ABSTRACT
Introduction
Implantable cardioverter-defibrillators (ICD) terminate potentially deadly ventricular arrhythmias in at-risk patients. Retrosternal extravascular ICDs (EV-ICD) place a lead in the mediastinal space to provide both shock and antitachycardia pacing therapy without the need for venous access. A systematic review was performed to assess acute outcomes in patients implanted with a retrosternal EV-ICD.
Methods
A query of EMBASE and MEDLINE databases was performed to identify unique retrosternal EV-ICD cohort studies published between JAN-1–2020 and JAN-31–2026. Two known retrosternal EV-ICDs were included, the market-approved Medtronic substernal EV-ICD system and investigational AtaCor intercostal EV-ICD lead. Procedure and defibrillation testing details and acute safety data were collected.
Results
Fourteen articles (n = 806 patients) met criteria for inclusion. Lead placement success was 99.6% and 97.3% while defibrillation testing success was 98.5% and 97.2% for the substernal EV-ICD system and intercostal EV-ICD lead, respectively. Periprocedural complications were manageable with a rate of 1.5% (n = 11/731) for the substernal EV-ICD system and 8.2% (n = 6/73) for the intercostal EV-ICD lead.
Conclusions
The available evidence suggests that periprocedural outcomes and acute safety of retrosternal EV-ICD systems, particularly the substernal EV-ICD system, is an effective and safe strategy that provide potentially lifesaving therapy while preserving the vasculature.
Protocol registration: www.crd.york.ac.uk/prospero identifier is CRD420261328371.
KEYWORDS: EV-ICD, sudden cardiac death, systematic review, retrosternal ICD, defibrillation testing
Plain Language Summary
Implantable cardioverter-defibrillators (ICD) are devices designed to prevent sudden cardiac death by delivering an electric shock to correct a dangerous, irregular heartbeat in the in the lower chambers of the heart. Traditional ICDs, called transvenous ICDs, are implanted under the skin in the upper chest and connected to wires, called leads, that run through the veins to the heart. These leads can cause complications. To address this, extravascular ICDs (EV-ICDs) were recently developed. EV-ICDs have leads placed outside the veins and the heart. This systematic review describes short-term and procedure outcomes from two different types of EV-ICDs that place the lead in a novel location under the breastbone. This review identified 14 studies that reported on 806 patients implanted with one of the two EV-ICDs. Overall, the studies showed the devices were successfully implanted in most patients, and that the number of complications during or soon after the procedure was low. The studies also showed that defibrillation testing (ensuring the device can terminate a life-threatening irregular heartbeat with an electrical shock) was successful in most patients. This evidence shows short-term safety of EV-ICDS and that they can provide lifesaving therapy without the risk of leads in the veins and heart.
1. Introduction
For decades, implantable cardioverter-defibrillators (ICD) have provided protection from sudden cardiac death by using intravascular leads into the heart to detect and terminate sustained ventricular arrhythmias [1]. Advancements in transvenous ICD (TV-ICD) technology have improved outcomes by refining both the design and diagnostic capabilities of the system [2]. Despite these improvements, transvenous ICDs are still limited by the need for vascular access, which can result in complications including vascular injury, sepsis, thrombosis, worsening tricuspid regurgitation, and those from high-risk lead extraction procedures [3]. To address this, non-transvenous systems have been developed to avoid the vasculature and the associated risks.
The first non-transvenous ICD was the subcutaneous ICD (S-ICD) (Boston Scientific), approved by the United States Food and Drug Administration (US FDA) in 2012, with a lead placed under skin adjacent to the sternum to provide sensing and shock therapy but without the ability to provide antitachycardia pacing (ATP) [4,5]. More recently, the Medtronic substernal extravascular ICD (EV-ICD) system, hereafter referred to as the substernal EV-ICD system, received regulatory approval across several regions including US FDA approval in 2023 [6,7]. The substernal EV-ICD system involves placing a lead under the sternum allowing for both shock and ATP therapy. The AtaCor intercostal EV-ICD lead, hereafter referred to as the intercostal EV-ICD lead, is a second retrosternal EV-ICD that is currently in early pre-market studies and uses a lead that is delivered through the intercostal structures of the ribcage into the mediastinal space and meant to be compatible with pulse generators from other manufacturers [8]. While a systematic review of available S-ICD human clinical data has been performed [9–12], to our knowledge, no such review has been performed for retrosternal EV-ICDs. Owing to the relative novelty and unique implant location of the substernal EV-ICD system and intercostal EV-ICD lead, a systematic review of the literature was performed to evaluate periprocedural outcomes and acute safety of retrosternal EV-ICD systems.
