OUP user menu

Ventricular arrhythmias in patients with implanted ventricular assist devices: a contemporary review

Jonathan T. Shirazi, John C. Lopshire, Irmina Gradus-Pizlo, Mohammed A. Hadi, Thomas C. Wozniak, Adnan S. Malik
DOI: http://dx.doi.org/10.1093/europace/eus364 11-17 First published online: 11 December 2012

Abstract

Ventricular arrhythmia (VA) is a significant factor in the clinical management of patients with congestive heart failure (CHF). Understanding the implications of VA in ventricular assist device-supported CHF patients is critical to appropriate clinical decision making in this special population. This article details research findings on this topic, and attempts to link them to practical patient management strategies.

  • Ventricular tachycardia
  • Ventricular fibrillation
  • Congestive heart failure
  • Left ventricular assist device

Introduction

As implantable mechanical circulatory support emerges as a new standard therapy in end-stage congestive heart failure (CHF) patients, bridging to recovery, bridging to transplant (BTT), or as destination therapy (DT), the study of adverse effects from ventricular assist devices (VADs) will be crucial in improving patient outcomes. There are several significant adverse events that occur in VAD patients. Table 1 represents the 10 most common adverse events that were listed in the first report from INTERMACS (Interagency Registry for Mechanical Circulatory Support).1 In subsequent annual reports from INTERMACS, the focus shifted to reporting risk factors and cause of death as opposed to absolute adverse event rates. In the second INTERMACS report, the adverse event categories of cardiac failure, including right ventricular failure, ventricular tachycardia (VT), and ventricular fibrillation (VF) were the most common causes of death in LVAD patients.2

View this table:
Table 1

Ventricular assist device adverse events—first annual INTERMACS report

Adverse event (AE)Patients with AE ≤30 days (% of total patients)Patients with AE >30days (% of total patients)Total patients with AE (% of total patients)
Infection100 (23.8%)107 (25.5%)160 (38.1%)
Bleeding114 (27.1%)64 (15.2%)146 (34.8%)
Other AEs60 (14.3%)90 (21.4%)136 (32.4%)
Respiratory failure74 (17.6%)14 (3.3%)84 (20.0%)
Cardiac arrhythmia62 (14.8%)22 (5.2)%77 (18.3%)
Neurological dysfunction51 (12.1%)33 (7.9%)77 (18.3%)
Hypertension28 (6.7%)41 (9.8%)66 (15.7%)
Renal dysfunction54 (12.9%)12 (2.9%)64 (15.2%)
Device malfunction12 (2.9%)42 (10.0%)53 (12.6%)
Hepatic dysfunction27 (6.4%)18 (4.3%)41 (9.8%)
  • Data collected from June 2006 to December 2007 (n = 420).

Methods

A review of the literature was performed via an OVID Medline search. The exploded keyword terms ‘Heart Assist Device’ and ‘Assisted Circulation’ were combined with the following exploded keyword terms: ‘Ventricular Tachycardia,’ ‘Cardiac Arrhythmias,’ ‘Ventricular Fibrillation,’ and ‘Implantable Defibrillators.’ Studies were limited to English language and those with abstracts. All abstracts were reviewed. The studies cited shared the focus of this review article, VT or VF in patients with implanted VADs. All efforts were made to evaluate the primary references within every article with this focus. There were studies that were assessed, but generally omitted from this review. Articles evaluating the use of VADs with the primary purpose of supporting a patient with recurrent or refractory VT, as well as studies utilizing temporary (e.g. Impella, TandemHeart) or extracorporeal systems were reviewed and determined to be outside the primary scope of this review.

Ventricular arrhythmia in ventricular assist device-supported patients

Sustained VT and VF are haemodynamically unstable heart rhythms and often life threatening if not terminated. However, many VAD patients are able to tolerate VT specifically for periods of several hours or longer.3,4 In fact, there have been multiple reports of VAD implantation for patients with incessant VT as the indication.510 Ventricular arrhythmia does cause a drop in VAD output, confer a risk of thrombus formation,3 and increase the risk of right ventricular failure when sustained.11 Furthermore, even among patients who recover cardiac function and can be weaned from mechanical circulatory support, VA can still occur with fatal outcomes.12

Due to the incomplete haemodynamic support offered by LVADs, mechanical circulatory support purely for short-term management of incessant, medically refractory VT or VF often requires biventricular support. Additional options might also include total artificial heart (TAH) or veno-arterial extracorporeal membrane oxygenation in select situations. Experience with these latter two advanced therapies for this indication will hopefully begin to accrue as technical advancements make them increasingly viable.

