Left ventricular assist device implantation outcomes in patients with subcutaneous implantable cardioverter-defibrillators: A case series

Abstract

Background

Left ventricular assist devices (LVADs) may induce electromagnetic interference (EMI) affecting implanted cardiac devices, including more novel subcutaneous implantable cardiac defibrillators (S-ICDs).

Methods

In this case series, the authors retrospectively reviewed courses of 6 patients with S-ICDs who underwent LVAD implantation at a single center.

Results

Of the 6 patients reviewed, 4 experienced inappropriate ICD shocks, of which 3 resulted from EMI. Five of the 6 patients ultimately had S-ICD therapies disabled.

Conclusions

Due to EMI resulting in inappropriate shocks and improved tolerability of malignant arrhythmias, deactivation or removal of S-ICDs should be considered in patients undergoing LVAD implantation.

1. Introduction

Electromagnetic interference (EMI) is a sequela of left ventricular assist device (LVAD) implantation in patients with preexisting cardiac implantable electronic devices (CIEDs) [1,2]. In LVAD-related EMI, the spinning impeller and magnet of the LVAD generate an electromagnetic field, which interferes with nearby electronic devices. Effects of such EMI include inappropriate device shocks, increased device battery usage, and issues with telemetry connectivity. As a majority of LVAD candidates have undergone implantation of a CIED, EMI may pose a substantial risk to this vulnerable and growing patient population.

LVAD-related EMI was first described in patients with traditional transvenous CIEDs, and more recently has been noted [3] among patients with subcutaneous implantable cardioverter defibrillators (S-ICD). The S-ICD involves a subcutaneous generator and lead system [4], an increasingly popular option for adult patients with challenging cardiac anatomy, venous access issues, and high infection risk [5]. There is yet no consensus on management of S-ICDs following LVAD implantation, and clinical practice remains heterogenous. LVAD-related EMI in patients with S-ICDs is postulated to interfere with ventricular tachycardia detection and result in inappropriate delivery of shocks from oversensing. In this work, the authors present a single-center case series of patients with S-ICDs undergoing LVAD implantation, and subsequent incidents of EMI and inappropriate device therapies, with the goal of adding to the available literature on the subject and describing a proposed approach to management.

2. Cases

2.1. Case 1

A 21-year-old man with non-compaction cardiomyopathy and worsening left ventricular ejection fraction (LVEF) was referred for consideration of advanced therapies of heart failure. After experiencing an episode of polymorphic ventricular tachycardia while hospitalized, he was referred for ICD implantation. Due to patient age and non-compaction related intracardiac thrombi, a subcutaneous system was selected. The patient underwent uncomplicated S-ICD implantation (EMBLEM Model A209, Boston Scientific, Marlborough, Massachusetts; subcutaneous electrode lead 3401) with midsternal lead positioning and successful defibrillation threshold testing (DFT) at 50 J. A shock impedance of 103 Ohms was observed during the DFT. The S-ICD was programmed with tachyarrhythmia and conditional shock zones of 220 and 200 bpm, respectively. In the follow up period the patient experienced progressive cardiomyopathy culminating in the ultimate implantation of right and left ventricular assist devices (HeartMate 3, Abbott, Chicago, IL) approximately one month later as his BMI precluded transplantation. During LVAD implantation, the substernal portion of the ICD lead was intentionally disrupted though the device was not extracted. Approximately 4 months later the patient presented to the ED after experiencing a shock from his S-ICD. Therapies were subsequently deactivated and elective removal was planned.

2.2. Case 2

A 30-year-old man was referred for non-compaction cardiomyopathy. He had previously undergone implantation of an S-ICD (EMBLEM Model A209, Boston Scientific, Marlborough, Massachusetts; subcutaneous electrode lead 3401) 3 years prior to referral after he was found to have a severely reduced ejection fraction and experienced syncope. At the time of implantation, DFT was successful at 65 J. The patient later experienced progressive cardiomyopathy with a LVEF of <10 % despite medical management, and at the time of referral was dependent on continuous inotrope infusion with NYHA class IV symptoms. Soon after referral he underwent uncomplicated LVAD (HeartMate 3, Abbott, Chicago, IL) implantation. A preoperative chest x-ray revealed left sternal S-ICD lead positioning. In the follow-up period the patient did not experience any ICD therapies and the device was found to be functioning appropriately on interrogation.

2.3. Case 3

A 59-year-old woman with nonischemic dilated cardiomyopathy underwent S-ICD (EMBLEM Model A209, Boston Scientific, Marlborough, Massachusetts; subcutaneous electrode lead 3401) implantation for primary prevention of sudden cardiac death. She was referred to our institution due to progression of her heart failure, and underwent LVAD (HeartMate 3, Abbott, Chicago, IL) implantation 21 months later. Her S-ICD demonstrated appropriate R wave sensing and normal lead impedance across all vectors prior to implantation (Fig. 1A). A preoperative chest x-ray revealed right sternal S-ICD lead positioning. Approximately 17 h after LVAD implantation the patient received 12 inappropriate S-ICD shocks while in normal sinus rhythm. The device was found to have inappropriately recorded and delivered therapies due to EMI causing ventricular oversensing. Lead impedance was unchanged. Tachycardia therapies were immediately disabled. Ventricular oversensing was noted in all vectors and at all gain settings. Tachycardia therapies were left disabled and elective S-ICD removal following the rehabilitation period was planned.

Fig. 1.

S-ICD electrograms from patients 3 (A), 4 (B), and 5 (C) prior to LVAD implantation (above) and at the time of inappropriate shocks (below).

