Contemporary Outcomes of Ventricular Tachycardia Ablation in Left Ventricular Assist Device Therapy: A Meta-Analysis

Abstract

Background and Objective:

Ventricular arrhythmias (VAs) remain a pervasive and deadly arrhythmia in patients with left ventricular assist devices (LVADs). Catheter ablation has emerged as a treatment option for refractory VAs, yet evidence in the era of the HeartMate 3 (HM3) remains limited. This review aims to synthesize contemporary evidence for VA ablation in LVAD recipients.

Methods:

A systematic review was performed across major electronic databases. The primary efficacy outcome was the recurrence of ventricular tachycardia (VT), and the primary safety outcome was the rate of procedural complications. The secondary outcomes were inability to induce any VT, all-cause mortality at 12 months, orthotropic heart transplantation (OHT). Sub-analyses were performed for patients with HM3 LVADs.

Results:

Twenty-seven studies encompassing 300 LVAD recipients undergoing 325 VT ablations, after a mean follow-up of 327 ± 175 days post VT ablation, VT recurred in 38% (95% CI, 28% to 49%) of cases and the complication rate was 8% (95% CI, 1.6% to 15.7%). VT was non-inducible in 61% of cases. One-year all-cause mortality was 26%, and 16% had OHT. Among HM3 recipients, electromagnetic interference (EMI) occurred in 51%, and no cases of device thrombosis were reported; one stroke was observed.

Conclusions:

Catheter ablation is a safe and feasible treatment for refractory VAs in LVAD patients as evidenced by low complication rates and reasonable acute success. Yet, the persistence of considerable VT recurrence and all-cause mortality reflects the clinical complexity of this population. Procedural challenges include mapping limitations caused by EMI, particularly in the HM3 era.

CENTRAL ILLUSTRATION 1: |

The most common VT mechanism was scar-related macro-reentry (80%), with inflow cannula circuits in 18%. Scar localization was most frequent at the apex/pericannula (30%) and septum (29%).

1 |. Introduction

The use of left ventricular assist devices (LVADs) has increased in recent years in parallel to the rising prevalence of advanced heart failure (HF). Advanced HF patients who receive an LVAD experienced improved outcomes with survival rates of 80% at 2 years [1, 2]. However, LVADs are associated with major complications, including ventricular arrhythmias (VAs), particularly among patients with a history of VAs pre-LVAD [3, 4]. Antiarrhythmic medications and implantable cardioverter-defibrillators (ICDs) are typically the first-line treatment for VAs; nonetheless, refractory arrhythmias can still occur after LVAD implantation [4, 5]. In such cases, catheter ablation has emerged as a promising treatment option for drug-refractory VAs.

In LVAD recipients, ventricular tachycardia ablation (VTA) presents specific challenges, including selecting appropriate candidates, determining the optimal timing for the procedure, and overcoming technical limitations such as those affecting mapping and ablation due to the device's magnetic field. Current evidence supporting the use of catheter ablation in LVAD patients is limited and derived primarily from high-volume, pooled analyses of single-center studies [6]. The recent clinical consensus statement from the European Heart Rhythm Association and the Heart Failure Association of the European Society of Cardiology (ESC) recommends catheter ablation for recurrent, symptomatic VAs that do not respond to antiarrhythmic drugs and/or ICD reprogramming with a moderate level of evidence [7].

A better understanding of the role of VTA in LVAD patients remains warranted. This systematic review and meta-analysis aims to consolidate current evidence on the efficacy and safety of VTA in this population, emphasizing procedural outcomes and complication rates in the era of contemporary LVAD technology.

2 |. Methods

2.1 |. Protocol and Registration

This systematic review was conducted according to the guidelines of PRISMA 2020 (Preferred Reporting Items for Systematic Reviews and Meta-Analysis [8]. The PRISMA checklist is presented in Supporting Information: Table S1. The review protocol was registered in Prospero [CRD42025630461].

2.2 |. Literature Search

The literature search was conducted by two reviewers (N.B. & M.S.) using five electronic databases (Science Direct, PubMed, Google Scholar, CENTRAL, and EMBASE). The keywords used included (Ventricular Tachyarrhythmias, Ventricular Tachycardia, Ventricular Fibrillation OR VT) AND (Catheter Ablation, Radiofrequency Ablation OR RFA) AND (Left Ventricular Assist Devices OR LVAD OR mechanical support), with slight modifications per database to maximize results. Reference lists of studies were also reviewed for additional relevant studies on LVAD recipients undergoing VTA, and the clinicaltrial.gov registry was checked for unpublished trials.