Among common VAD parameters, VA may manifest as a drop in device pulsatility index and/or flow due to inadequate preload. In ambulatory VAD patients, this may be represented clinically by palpitations, presyncopal or syncope symptoms, or CHF symptoms such as dyspnoea or fluid retention, but may be asymptomatic if recent in onset. In the setting of an LVAD, recurrent VT or VF can precipitate de novo, or compound pre-existing, right ventricular dysfunction, sometimes necessitating urgent RVAD support.3

An important early observation was that BTT VAD patients experiencing VA were less likely to survive to transplant. In a study of 111 VAD patients, Bedi et al. identified a 33% mortality rate in patients who experienced VA compared with an 18% mortality in those who did not experience VA (P < 0.001). Furthermore, in patients who experienced VA within the first week after VAD implant, a mortality rate of 54% compared with 9% in those who did not experience VA was identified (P < 0.001).13 Other investigators found increased mortality in VAD patients who experienced sustained VA requiring ICD therapy.14 A study by Ziv et al.11 also showed a trend towards VAD patients with VA having increased all-cause mortality, but the study was underpowered to demonstrate significance. These findings may have been related to the type of VAD device being studied. For instance, VA in a patient with pulsatile VAD support might confer higher mortality risk compared with VA in a patient on contemporary continuous flow device support. Investigation of this assumption can be performed through the analysis of INTERMACS data.

The first INTERMACS report reviewed data obtained from June 2006 to December 2007, which pertained exclusively to pulsatile flow devices.1 After US Food and Drug Administration approval of the first continuous flow VAD for BTT in April 2008, and subsequent approval of the device for DT in January 2010, the shift towards continuous flow LVADs was marked. The second annual INTERMACS report, comprised of data from June 2006 to March 2009, included 528 pulsatile VADs (48.4%) and 564 continuous flow VADs (51.6%).2 The third annual INTERMACS report included data through September 2010, wherein continuous flow VADs comprised 77.3% (1936/2506) of implants, compared with pulsatile VADs, which comprised 22.7% (570/2506).15 The increased use of continuous flow VADs is likely due the devices' smaller size, their proven efficacy, prolonged survival, and reductions in device-related malfunction in large trials for both BTT16,17 and DT patients.18 The most recent, fourth annual INTERMACS report included data from June 2006 to June 2011 and reiterated the trends seen in the third report. The fourth annual report found a trend towards implanting devices in patients before multiple organ dysfunction occurred. Of the 4366 total patients receiving LV support from devices, 2.2% (99 patients) had the TAH, 19.7% (862 patients) had pulsatile devices, and 77.9% (3405 patients) had continuous flow devices.19 It is important to note that the second, third, and fourth annual reports did not find VT or VF to be a statistically significant risk factor for death after VAD implantation.2,15,19 Although it is attractive to attribute this discrepancy in findings between the earlier and later INTERMACS analyses to specific characteristics of pulsatile versus continuous flow devices, or to greater experience with VAD patient management, a definitive explanation has not been reported in the literature.

Ventricular arrhythmia morphology and onset

One retrospective study evaluated the types of VA observed in 91 patients both before and after pulsatile VAD placement.11 It is important to note that because of the retrospective nature of this study, there may have been a selection bias potentially excluding VT-prone patients from VAD therapy. Nevertheless, 30 patients experienced VA before LVAD placement, and 32 patients experienced VA after LVAD placement. The study identified a statistically significant (P = 0.001) increase in new-onset monomorphic ventricular tachycardia (MVT) post-VAD. More specifically, 9 patients experienced MVT before VAD placement, and 23 experienced MVT post-VAD placement. Only four patients experienced MVT both before and after VAD. The post-VAD MVT was also significantly (P = 0.04) faster, with a cycle length of 308 ± 88 ms compared with an average pre-VAD cycle length of 355 ± 82 ms. The authors discussed several theories for the mechanism of increased MVT, including myocardial scarring from the apical insertion of the VAD. In contrast, they also discussed improvement in cardiac output being a possible mechanism in the suppression of VA in some patients post-VAD implant. There was no statistically significant de novo polymorphic VT or VF post-VAD. This study also identified an increased risk of VA in the early post-operative time period. Thirty-one out of 32 patients with post-VAD VT experienced the arrhythmias within the first 14 days after VAD placement.11