2. Methods
2.1. Search strategy
A systematic review was performed by searching the EMBASE and MEDLINE databases using key terms related to retrosternal EV-ICDs (Supplemental Table S1). The date range for the search query was 1 January 2020 (first substernal EV-ICD publication published in 2020) to 31 January 2026. An initial screen of captured articles was performed (AS) to exclude those that were either unrelated, pre-clinical animal studies, reviews, or single-case studies (Supplemental Table S2). Studies that reported data from unique cohorts of at least five patients, included procedure and/or defibrillation testing details, and indicated implantation of a retrosternal EV-ICD system were selected. Articles that appear to meet the inclusion criteria, as they may indicate implantation of a retrosternal EV-ICD system, but were excluded from the systematic review, are provided in Supplemental Table S3 along with rationale for exclusion. Separately, a manual search of PubMed Central was performed (TRH) to identify any studies that may have been missed by the automated query. One article identified during the manual search required translation to English which was performed using Google translate. Two articles reported data from the same study (Continued Access Study) [13,14], but only unique datapoints (those reported in one study but not the other) were included to prevent reporting of overlapping data. Ethical approval was not required as all data used was anonymous and taken directly from the published literature. This systematic review was prospectively registered on PROSPERO (ID: CRD420261328371) and was designed using the Oxford Center for Evidence Based Medicine guidance for systematic review of cohort studies [15]. There were no major deviations from the protocol registered on PROSPERO.
2.2. Datapoints reported
The following data from each study was collected as available: baseline patient characteristics, implant procedure success (defined as completing the procedure with a system still in place), lead placement success, defibrillation testing success (at any energy, or per the study protocol), energy at which defibrillation testing was successful, R-wave amplitude at time of implant (ring 1 to ring 2 for the substernal EV-ICD system and dedicated bipolar for the intercostal EV-ICD lead), pacing capture threshold and lead impedance at time of implant (ring 1 to coil 2 for the substernal EV-ICD system and integrated bipolar for the intercostal EV-ICD lead), total procedure time, time to lead placement (time from first incision to final lead placement including lead suturing where noted), total fluoroscopy time, acute major complications (e.g., pneumothorax, cardiac perforation, infection, lead migration, pleural lead placement), acute system revisions, and any troubleshooting performed to improve the outcome of the procedure or defibrillation testing.
2.3. Assessment of bias and quality
Two independent reviewers (AS, TRH) assessed each study for bias and quality. Only outcomes from single-arm studies were assessed, so bias was evaluated using a modified version of the Ottawa-Newcastle questionnaire to exclude questions related to the bias of comparing one treatment to another [16]. A maximum of six stars could be awarded to each study, three for cohort bias and three for outcome bias, with six stars indicating the lowest risk of bias. To assess quality, the Joanna Briggs Institute (JBI) critical appraisal tool was used [17]. Out of 10 questions, a study was deemed high quality for eight or more ‘yes’ responses, medium quality for five to seven ‘yes’ responses, and low quality for less than five ‘yes’ responses. When required, the two reviewers discussed both their bias and quality assessments to resolve any differences and come to a consensus opinion.
2.4. Description of retrosternal EV-ICDs
The substernal EV-ICD system (Medtronic, Minneapolis, MN, USA) is a market-approved ICD system that deploys an epsilon-shaped lead with two pace/sensing electrodes and two defibrillation coils that all connect to a dedicated pulse generator (33 cm3 volume) [6]. The substernal EV-ICD has three sensing configurations, one near-field (ring 1 to ring 2) and two far-field (ring 1 to can and ring 2 to can) vectors. The lead is placed under the sternum via a subxiphoid incision and subsequent tunneling into the mediastinal space using dedicated tunneling and sheath delivery tools. Lateral tunneling is then performed to connect the lead to the pulse generator. The substernal EV-ICD lead is typically placed just under or to the left margin of the sternum over the apex of the heart in a position to optimize R-wave amplitude and eliminate P-wave sensing, while the generator is placed in a left mid-axillary position. The substernal EV-ICD system can provide shock, ATP, pause prevention, and post-shock pacing therapy. The market-approved Aurora EV-ICDTM system (Model DVEA3E4) differs from those used in pre-market studies (Models DVEX2E4 or DVEX3E4) due to the addition of a novel algorithm intended to reduce inappropriate detections due to P-wave oversensing (Smart Sense).
The intercostal EV-ICD lead (AtaCor Medical, San Clemente, CA, USA) is an investigational ICD lead that is designed to be compatible with pulse generators from multiple manufacturers via a DF-4 connector. Two pacing and sensing electrodes are present on the lead as well as a high energy coil that is divided across two symmetrical arms of the lead. Three pace/sense vectors can be used, including two near field vectors (dedicated bipolar and integrated bipolar) and a far field vector. The lead is implanted into the anterior mediastinum through intercostal tissue adjacent to the left sternal margin using a dedicated delivery tool and then connected to any DF-4 compatible pulse generator in either the left pectoral or left mid-axillary position [8]. Strict training protocols are in place for both retrosternal EV-ICDs.