2.3 |. Eligibility Criteria and Data Extraction

Duplicate studies were removed, and the remaining studies were evaluated against several exclusion criteria: (1) not published in English; (2) no documented VT during or after LVAD implant; (3) post-implant VTA not reported as an intervention; (4) lack of outcome data; or (5) secondary analysis. Included studies comprised case reports, case series, case-control studies, cohort studies, and clinical trials. Extracted data included study author, design, sample size, age, sex distribution, HF etiology, mean LVEF, history of VT episodes, presence of ICD, prior VTA history, and VT storm.

2.4 |. Outcomes

Analysis was landmarked at the time of VTA. The primary efficacy outcome was recurrence of VT, and the primary safety outcome was rate of procedural complications. Secondary endpoints were non-inducibility of VAs at the end of ablation, successful bridging to orthotopic heart transplant (OHT), and mortality at 6 and 12 months. Clinical variables were compiled across all studies, including patient characteristics, indications for LVAD implantation, and details of VTA procedure such as approach, mapping technique, scar location, and VT mechanism. Procedure-related complications were also measured and compiled among all studies. We performed pre-specified subgroup analyses in recipients of HeartMate 3 (HM3) and in patients who underwent VTA at the time of LVAD implantation.

2.5 |. Statistical Analysis

We conducted a single-arm meta-analysis of proportions for primary and secondary endpoints, reported as event rates per 100 observations with 95% confidence intervals. Procedural techniques, scar location, VT mechanism, and the breakdown of complications were summarized as percentages. Continuous variables were presented in means with standard deviations or medians with interquartile ranges. We assessed heterogeneity using a random-effects model and the Cochrane Q test, identifying high heterogeneity with I² values over 75%. Evidence quality was evaluated with the ROBINS-1 tool for observational studies and the Joanna Briggs Institute tools for case series and reports. Publication bias was analyzed using the Egger test, considering p < 0.05 as indicative of bias. A leave-one-out sensitivity analysis was performed with the “metainf” function in the meta package, and data analyses utilized Microsoft Excel and R Studio.

3 |. Results

3.1 |. Study Eligibility

Literature search yielded 735 articles, of which 556 were duplicates. Of the remaining 179 publications, 65 were excluded based on review of the title & abstract and 114 were retrieved for detailed evaluation and application of our eligibility criteria. Ultimately, 35 articles were included as summarized in Figure 1.

FIGURE 1 |.

A PRISMA Flow diagram summarizing the search strategy. Flow diagram summarizing the systematic search process: 179 records were screened from five databases after deduplication, 35 studies were included in the final review following full-text assessment. (CENTRAL = Cochrane Central Register of Controlled Trials).

3.2 |. Study Characteristics

Table 1 summarizes the studies included in the systematic review and their characteristics. Of the 35 included studies, three were case-control studies, seventeen were retrospective case series, and fifteen were case reports. Twenty-seven studies examined patients with post-implant VTA, while eight studies focused on patients experiencing ablation during the LVAD implantation.

TABLE 1 |.

Characteristics of patients with LVAD who underwent catheter ablation for ventricular tachycardia after LVAD implantation.