The increased occurrence of VA in the early post-operative period is corroborated by multiple studies, in both pulsatile and continuous flow device types.1,16,2023 Table 2 displays the number of patients who experienced early (≤30 days) versus late (>30 days) VA after VAD implantation. The first report from INTERMACS comprised adverse event data almost entirely from patients with pulsatile VADs implanted from June 2006 to December 2007. The incidence of early VA was 14.7%, while late VA incidence was 5%. The Heartmate II investigators in a multi-centre, prospective trial published in 2009, investigated continuous flow LVADs in BTT patients. While total VA incidence was 19.9%, early arrhythmias still accounted for the majority of instances, with an early arrhythmia incidence of 13.2% compared with late incidence of 8%.20 An earlier study from the Heartmate II investigators, reviewing data from 133 patients from March 2005 to May 2006, elicited a higher occurrence of early VA.16 Expectedly, over time and across multiple centres, there was a reduction in both early and overall VA. There was another, albeit smaller, prospective study from Denmark evaluating VA in patients with continuous flow VADs at one center from March 2006 to July 2008. This study also found a higher incidence of early VA. Interestingly, this centre had a higher overall incidence of VAs (52%) than other data reviewed.22 Other studies reported the incidence of experiencing sustained VT or VF in the range of 18–52%.1,11,16,20,22,23

View this table:
Table 2

Early vs. late ventricular arrhythmia incidence in ventricular assist device patients

0–30 days arrhythmia incidence>30 days arrhythmia incidenceTotal study incidence
Kirklin et al.162/421 patients (14.7%)22/421 patients (5%)127/421 patients (18.3%)
Pagani et al.2037/281 patients (13.2%)23/281 patients (8%)56/281 patients (19.9%)
Miller et al.1624/133 patients (18.0%)8/133 patients (6.0%)32/133 patients (24.0%)
Andersen et al.229/23 patients (39.1%)3/23 patients (13.0%)12/23 patients (52%)

Risk factors and mechanisms of arrhythmia

Several statistically significant risk factors have been identified for the development of VA post-VAD implantation. However, many of these risk factors have not been corroborated by subsequent studies. Table 3 is a listing of all the statistically significant risk factors for developing post-VAD VA that have been identified.

View this table:
Table 3

Risk factors for developing ventricular arrhythmias post-ventricular assist device implant

Older patientsAmbardekar et al.14
Ischaemic heart diseaseZiv et al.11, Bedi et al.13
Nonischaemic heart diseaseOswald et al.25
Destination therapyAmbardekar et al.14
Pre-VAD history of VAsOswald et al.25, Cantillon et al.23
Non-usage of beta-blockersRefaat et al.24
Post-Op electrolyte abnormalityZiv et al.11
Increase in QTc immediately post-VAD placementHarding et al.21
Low LVAD flows immediately post-OpZiv et al.11

While ischaemic heart disease (ICM) was a significant risk factor for post-VAD implant VA in some studies,11,13 this was not confirmed by other studies.14,2224 In fact, other investigators identified non-ischaemic cardiomyopathy (NICM) as a risk factor for post-VAD VA.25 Sixty-one patients who received VADs for medically refractory heart failure included 49% with ICM and 51% with NICM. Non-ischaemic cardiomyopathy patients experienced a 50% VA incidence versus 22% in the ICM group. The study also identified pre-VAD history of VA as a risk factor for post VAD VA (50 vs. 24%).25 Pre-VAD VA history has also been found to be a statistically significant risk factor by other investigators.23 The study that identified non-usage of beta-blockers as a risk for post-VAD VA included patients with non-sustained VT along with those who experienced sustained VT or VF.24 Other studies did not find non-usage of beta-blockers as a significant risk factor. Other investigators have identified older age and destination therapy as the only statistically significant risk factors for VA.14 An increase in the QTc interval 1 day post-operatively was also a statistically significant risk factor for developing VA in one study.21

