were described in the high-risk subgroup.15
The PAINESD risk score has been developed by the University of Pennsylvania group. The predictors of acute heart deterioration that were incorporated in the PAINESD risk score included age, diabetes, ischaemic cardiomyopathy, reduced LVEF, chronic obstructive pulmonary disease, presentation with VT storm, and New York Heart Association functional class III/IV. An additional variable that was also found to be associated with increased risk of AHD (Acute Heart Decompensation) was the use of general anaesthesia.7
Prophylactic p-VAD insertion guided on the base of the PAINED risk score in a series of 75 patients undergoing scar-related VT ablation proved to be effective producing better outcomes as compared to a propensity-matched population of 75 patients who undergo CA of VT without p-VAD insertion.14 Despite comparable ablation results, the group who underwent CA supported with HMS showed higher survival and transplant-free survival and a lower incidence of acute HF after ablation, as compared to the group who underwent CA without HMS.14
These results show that VT patients should undergo an accurate risk stratification before the ablation procedure, because, even when approaching ablation in an apparent state of haemodynamic stability, the most complex patients might eventually face a periprocedural acute haemodynamic impairment which is a factor severely impacting on patients’ survival. Patients stratified to be at high risk for VT ablation, in absence of contraindications, should be protected with preemptive haemodynamic support independently from the aetiology of heart disease, the strategy designed for VT ablation, and haemodynamic status at the beginning of the procedure.
Conclusion
Haemodynamic support is an effective treatment of cardiogenic shock secondary to refractory VTs; it allows patient’s stabilization when a bailout CA procedure is needed to treat refractory arrhythmias. It is also useful to preemptively support high-risk patients who require an ablation of unstable VTs. The choice of the device should be considered according to team expertise and the availability of tertiary referral centres. Aims of HMS should be the restoration of adequate circulatory status, allowing CA and bridging patients to HF destination therapies.
Conflict of interest: none declared.
References
Notes
Biography: Following graduation in Medicine, Dr Paolo Della Bella spent several years in basic research focusing on cardiac sympathetic reflexes under the head of Prof. Malliani. Later moved to Clinical Electrophysiology with Prof. Wellens in Maastricht, and later in London-Ontario and Oklahoma City. Since 1992, he was Head of Arrhythmia Unit at CC-Monzino, and in 2010, he moved to San Raffaele Hospital, where is currently head of Arrhythmia Department. His research focused on VTs in ischaemic and non-ischaemic cardiomyopathy, multielectrode substrate and VT activation mapping, bipolar ablation of septal VTs, haemodynamic support in heart failure VT patients. He first envisioned the concept of the VT Unit, a dedicated structure for the treatment of VT patients. He performed over 5000 electrophysiological procedures as main operator. He is author of over 150 clinical articles and member of the editorial board of several international journals. He’s Scientific Director of learning programmes addressed to international Electrophysiologists.
Biography: Dr Andrea Radinovic graduated in Medicine in 2006 at the San Raffaele University Hospital in Milan. He specialized in cardiology and works in the Arrhythmia Unit and Electrophysiology Laboratories at the San Raffaele Hospital with Dr Paolo Della Bella since 2011. His fields of interest include management of cardiac arrhythmias and prevention of sudden cardiac death. His clinical practice focuses on the treatment of heart rhythm disorder with catheter ablation and his research has been published on international Journals and book chapters. He is a co-investigator and coordinator in international and national multicentric clinical trials on ventricular arrhythmias.
Biography: Dr Luca Rosario Limite works at San Raffaele Hospital, Milan as cardiologist in the Arrhythmia Department led by Dr Paolo Della Bella. His day-to-day activities involve electrophysiological procedures and devices implantations. Dr Limite graduated in Medicine in 2013 at San Raffaele University under the supervision of Professor Cianflone and later moved to Sapienza University of Rome as resident in the Cardiology Department, chaired by Professor Volpe. Here, he developed his interest in hypertrophic cardiomyopathy, with particular focus in the management of arrhythmic risk and sudden cardiac death. He is author and co-author of over 20 peer-reviewed papers focused on sudden cardiac death, cardiomyopathies, atrial fibrillation, and ventricular tachycardia.
Biography: Dr Francesca Baratto graduated in Medicine in 2006 at San Raffaele University Hospital, Milan with Prof Maseri, and later moved to Centro Cardiologico Monzino, as resident in Cardiology under the head of Dr Paolo Della Bella, in 2007, where she started her training in Electrophysiology, Arrhythmia and Device Implantation. She later moved to San Raffaele Hospital, where she currently works as Cardiologist in the Arrhythmia Department directed by Dr Della Bella; she performs clinical activity and attends the Electrophysiology laboratories as first and second operator, performing electrophysiological procedures and devices implantations. She is author and co-author of over 25 clinical articles on cardiac arrhythmias mainly focused on ventricular tachycardia, the role of haemodynamic support in advanced heart failure ventricular tachycardia patients and ventricular tachycardia patients’ management, and exposed her research work during some international meetings. She is reviewer for some international journals in the field of Electrophysiology.