Author ID	Study design	Setting	Sample size (patients)	Mean age (Years)	Number of Male/Female	Etiology		LVEF (%)	History of VAs	ICD	History of ablation	History VT storm
Moss et al., 2017 [9]	Case series	USA	21	62 ± 10	19/2	ICM	14 [67]	18 ± 5	15 [71]	21 (100)	4 (19)	6 (28)
						NICM	7 (33)
Cantillon et al., 2012 [10]	Case series	USA	21	56 ± 13	18/3	ICM	12 [57]	14 ± 6	8 (38)	11 [52]	—	—
						NICM	9 (43)
Whang et al., 2014 [11]	Case report	USA	1	70	1/0	ICM	1(100)	10– 15	—	1 (100)	0	0
Thosani et al., 2017 [12]	Case report	USA	1	72	1/0	ICM	1(100)	—	—	1 (100)	0	1(100)
Schade et al., 2014 [13]	Case report	Germany	1	66	1/0	ICM	1(100)	20	1(100)	1 (100)	0	1(100)
Komeyama et al., 2022 [14]	Case report	Japan	1	52	1/0	ICM	1(100)	—	—	1 (100)	—	1(100)
Herweg et al., 2010 [15]	Case report	USA	1	72	1/0	ICM	1 (100)	—	0	1 (100)	0	1(100)
Romano et al., 2015 [16]	Case report	Italy	1	63	1/0	ICM	1 (100)	—	0	1 (100)	0	1(100)
Herweg et al., 2012 [17]	Case series	USA	6	61.5	6/0	ICM	4 [67]	20	4 [67]	6 (100)	1 (16)	5 [83]
						NICM	2 (33)
Szegedi et al., 2015 [18]	Case report	Hungary	1	38	1/0	ICM	1 (100)	—	1	1 (100)	0	1(100)
Pedretti et al., 2017 [19]	Case report	Italy	1	58	1/0	ICM	1 (100)	—	0	0	0	0
Lü et al., 2013 [20]	Case series	USA	6	65 ± 13	5/1	ICM	4 [67]	14 ± 6	—	0	6 (100)	6(100)
						NICM	2 (33)
Snipelsky et al., 2016 [21]	Case series	USA	6	64.5	4/2	ICM	3 (50)	21.2	3 (50)	6 (100)	0	0
						NICM	3 (50)
Osaki et al., 2008 [22]	Case report	USA	1	65	1/0	ICM	1(100)	10	1(100)	1 (100)	0	0
Hottigoudar et al., 2011[23]	Case series	Canada	3	61	3/0	ICM	1 (33)	10–15	2 [67]	2 [67]	1 (33)	0
						NICM	2 [67]
Dandamudi et al., 2007 [24]	Case series	USA	3	55	3/0	ICM	1 (33)	20	1 (33)	2 [67]	1 (33)	1 (33)
						NICM	2 [67]
Garan et al., 2014 [25]	Case series	USA	7	66	7/0	ICM	5 [71]	—	7 (100)	7 (100)	5 [71]	1 (14)
						NICM	2 (29)
Sacher et al., 2015 [26]	Case series	France	34	58 ± 10	30/4	ICM	21[62]	17 ±5	27 [79]	32 [94]	5(15)	5 (15)
						NICM	13 (28)
Bergau et al., 2021[27]	Case series	Germany	9	60 ± 8	9/0	ICM	4 (44)	22 ± 6	7 [77]	9 (100)	4 (44)	9 (100)
						NICM	5 [56]
Izumida et al., 2023 [28]	Case report	Japan	1	53	0/1	NICM	1(100)	—	1(100)	1 (100)	0	0
Brack et al., 2024 [29]	Case control study	Germany	69	61 ± 8	43/6	ICM	41[59]	21 ± 6	44 [64]	63 [91]	—	47 [68]
						NICM	28 (41)
Niknam et al., 2024 [30]	Case series	USA	7	64 ± 6	—	ICM		—	—	7 (100)	—	—
						NICM	6 [86]
						Combined	1(14)
Grinstein et al., 2020 [31]	Case-control study	USA	43	63 ± 10	39/4	ICM	28 [65]	—	29 [67]	31 [72]	3 (7)	27 [63]
						NICM	15 (32)
Oates et al., 2022 [32]	Case report	USA	1	74	1/0	NICM	1(100)	—	1 (100)	1 (100)	1 (100)	1 (100)
Wang et al., 2023 [33]	Case report	USA	1	59	1/0	ICM	1(100)	20	1 (100)	1 (100)	—	—
Nof et al., 2022 [34]	Case series	France	19	65 ± 8	18/1	ICM	15 [79]	17 ± 5	17 [89] VT	—	8 (42)	10 [53]
						NICM	4 (21)		4 VF (21)
Lynch et al., 2024 [35]	Case control study	USA	34	62 ± 9	32/2	ICM	23 [68]	—	24 [71]	28 [83]	10 (29)	—
						NICM	11 (32)

Note: Values are mean ± SD or n (%).

Abbreviations: ICM = ischemic cardiomyopathy, LVEF = left ventricular ejection fraction, NICM = nonischemic cardiomyopathy.

Among the 300 patients with post-implant VTA ablation, the mean age was 61.8 ± 7.1 years, and 82% were male. Etiology of HF was ischemic cardiomyopathy (ICM) in 62% and nonischemic cardiomyopathy (NICM) in 38%, and one patient (0.01%) had both. Left ventricular ejection fraction ranged from 10% to 22%, and 82.1% of patients had an ICD; 76%had a history of VAs before LVAD implantation. Among these, 16% had undergone a VTA before implantation, 45% had prior VT storm, and 8% underwent redoablation for VT.

3.3 |. Quality Assessment

We found that three non-randomized studies had a moderate risk of bias, while only the studies from the case series and case report had an intermediate risk of bias, as described in Supporting Information S1: Tables 2A, 2B, and 2C.