Investigators have proposed multiple mechanisms for VA in VAD patients that may be unique to device therapy (Table 4). Apical scarring at the LVAD inflow cannula site has been proposed as one mechanism.11,13,26 Because of the nature of continuous flow devices providing cardiac output regardless of ventricular preload, the possibility for arrhythmia during a suction event, or complete decompression of the ventricle, has also been described.22,27

View this table:
Table 4

Proposed mechanisms for ventricular arrhythmia in ventricular assist device patients

Apical scarring from LVAD inflow siteDrakos et al.26, Ziv et al.11, Bedi et al.13
Persistent or recurrent myocardial ischaemiaZiv et al.11, Bedi et al.13
Intrinsic arrhythmogenicity due to fibrosis or myocyte remodellingBedi et al.13
Use of inotropic drugs post-operativelyZiv et al.11, Bedi et al.13
Suction events in continuous flow devicesVollkron et al.27, Andersen et al.22
QTc prolongation from unloading of cardiomyopathic heartsHarding et al.30
Changes in ion channel and gap junction regulationRefaat et al.24

Molecular and accompanying electrophysiological changes resulting in VA have also been implicated. One retrospective study performed a comparative genetic analysis on the myocardium of six patients who experienced post-VAD VA to six patients who did not experience post-VAD VA. The tissue samples were taken from the LV apex at the time of LVAD implantation; tissue samples were also obtained at the time of cardiac explantation from the same patients. In the patients who experienced post-VAD VA the following was discovered: (i) upregulation of the sodium–calcium exchanger (NCX), (ii) downregulation of the voltage gated potassium channel (Kv4.3), (iii) downregulation of the sodium/potassium ATP pump, and (iv) downregulation of connexin 43 (Cx43). The investigators noted that the increase in NCX could predispose to delayed after-depolarizations by increasing intracellular sodium influx. The combination of the increase in NCX and decrease in Kv4.3 channels has been shown to increase action potential duration in animal studies. The investigators commented that the downregulation of Cx43 alone has been linked to VT in animal studies.24 Another study investigated molecular changes that occurred in the myocardium of CHF patients who underwent VAD implantation. Tissue samples of seven patients taken before and after VAD support (a median duration of 59.7 ± 36.1 days) were analysed. There was a statistically significant upregulation in calcium handling and sarcomeric genes, including calcium ATPase and ryanodine receptor 2. These investigators also found an increase in cardiofibroblast genes as well.28

Several studies have investigated the electrocardiographic (ECG) changes that take place in cardiomyopathic hearts when the heart is unloaded, which in a physiological sense might be pertinent to VAD-supported patients. One study monitored the ECG changes that took place during cardiac unloading in patients with cardiomyopathy (LV ejection fraction < 30% and enlarged LV end-diastolic dimension by echocardiogram) and 23 normal (control) patients. The investigators utilized head-up tilt table testing to unload the ventricles of the patients and recorded ECG at 5 and 25min. The control group had statistically significant shortening of the QRS complex at 5 and 25min of head-up tilt table testing whereas the cardiomyopathic patients had no statistically significant change in QRS duration. Furthermore, the control group had QTc shortening at 5min and no change at 25min. In contrast, the cardiomyopathic patients experienced statistically significant QTc prolongation at both 5 and 25min of head-up tilt table testing. Subgroup analysis revealed that bundle branch block, antiarrhythmic therapy, or baseline prolonged QTc interval did not account for QTc prolongation seen in the cardiomyopathic patients.29 An earlier, retrospective study by Harding et al. evaluated QRS and QT durations in patients undergoing VAD implantation. Evaluation of ECG taken within 6 h post-operatively demonstrated statistically significant shortening of QRS duration and increase in QT and QTc duration (379 ± 10 to 504 ± 11 ms). The authors proposed two possible mechanisms for the QTc prolongation in the early post-VAD period: inactivation of the swelling-activated chloride channel and an increase in the NCX.30 An increase in the NCX was confirmed by Refaat et al.24