2.4. Case 4

A 70-year-old man with non-ischemic dilated cardiomyopathy was referred to our institution for evaluation of advanced heart failure therapies. He underwent implantation of a S-ICD (EMBLEM Model A209, Boston Scientific, Marlborough, Massachusetts; subcutaneous electrode lead 3401) for sustained VT 27 months prior to referral. After evaluation, he subsequently underwent uncomplicated LVAD (HeartMate 3, Abbott, Chicago, IL) implantation. A preoperative chest x-ray revealed left sternal S-ICD lead positioning. Preprocedural device interrogation was unremarkable and demonstrated appropriate R wave sensing and lead impedance across all vectors (Fig. 1B). Nearly 11 h after LVAD implantation the patient experienced 4 S-ICD shocks while in normal sinus rhythm. On interrogation, the device shocks were noted to be inappropriate and related to oversensing from EMI. Lead impedance was unchanged. There was significant EMI noted causing ventricular oversensing in all vectors and gain settings. Device therapies were immediately disabled and elective removal was planned.

2.5. Case 5

A 73-year-old man was referred for advanced heart failure therapies evaluation for progressive non-ischemic dilated cardiomyopathy. He had undergone implantation of a S-ICD 25 months previously and had experienced sustained VT requiring an ICD shock in the intervening time. After evaluation the patient underwent uncomplicated LVAD (HeartMate 3, Abbott, Chicago, IL) implantation and recovered well. A preoperative chest x-ray revealed left sternal S-ICD lead positioning. Preprocedural device interrogation demonstrated appropriate R wave sensing and lead impedance across all vectors (Fig. 1C). Unfortunately, 1 month after his LVAD implantation the patient experienced an ICD shock which was found to be inappropriate and related to EMI oversensing. The patient's S-ICD was deactivated, and he was referred for elective extraction.

2.6. Case 6

A 24-year-old female with postpartum cardiomyopathy was referred for progressive heart failure despite maximally tolerated medical therapy. A S-ICD had been implanted 5 years previously for primary prevention. She subsequently underwent uncomplicated LVAD (HeartMate 3, Abbott, Chicago, IL) implantation. A preoperative chest x-ray revealed left sternal S-ICD lead positioning. Pre-operatively her S-ICD was deactivated and during her immediate recovery the decision was made to leave her device therapies off due to concern for EMI and risk of inappropriate shock.

3. Discussion

In this study, the authors retrospectively reviewed a single center's experience caring for patients with S-ICDs who underwent LVAD implantation. Of the 6 patients meeting these criteria, none were found to have ischemic cardiomyopathy as the etiology of their heart failure. There were 4 men and 2 women in this group and in all cases the same model ICD generator, lead, and LVAD were used. The mean age was 46 years old, and two patients (33 %) had experienced VT/VF prior to ICD implantation. Mean time between S-ICD implantation and LVAD implantation was 28 months. Inappropriate shock was experienced by 4 (66 %) of the patients, with clear documentation of EMI in 3 of the 4 patients receiving inappropriate shocks. S-ICD lead positioning varied with 4 patients (66 %) having a left sternal position of their subcutaneous lead, one midsternal and one right sternal. Inappropriate shocks were observed in patients with all 3 subcutaneous lead positions. The time between LVAD implantation and inappropriate shock varied widely in these cases, between 0 and 130 days, with a mean of 40 days from LVAD implantation to inappropriate shock. Proposed mechanisms for this include direct EMI from the LVAD as has been previously described [1,2] with transvenous systems. This may be especially notable given the physical proximity between the LVAD impeller and the S-ICD generators in situ (Fig. 2). An additional mechanism for inappropriate shocks and oversensing due to lead noise may be subcutaneous lead insulation breach during thoracotomy, given variation in subcutaneous lead placement [4]. Ultimately, 5 (83 %) of the patients in this series had S-ICD therapies permanently disabled, with one patient undergoing device extraction and 3 more planned for extraction.

Fig. 2.

AP CXR of patient 3 demonstrating physical proximity of LVAD and SICD.

4. Conclusions

Among patients with S-ICDs who underwent LVAD implantation, an extremely high incidence of EMI resulted in oversensing and inappropriate device therapies. As observed in this series, a wide variation in the period after LVAD implantation until delivery of an inappropriate shock may lead to false dismissal of EMI. Given the variation in subcutaneous sternal lead position in patients with S-ICDs, there is also an additional risk of insulation breach during sternal incision as a contributing factor to this problem. Due to the high rate of inappropriate shocks observed and the improved tolerability of malignant arrhythmias in patients with VADs, we suggest deactivation of S-ICDs prior to VAD implantation. Frequent serial evaluation including provocative maneuvers and ICD reprogramming should be considered prior to reactivation of therapies to verify the absence of EMI, and permanent deactivation with or without device removal following or during LVAD implantation may also be deliberated.

Ethics statement

The authors declare that the work described has not been published previously.

The article is not under consideration elsewhere.

The article's publication is approved by all authors and tacitly or explicitly by the responsible authorities where the work was carried out.

Id accepted, the article will not be published elsewhere in the same form, in English or in any other language, including electronically without the written consent of the copyright holder.

All authors made substantial contributions to the work including conception, drafting, critical revision, and final approval of the submitted work.

Funding

No funding was received for this research.

Ethical statement

All authors attest that this work is original and not under consideration for publication elsewhere.

CRediT authorship contribution statement

Alex J. Nusbickel: Conceptualization, Data curation, Writing – original draft, Writing – review & editing, Investigation. Stephen Allan Petty: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. Steven J. Ross: Conceptualization, Writing – original draft, Writing – review & editing. Alex Parker: Conceptualization, Writing – original draft, Writing – review & editing. Juan Vilaro: Conceptualization, Writing – original draft, Writing – review & editing. Mustafa M. Ahmed: Conceptualization, Project administration, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.