The leave-one-out sensitivity analysis demonstrated that no individual study significantly altered the pooled estimate of all outcomes, indicating the robustness of the results—Supporting Information S1: Figure 1. When assessing publication bias, our funnel plot exhibited a symmetrical distribution, and the Egger's test result was nonsignificant (p > 0.05) for all outcomes, indicating no evidence of publication bias, as shown in Supporting Information S1: Figure 2.

3.4 |. Characteristics of LVAD Therapies in the Included Studies

A total of 285 individuals were reported with information and details regarding varying types of LVADs. The characteristics of the LVAD therapies are presented in Supporting Information S1: Table 3. The predominant devices utilized in this study included the HeartMate II (HMII), accounting for 47% of cases, followed by the HeartWare ventricular assist device (HVAD), which comprised 24%. The HM3 also accounted for 24% of the devices used. Older LVAD models (5%) were considered obsolete and therefore not included in the analysis. Additionally, the device type was not reported in 15 patients. LVADs were used as a bridge to transplant therapy in 46.9% and as destination therapy in 53.1% cases. The time interval between LVAD implantation and VT ablation was highly variable across studies, with a median of 181 days (IQR: 66–365), reflecting both early and late arrhythmic presentations following LVAD placement.

3.5 |. Ventricular Tachycardia Ablation Characteristics

Among patients undergoing VTA, 75% underwent a transseptal approach, 19% a retrograde aortic approach, 1% an epicardial approach, and 5.2% a combined trans-septal and retrograde approach (Supporting Information S1: Figure 3A). Entrainment or activation mapping was the most implemented technique for mapping, reported in 36.1% of studies, while substrate and pace mapping were each used in 19.4% of studies. Coherent mapping (vendor-specific activation mapping algorithm, Abbbot Ensite Precision) was reported in a single study, representing 3% of studies, while in 22.2% of studies, multimodal approaches were described, most often substrate mapping combined with activation or entrainment (Supporting Information S1: Figure 3B).

A total of 540 VAs were mapped, with a mean of 2 ± 1 VT per patient. The most common mechanism was scar-related macro-reentrant, observed in 80% of cases. Inflow cannula–related reentry occurred in 18%, bundle branch reentry in 3%, and focal micro–reentry in 2%. The most frequent scar locations were the septum (30%), the apex (30%), the anterior wall (27%), and the lateral wall (22%). A summary is provided in the Central Illustration 1.

3.6 |. Outcomes

The mean duration of follow-up for the case series was 327 ± 175 days, whereas for the case reports, it was 196 ± 157 days.

3.6.1 |. Primary Efficacy Endpoint: VT Recurrence

The primary endpoint for VT recurrence in the pooled analysis of case series and retrospective studies was 38% (95% CI, 28% to 49%), with moderate heterogeneity observed across studies (I² = 59%) (Figure 2A), and 21 patients underwent redo-ablation. Among the 12 patients described in case reports, VT recurred in 5 individuals, and 4 underwent redo ablation; among these, 2 underwent mini-thoracotomy for epicardial ablation and one had VT originating from a canula scar. Notably, in one case, VT recurred after the first ablation and resolved thereafter; this patient underwent orthotopic heart OHT before the planned second ablation.

FIGURE 2 |.

(A and B) Primary Efficacy and Safety Outcome After VTA in LVAD recipients. (A-B) In patients with LVADs, VT recurrence was 38% after catheter ablation (95% CI: 28%–49%; I² = 59%), showing a high recurrence rate. The pooled complication rate from 309 procedures was 7.4%, indicating low procedural risk for LVAD patients undergoing VTA.

3.6.2 | Primary Safety Endpoint: Procedural Complication

A total of 325 procedures were performed: among the 309 procedures in the case series, 45 of 309 (14%) experienced procedure-related complications, yielding an overall pooled event rate of 7.4% (95% CI, 1.6% to 15.7%; I²:71%) (Figure 2B). Procedural complications were reported in 5 of 16 procedures from case reports (31%). Major complications occurred in 38 of 325 procedures (12%), whereas minor complications were observed in approximately 12 (4%) cases. The most frequently reported major complication was stroke or embolism, which occurred in 19 patients (6%), followed by device thrombosis in 10 patients (3%). Severe hypoxemia, attributed to an atrial septal defect, was reported in 3 patients (1%). Cardiogenic shock and volume overload each occurred in 1 patient (0.3%). There was one procedural-related death (0.3%) in which the patient experienced post-procedure hypoxemia requiring intubation and later died from hypoxic brain injury. No catheter entrapment was reported. Minor complications included groin hematoma in 5 patients (1.5%) and access site complications in 6 patients (2%). Table 2 provides a detailed breakdown of procedural complications.