The Harding et al. study also evaluated ECGs taken ≥1 week post-operatively. A statistically significant reduction in QTc from 504 ± 11 to 445 ± 9 ms was discovered. This QTc interval reduction after the immediate post-operative period has been corroborated by two prospective studies of VAD patients.26,31 Xydas et al. evaluated ECG, echocardiographic, neurohormonal, and histological data in 36 VAD patients. The authors found QRS duration shortening at 2, 30, 60, and 90 days post-VAD implantation. QTc duration shortening became statistically significant at 30 days. The reduction in QTc over the first 60 days correlated with echocardiographic reduction in left ventricular end-diastolic dimension.31 Drakos et al. noted statistically significant QRS and QTc duration shortening at both 1week and 6 months post-VAD implantation. These investigators suggest that the finding of sustained QRS and QTc shortening seen at the later, 6-month time point demonstrated the presence of reverse electrophysiological remodelling.26

Risk reduction and interventions for ventricular arrhythmias

After acknowledging the identified risk factors and proposed mechanisms for VA in the VAD population, interventions can be more accurately pinpointed. Cost-effective and straightforward interventions such as protocols for post-operative electrolyte replacement and monitoring as well as consideration for early initiation of beta-blocker therapy seem reasonable. Improvements might also be achieved with increased physician experience, refined hospital protocols, and system-based improvements. This seems a reasonable conclusion when evaluating the improvements seen by the Heartmate II investigators' subsequent, multicentre study outcome data.16,20 There is no intervention for advanced age, designation of destination therapy, or CHF aetiology. However, a history of VA before VAD placement may direct us to more carefully manage and monitor such patients post-operatively.

There have been several studies evaluating ICDs in VAD patients. One retrospective study evaluated 33 patients with ICDs and either a continuous flow VAD or pulsatile VAD. These investigators found that patients who experienced appropriate ICD therapy for VA were 5.3 times less likely to survive to transplant or with LVAD support when compared with those who did not receive an ICD shock (P = 0.023). Interestingly, this study also found an increased frequency of inappropriate ICD shocks in patients with the pulsatile VADs compared with those with continuous flow VADs. The finding of any ICD shock, appropriate or inappropriate, was also a statistically significant hazard for survival (HR 4.5, P = 0.027). Of note, 18% of the patients in this study received inappropriate ICD therapy vs. 24% receiving appropriate ICD shocks.14 A separate study also found evidence linking ventricular arrhythmias and survival. These investigators performed a retrospective study of 478 patients who underwent VAD placement, including a majority of pulsatile VADs with continuous flow and extracorporeal systems also included. There was a focus on 90 patients with concomitant ICD in place. Twenty-six out of 90 (28.9%) of the ICD patients experienced VA. Crude mortality was lower in the ICD group compared with the non-ICD group (24.4% compared with 36.9%, P = 0.026). Cardiovascular death was more common in the non-ICD group when compared with the ICD group (82 vs. 55%, P = 0.02). The ICD group had longer median survival (295 vs. 226 days, P = 0.24) and, concurrently, a larger proportion of the ICD group patients survived to transplant. Of note, only 3.3% of ICD patients (3/90) in this study received inappropriate ICD shocks. The investigators concluded that concomitant ICD in VAD recipients is associated with extended survival.23 With acknowledgement of this study's finding that pre-VAD VA is associated with post-VAD VA, the investigators recommended further studies to focus on sub-groups who benefit from ICD therapy with VAD support.

One prospective study of patients with ICDs and continuous flow LVADs followed 61 patients for 365 ± 321 days. The majority of the patients, 46, had ICDs placed for primary prevention, while 15 patients had ICDs previously placed for secondary prevention of VT or VF. Twenty-one patients (34%) received appropriate ICD therapy for VA, while 15 patients (25%) experienced inappropriate ICD therapy; specifically, 13 patients experienced inappropriate antitachycardia pacing, and 2 patients experienced ICD shocks.25 This study corroborated earlier findings that there is a higher likelihood of appropriate post-VAD ICD therapy in patients who had the ICDs placed for secondary prevention rather than primary prevention.23 The investigators, Oswald et al.25 recommended that patients with ICDs at the time of VAD placement would benefit from the ICD remaining, but that a general statement in favour of ICD implantation in all patients receiving VADs could not be made. Finally, the 2008 ACC/AHA/HRS guidelines for device-based therapy for arrhythmias are without explicit mention of VADs, but include a IIa recommendation (‘reasonable to consider’) in non-hospitalized BTT patients.32

There have been studies reporting a role for VA ablation in patients with durable implanted VADs.3335 In a retrospective study, Mulloy and colleagues analysed 14 patients with recurrent pre-VAD VA; half of these patients received a peri-procedural cryoablation while the other 7 received no ablation. Of the seven who received cryoablation, four were stable enough to have received an electrophysiology (EP) study mapping the location of the arrhythmic origin. The remaining three patients, deemed too unstable for a pre-VAD EP study, received systematic ECG analysis to determine arrhythmia origin. The epicardial and endocardial cryoablation was performed immediately after sternotomy and after initiation of cardiopulmonary bypass, respectively. The group who underwent peri-procedural cryoablation enjoyed a statistically significant reduction in post-VAD recurrent VT, with 0/7 patients in the cryoablation group compared with 4/7 patients in the no-cryoablation group experiencing post-VAD recurrent VA (P = 0.02). The group treated with cryoablation also benefited from fewer reintubations (0/7 vs. 3/7 patients, P = 0.05), reduction in total time in ICU (6.9 ± 4.6 vs. 18.4 ± 13.3 days, P = 0.01), and a reduction in time from surgery to discharge (26 ± 19 vs. 57 ± 31 days, P = 0.03).33 There are also reports of catheter-directed radiofrequency (RF) ablation in patients with implanted VADs.34,35 Dandamudi and colleagues reported three very different clinical scenarios wherein patients underwent EP mapping and RF ablation for VA post-VAD placement. Briefly, the first case involved a young male who suffering from incessant, medically refractory VT 5 days after receiving an emergent three-vessel coronary artery bypass surgery. He underwent VAD placement as he clinically deteriorated, but continued to experience VA. He was taken to EP lab and underwent VT mapping and RF ablation. He experienced no post-ablation VA. The authors describe two other cases of post-VAD VA that were taken to EP lab for RF ablation; there was also a discussion of the technical difficulties that may be encountered when performing catheter-directed RF ablation in a patient with an implanted VAD, which included difficulty obtaining vascular access when no pulsatile flow is palpable, decreased catheter manoeuvrability in the left ventricle secondary to decreased LV volume, and the tendency for the mapping catheter to be pulled towards the inflow cannula.35

Discussion and management strategies

Ventricular arrhythmia is common in patients with VADs. There is an increased incidence of patients experiencing these arrhythmias in the early post-operative period. The majority of patients tolerate these arrhythmias well, owing to a more reliable cardiac output provided by the VAD. While patients with VADs can generally tolerate sustained VT and VF for a period of time, there should be an effort to place these patients back into normal rhythm, given the potential adverse consequences (haemodynamic compromise, thrombus formation, and RV dysfunction) of prolonged VA. While early studies found increased mortality in those patients experiencing VT or VF after VAD placement, larger more recent reviews do not find VA as a risk factor for death. Currently, no prospective, authoritative data exist on which to base practice recommendations.

Typical VA assessment and management should consider the following. Conventional pharmacotherapy with beta-blockers and perhaps spironolactone, which has been demonstrated to reduce sudden death in CHF patients, are low-risk interventions that can be applied in most pre- and post-VAD patients. Similarly, limitation of proarrhythmic inotropic therapies, as well as mindfulness to avoiding QT-prolonging drugs such as antipsychotics or associated antibiotic classes could be given special emphasis. Careful attention should be given to the possibility that VAs are being precipitated by VAD suction events. These may correlate to patient volume depletion (due to decreased pre-load) or to excessive unloading by the assist device. In the former case, volume loading would be appropriate, and in the latter, consideration towards decreasing VAD pump speed. If problems are recurrent, evaluation with echocardiogram, chest X-ray, and/or chest computed tomography scan to rule out VAD inflow cannula misalignment may be necessary. Antiarrhythmic therapy with amiodarone often has a role for refractory arrhythmias (both ventricular and atrial) in this population as well; for patients with recurrent pre-VAD VA in particular, continuation of amiodarone in the post-operative state is reasonable. Furthermore, patients who experience recurrent VT or any episode of VF post-operatively should receive antiarrhythmic therapy, barring traditional contraindication or identification of suction event-induced arrhythmia. Routine prophylactic antiarrhythmic medication is not necessary. Finally, any patient who becomes haemodynamically unstable from VA should be treated with DC cardioversion. In cases of recurrent VA pre-VAD placement, endo- and epicardial cryoablation may be reasonable at centres experienced with cryoablation, emphasizing cooperation among the advanced heart failure, electrophysiology, and cardiothoracic teams. The other approach to patients experiencing recurrent or symptomatic VA post-VAD placement, refractory to the aforementioned medical therapies, is the use of catheter-directed RF ablation. Table 5 suggests a phased approach to VA assessment and management in VAD patients.