TABLE 2 |.

Procedural complications of VT catheter ablation in the case series and case reports.

Classification of complications	Types	Number (%)
Minor	Groin hematoma [9, 13, 25, 26, 35]	5 (1.5)
	Access site complications [13, 29]	6 (2)
	Pericardial effusion[29]	1 (0.3)
Major	Groin pseudoaneurysm/fistula requiring surgical repair [10]	2 (1)
	Cardiogenic shock [26]	1 (0.3)
	Stroke/embolism [26, 31, 34]	19 (6)
	Hypoxia /Hypoxemia [14, 32, 33]	3 (1)
	Volume overload [23]	1 (0.3)
	Device Thrombosis [31]	10 (3)
	Procedure Related Death [29]	1 (0.3)
	Aortic Dissection [29]	1 (0.3)

3.6.3 | Secondary Outcomes

3.6.3.1 | No Inducibility of Any VT After VTA.

Acute non-inducibility of VT was achieved in 61.5% (95% CI [47.5%, 74.8%]) of patients in the case series (Figure 3A). Outcomes had moderate heterogeneity between studies (I² = 73.1%). In the case reports, clinical VT was rendered non-inducible in 11 out of the 12 patients.

FIGURE 3 |.

(A and B) No inducibility of VT and bridge-to-OHT after VTA in LVAD. (A) Acute outcomes of VTA showed 62% achieved complete non-inducibility of arrhythmias, indicating procedural success. (B) Additionally, 17% of patients t moved on to heart transplantation.

3.6.3.2 |. Heart Transplant After VTA.

The rate of successful bridging to OHT was 17% (95% CI [6.2%, 30%]). The outcome had moderate heterogeneity across the studies, I² = 72.3% (Figure 3B). In the case reports, 3 of 12 patients received OHT during the follow-up period.

3.6.3.3 |. Mortality Outcomes.

Aggregate mortality rate combined among case series and observational studies was 12% (95% CI [3.6%, 23.1%] I²:52%) at 6 months after ablation and increased to 26% (95% CI [13.5%, 40.7%] I² :68%) at 12 months (Figure 4A,B). Among the 13 patients included in the case reports, only one had died within 1 year of follow-up. The common causes of death included advanced HF, infectious complications, multiorgan failure, and stroke. A detailed summary of the causes of death is presented in Table 3.

FIGURE 4 |.

Mortality following VTA in patients with LVADs: 6- and 12-Month. (A-B) The pooled mortality rate was 13.3% at 6 months and 26% at 1 year, based on data from case series and observational studies.

TABLE 3 |.

Breakdown of mortality outcomes at 6 and 12 months after VT catheter ablation.

	Mortality
Author ID	6 months	12 Months	Cause of death
Bruck et al., 2024 [29]	9	26	Heart failure, intracranial bleeding, sepsis, procedure related hypoxemia.
Hottigoudar et al., 2011 [23]	1	1	Multiorgan failure.
Garan et al., 2014 [25]	—	3	Device-related death, RV failure, and Multiorgan failure.
Niknam et al., 2024 [30]	—	2	NR
Bergau et al., 2021 [27]	1	2	Septic shock, paralytic ileus and Severe pneumonia.
Lü et al., 2013 [20]	1	—	End-stage HF
Moss et al., 2017 [9]	6	10	Hemolysis, infection, multiorgan failure, LVAD pump failure, and intracranial hemorrhage.
Herweg et al., 2012 [17]	3	5	RV failure, stroke, pump pocket failure, emphysema, and pulmonary embolism.
Dandamudi et al., 2007 [24]	1	1	Infection
Romano et al., 2015 [16]	—	1	Sepsis and ischemic stroke (device thrombosis)
Sacher et al., 2015 [26]	—	10	End-stage HF, septic shock, LVAD deactivation, Intracranial hemorrhage, sudden death, Hemolysis, Cable failure/redo surgery.
Snipelisky et al., 2016 [21]	—	1	NR
Lynch et al., 2024 [35]	4	6	NR

3.7 |. HeartMate 3 Outcomes

We identified 69 patients with a HM3 who underwent VTA. Only one study specifically focused on HM3 recipients, while the other two were observational studies that included data on various devices. Notably, four out of six studies reported technical limitations during the ablation procedures, which ranged from temporary visualization loss of the catheter at some point during the procedure to incomplete mapping and even severe electromagnetic interference (EMI). However, no procedures were aborted due to EMI. To mitigate EMI, approaches included using CARTO patches placed away from the LVAD and the complementary use of intracardiac echocardiography. Table 4 presents a summary of the HM3 included studies.