View this table:
Table 5

Ventricular arrhythmia management in ventricular assist device-supported patients

Pre-implant:
  • Assessment of VA history and risk factors as well as presence and settings of ICD

  • Electrophysiology consultation in patients with high pre-VAD VA burden

  • Options based on individual center experience: TAH rather than VAD, post-VAD RF ablation, post-VAD ICD placement, combined peri-procedural cryoablation with VAD implantation

Early post-operative period (<30 days):
  • ICD therapy reactivation after surgery

  • Weaning of inotropic and pressor support as tolerated

  • Early initiation of beta-blocker ± aldosterone antagonist

  • Early electrocardiographic monitoring of cardiac rhythm and QTc

  • Documentation of VA and associated therapies as part of daily assessment

  • Avoidance of QT prolonging agents

  • Assessment of patient volume status, VAD speed

  • Assessment of cannula position

  • Early cardioversion-defibrillation, antiarrhythmic therapy for unstable VA

  • Electrophysiology consultation for refractory VA

Late post-operative period (>30 days):
  • Review pre- and post-VAD VA history

  • If recurrent VA, electrophysiology consultation to discuss ablative therapy or, if not already present and consistent with long-term goals of care, ICD placement

  • In patients with pre-existing ICD, routine device follow-up and interrogation per protocol

  • Continue to optimize HF therapy in patients at risk of developing VA

  • Evaluate for VA when presented with worsening CHF or syncopal symptoms

There do appear to be some sub-groups of patients that have improved survival with ICDs and continuous flow VADs. Moreover, the proportion of patients experiencing VT or VF has not dramatically changed despite the evolution from pulsatile to present day continuous flow VADs.

For the patient with a pulsatile VAD, concomitant ICD therapy would be a reasonable approach, particularly with this population's higher mortality when VAs are experienced. However, as the majority of devices being utilized today are continuous flow devices, ICD therapy can likely be reserved for those patients who have a documented history of VT or VF before VAD, given their much higher likelihood for VT or VF post implantation, or for those who have new VT or VF beyond the immediate post-operative period. If an ICD is already in place it should probably remain implanted, given the slight incremental risk of generator and/or lead removal. For DT patients, it also may be appropriate to consider abstaining from ICD therapy in patients without significant VA history, recognizing that most patients with VAD support and VT or VF will not experience sudden incapacity, and that painful shocks could be incongruent with perceived quality of life. (Similarly, in end-of-life situations involving patients with pre-existing ICDs, deactivation of the defibrillator should be discussed in tandem with discontinuation of VAD support.) In the presence of VA, maximization of oral CHF and antiarrhythmic therapy would be an appropriate initial approach. In our institutional experience, DT patients, because of their higher comorbidity burden, may be more prone to complications of ICD placement in the early post-VAD implant period. Such complications can include pocket haematoma due to VAD-related anticoagulation; infection of the ICD pocket, or even of the VAD due to transient bacteremia; and wound-healing abnormalities due to poor nutritional status. As such, watchful waiting and resorting to ICD placement at a later time, and only if necessary, seems to make sense.

Analysis of contemporary VAD patient populations vis a vis VA will be imperative going forward to afford them the best outcomes. Careful study and refinement of patient management must continue in order to further reduce the potential risks and morbidity to VAD-supported patients of sustained VA.

Conflict of interest: none declared.

References

View Abstract