TABLE 4 |.

Summary of VTA procedural complications in patients supported with HeartMate 3.

Author name	Sample size (HM3 patients)	Procedure related complications	Device-related issues or technical limitations	Approach to EMI
Bruck et al., 2024 [29]	42	Stroke: 0	12 EMI mild Transient loss of visualization catheter	No LVAD flow adaptation performed
		DT: 0
		†ARF: 1
Bergau et al., 2021 [27]	5	Stroke: 0	Minor EMI 1 Difficult trans-aortic approach due to closed aortic valve	Temporal LVAD pump stop
		DT: 0
		—
Oates et al., 2022 [32]	1	Stroke: 0	1 Significant EMI	NR
		DT: 0
		Iatrogenic ASD with R-L Shunting
Nof et al., 2022 [34]	19	Stroke: 1	Significant EMI in all cases. No abortion due to EMI Reported	– CARTO system patches distribution – LVAD Battery removal – CARTO system low pass filter was lowered – Intracardiac signals as reference – Adjust the ECG to the 40–20 Hz range.
		DT:0
		CAVB
Wang et al., 2023 [33]	1	Stroke: 0	NR	—
		DT:0
		ASD with R-L Shutting
Lynch et al., 2024 [35]	1	NR	NR	—

Abbreviations: ARF = acute respiratory failure, ASD = atrial septal defect, CAVB = complete atrioventricular block, EMI = electromagnetic interference.

HeartMate 3 was associated with low postablation complication rates; only one patient experienced a stroke that was considered secondary to profound hypoxemia caused by an iatrogenic atrial septal defect. No cases of device thrombosis or catheter entrapment were reported.

3.8 |. Characteristics of VTA Performed Concurrently with LVAD Implantation

Table 5 summarizes the results of the included studies. Across five case series and 4 case reports, 29 patients underwent concomitant VTA during LVAD implantation. In these patients, there was nearly a universal history of VA before implantation (93%). Most individuals had an ICD at the time of surgery (79%), while a minority of patients had a prior ablation (38%). Six patients (32%) experienced VT recurrence after LVAD implantation. The most common complication during follow-up was sepsis (14%). Device thrombosis occurred in 2 cases; the recipients had a HM II, and one case of severe lateral wall edema resulting in partial inflow cannula obstruction was reported. Only one patient who underwent both LVAD implantation and VTA died within the 30-day postoperative period.

TABLE 5 |.

Catheter ablation of ventricular arrhythmias during left ventricular assist device implantation.

Author name	Sample	Follow updays	LVAD type	History of VA	ICD before LVAD	Prior Ablation	VTrecurrence	Preoperative condition	Mapping details	Complication# patients
Patel et al, 2016 [36]	5	363 ± 368	HMII: 2[30] Levitronix: 1(20) HM XVE: 2(40)	5 (100)	5 [80]	4 [80]	2 (40)	NR	Hybrid or during LVAD implant; epicardial mapping with EnSite (St. Jude); catheter-based ablation with Chilli™; substrate/activation mapping performed; no entrainment/pace mapping; 1 case Thermocool™ endocardial; 1 case Atricure™ surgical ablation tool	Hemorrhagic stroke 99 days after ablation :1 sepsis, multiorgan failure at month six months :1
Moss et al., 2019 [37]	2	311 (168–469)	NR	2 (100)	2 (100)	NR	1 (50)	NR	Epicardial high-density voltage mapping (EnSite NAVX, Abbott) performed intraoperatively before LVAD implant; duodecapolar catheters used; fractionated/late potentials identified; no routine VT induction/entrainment; epicardial cryoablation performed in 5 cases; no endocardial mapping/ablation	Stroke: 1
Mulloy et al., 2022 [38]	7	144.9 ± 76.6	HMII: 7 (100)	7 (100)	6 [86]	3 (43)	2 (29)	IABP: 2 Intubated before vad: 0	Intraoperative epicardial + endocardial cryoablation during LVAD implant (AtriCure Cryol probe); ablation guided by pre-op EP mapping (4 pts) or ECG review (3 pts); no intra-op mapping performed; lesions at −70°C for 2–2.5 min; cryo tracts extended to mitral annulus or cannula to prevent reentry	Cardiac Tamponade requiring intervention :1 Prolonged ventilation :3 Gastrointestinal bleeding :1
Tankut et al., 2022 [39]	10	360	HMII 4 (40) HM31 6 [60]	8 [80]	8 [80]	1 (100)	NR	3 (30) Mechanical Support 5 (50) Inotropes	Intraoperative endocardial (50%), epicardial (20%), or combined (30%) ablation during LVAD implant; radiofrequency or cryo used; substrate homogenization of scar targeting isthmus and exit sites; mapping system not specified; thoracotomy in 5 pts	Sepsis :3
Nishino et al., 2024 [40]	1	1080	HM3 (100)	1 (100)	1 (100)	1 (100%)	0	1 (100) Inotrope	Intraoperative epicardial RF ablation performed before LVAD implant; CARTO (Biosense Webster) used for 3D EAM; extensive low-voltage substrate mapped (posterior/inferior/apical walls); late/delayed potentials ablated with ThermoCool SmartTouch SF; 10–35 W, 15–30 s per lesion; endpoint = loss of capture/impedance drop; no post-induction performed	NR
Rao et al., 2020 [41]	1	NR	HVAD: 1	1 (100)	1 (100)	1 (100)	1 (100)	Treated with stellate ganglion block 1 (100) Inotrope	Intraoperative surgical cryoablation with CryoFlex (Medtronic); 4 applications × 2 min; targets = basal lateral LV epicardial VT exit + inferior wall (guided by prior endocardial ablation history); no intra-op mapping performed; ablation performed while on cardiopulmonary bypass during HVAD implant	Severe lateral wall edema causing partial inflow cannula obstruction
Orozco et al., 2020 [42]	1	NR	HM3: 1	1 (100)	1 (100)	1 (100)	1 (100)	IABP: 1	Intraoperative endocardial cryoablation performed on arrested heart during HM3 implant; apical ventriculotomy approach; lesion tracts extended to mitral valve to prevent reentry; no intra- or post-op EP mapping performed	NR
Mcllvenan et al., 2016 [43]	2	11 ± 2	HM2:2	2 (100)	NR	NR	NR	IABP :1 ECMO :1	Intraoperative open epicardial + endocardial surgical cryoablation before LVAD implant; pace activation mapping of LV used to identify VT focus; cryo lesions targeted VT site + ablation lines across scar; no electroanatomic system used	Device Thrombosis: 2 Both required device exchange

Abbreviations: ECMO = extracorporeal membrane oxygenation, IABP = intra-aortic balloon pump, NR = not reported.

4 |. Discussion

This meta-analysis summarizes the cumulative experiences of high-volume tertiary centers with VTA in LVAD recipients. Our pooled analysis provided several key clinical insights into the management of this population:

1. Despite the technical challenges of performing VTA, we identified a low rate of procedural complications and acceptable rates of procedural success and of VT recurrence during follow-up.

2. Ischemic cardiomyopathy was the most common underlying VT substrate. The predominant arrhythmogenic mechanism were scar-related macro–reentrant circuits, most frequently located in the septum or apical wall.

3. In contrast to older LVAD generations, no cases of thromboembolism were reported following VTA in the sixty-nine patients with HM3 LVADs. Only one case of postablation stroke was documented in a subgroup analysis, suggesting a favorable safety profile of the HM3 during VTA.

4. In patients with HM3 LVAD, mapping was hampered by EMI in 51% of cases, particularly when mapping near the inflow cannula. Despite this unique procedural challenge, no catheter entrapment or procedural abortion was reported, and the use of intracardiac echocardiography mitigated the impacts of EMI.

5. Despite high clinical complexity, those who underwent VTA during LVAD implantation had a low complication rate, although there was a notable recurrence rate of 32%.

4.1 |. Study Cohort and Ventricular Arrhythmia Characteristics

Several mechanisms contribute to the high incidence of ventricular arrhythmias in this population, including increased sympathetic tone, fluid and electrolyte disturbances, perioperative use of inotropes, transient alterations in cardiac repolarization, cannula suction events, and scar formation at the ventricular apex cannulation site, creating a substrate for reentrant VT [44, 45]. Consistent with prior reviews, our cohort was predominantly older males with an ischemic substrate, reflecting the population of LVAD recipients established to have the highest risk for developing ventricular arrhythmias [6]. We were surprised to find relatively low rates of VTA performed for the treatment of inflow cannula-related reentry VA, which accounted for only 18% of VTA cases. Interestingly, most arrhythmogenic scars were instead located in the anterior wall and septum, which corroborates recent evidence that a septal origin of VT may be associated with the highest risk of VT recurrence and development of intractable VT [34]. While we found that entrainment mapping remained the predominant technique for VT characterization during electrophysiology study, we noted a trend towards increased adoption of substrate mapping, pace mapping, or a combined approach when compared with earlier analysis, reflecting the complexity of management of VT in patients with numerous VT circuits [6].

4.2 |. Efficacy and Safety of VTA

As ventricular arrhythmias in LVAD patients are often well tolerated and sudden arrhythmic death is uncommon, the justification for VTA must be guided by a favorable risk–benefit profile, with particular attention to minimizing procedural harm. Consistent with prior systematic reviews [6], we observed a low complication rate, with only one death in the included studies that was directly related to procedural complications, which we feel is an appropriate level of risk considering the high burden of comorbidities in this patient population. Stroke emerged as the most common reported major complication; a single stroke, related to systemic hypoxia, occurred in a patient with HM3, while all other stroke events occurred in patients with earlier generation LVADs. We attribute this favorable safety profile in such a high-risk patient cohort to diligent multidisciplinary preprocedural planning, high operator expertise, significant advancements in mapping technology, and the introduction of the HM3, which has been associated with a major reduction in thromboembolic adverse events.

4.3 |. Contemporary Treatment of VA in LVAD Patients

The HM3 is distinct in its use of a magnetically levitated centrifugal pump, which reduces shear stress and the incidence of hemocompatibility-related complications. There was no reported stroke due to thromboembolism and device thrombosis following VTA in HM3 recipients, reaffirming that VTA is both safe and appropriate in this population.

Strategies to reduce VAs in LVAD patients, especially those awaiting transplant, remain under debate. In particular, the optimal timing of VTA—before, during, or after LVAD implantation—remains unclear. VTA before implantation offers has certain advantages, but may not account for VT that may originate from the cannula insertion site [46, 47]. Ongoing trials like CASTLE-VT will evaluate the efficacy and safety of this approach [48].

Intraoperative VTA during LVAD implantation remains an appealing strategy to lower post-implant arrhythmia burden as well. In particular, intraoperative VTA is appealing due to its direct access to epicardial tissue, which is of particular relevance in LVAD recipients as they have high levels of epicardial low-voltage zones, underscoring the critical role of epicardial substrate in this population. Still, intraoperative ablation presents challenges, including catheter tracking and impedance issues due to open-air exposure and prolonged surgical time [36, 49]. In our cohort, patients who underwent intraoperative VTA had a recurrence rate of 32%, though the procedure remained safe with no deaths linked to ablation. One of the studies included in our review indicated that this method is both safe and effective, with a median mapping time of just 12 min to identify arrhythmogenic areas, although the level of evidence remains low [37]. We look forward to the results of the ongoing PIVATAL trial, which aims to explore the role of a combined approach in reducing VA burden [50].

4.4 |. Limitations

This review has several limitations. First, all included studies were retrospective, introducing risks of selection bias, unmeasured confounders, and inconsistent outcomes reporting. Reporting of all-cause mortality during follow-up was variable; while overall rates were substantial and expected in this high-risk population, limited data precluded consistent subgroup analysis. Although no publication bias was detected, case reports and series describing complications are more likely to be published. Second, considerable heterogeneity was noted across studies, which may have been influenced by patient populations and procedural characteristics. In our primary efficacy outcome, postablation antiarrhythmic drug use was inconsistently reported and could not be systematically assessed. Given its potential effect on procedural outcomes, this also represents an essential source of heterogeneity. Moreover, variability in the reporting of procedural limitations—such as catheter visualization issues and EMI—may also have introduced bias into our analysis. Finally, nearly half of the included LVADs were older models, and only one study focused exclusively on HM3, limiting generalizability.

5 |. Conclusions

This systematic review summarizes the experience of multiple institutions with VTA in LVAD recipients, with a particular focus on the impact of the contemporary HM3 on VTA outcomes. Our findings indicate that VTA is a feasible and acceptably safe treatment strategy for managing refractory VAs, particularly in patients supported with the contemporary HeartMate 3. However, further research is needed to better understand the appropriate indications and optimal timing of VTA in LVAD recipients.

Supplementary Material

Supporting Information

Additional supporting information can be found online in the Supporting Information section.

Supplementary Appendix Final - Clean.

Abbreviatures:

EMI

electromagnetic interference

HF

heart failure

ICD

implantable cardioverter-defibrillator

LVAD

left ventricular assist device

OHT

orthotopic heart transplantation

VAs

ventricular arrhythmias

VT

, ventricular tachycardia

VTA

ventricular tachycardia ablation.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.