2.5. Statistical methods
Data was extracted for each desired endpoint and collected in a spreadsheet by three independent reviewers to ensure data accuracy (AS, TRH, BLR). Continuous variables were reported as either the mean plus or minus the standard deviation and/or median with interquartile ranges (where reported). Categorical variables were reported as the frequency and percentage. A weighted rate or mean value, to account for cohort sizes, was performed for select endpoints using the SUMPRODUCT/SUM equation in Excel (Microsoft, Redmond, WA, USA) to give a rough estimate for certain datapoint across multiple studies. Datapoints not reported in each study were designated as not reported (NR), except for major complications where they were designated as not occurring (0%) if sufficient follow-up time was available for the event to occur.
3. Results
3.1. Study selection and assessment
Out of 105 studies identified during the search process, 14 met the inclusion criteria, including three articles that were identified during the manual search of citations (Figure 1) [6,8,13,14,18–27]. Nine (64.3%) of the studies were considered high quality and all studies had either a moderate (3, 21.4%) or low (11, 78.6%) risk of bias. Eleven articles presented results from 11 independent cohort studies with the substernal EV-ICD system, which included three prospective multicenter studies, one prospective multicenter registry, six retrospective single center studies, and one prospective single center study [6,13,14,18–21,23–25,27]. As for the intercostal EV-ICD lead, three articles reported data from four studies, including two prospective single center feasibility studies, one prospective single center study, and one prospective multicenter study (Table 1) [8,22,26].
Figure 1.
Prisma flow diagram for retrosternal EV-ICD systematic literature review.
PRISMA flow diagram that details the identification methods of studies, screening, and studies included.
Table 1.
Overview of selected studies.
| Study reference | Study information | EV-ICD manufacturer and brand name | Study timeframe | Follow-up duration | Quality assessment | Risk of bias† |
|---|---|---|---|---|---|---|
| Medtronic EV-ICD system (substernal EV-ICD system) | ||||||
| Biffi et al. [18] | Retrospective, single-center case series | Medtronic Aurora EV-ICD™ system (Model DVEA3E4) | May 2024 to March 2025 | 9 ± 6 months | Medium | **** |
| Biffi et al. [19] | Retrospective, single-center study | Medtronic Aurora EV-ICD™ system (Model DVEA3E4) | November 2020 to September 2026 | 24 ± 19 months | Medium | **** |
| Billakanty et al. (Subgroup of Continued Access Study) [13] | EV-ICD Continued Access Study sub-cohort: Prospective, global, single-arm, premarket clinical study (NCT04060680) | Medtronic EV-ICD system (Model DVEX3E4) | NR | Acute follow-up | Medium | ****** |
| Coderre et al. [20] | Retrospective, single-center study | Medtronic Aurora EV-ICD™ system (Model DVEA3E4) | January 2021 to February 2025 | 11 ± 9 months | High | ***** |
| Crozier et al. [21] | EV-ICD Pilot Study: Prospective, nonrandomized chronic, first-in-human, multi-center study (NCT03608670) | Medtronic EV-ICD system (Model DVEX2E4) | 2018 | 3 months | High | ****** |
| Crozier et al. (Enlighten Study) [14] | Enlighten: Prospective, global, post-market registry study (NCT06048731) | Medtronic Aurora EV-ICD™ system (Model DVEA3E4) | NR | Through discharge | High | ****** |
| Crozier et al. (Continued Access Study) [14]* | Continued Access Study: Prospective, non-randomized, pre-market, multi-center study (NCT04060680) | Medtronic EV-ICD system (Model DVEX3E4) | 2023 | Through discharge | Medium | ****** |
| Dragisic et al. [23] | Retrospective, single-center case series | Medtronic Aurora EV-ICD™ system (Model DVEA3E4) | July 2024 to January 2025 | 5 months to 9 months | High | ***** |
| Falco et al. [24] | Retrospective, single-center case series | Medtronic Aurora EV-ICD™ system (Model DVEA3E4) | October 2023 to June 2024 | 3 months | Medium | **** |
| Friedman et al. [6] | EV ICD Pivotal Study: Prospective global, multicenter, single-group, nonrandomized, premarket approval study (NCT04060680) | Medtronic EV-ICD system (Model DVEX3E4) | September 2019 to October 2021 | 10.6 months (mean follow-up) | High | ****** |
| Lousinha et al. [25] | Prospective, observational, study at two centers | Medtronic Aurora EV-ICD™ system (Model DVEA3E4) | November 2024 to June 2025 | 3 months | High | ****** |
| Walsh et al. [27] | Retrospective, single-center study | Medtronic Aurora EV-ICD™ system (Model DVEA3E4) | May 2024 to April 2025 | 5 months (median follow-up) | High | ***** |
| AtaCor EV-ICD lead (intercostal EV-ICD lead) | ||||||
| Burke et al. [8] | PASS PULL EV ICD (Phase I): Prospective, nonrandomized, single-center, first-in-human study (NCT05099289) SECURE EV (Phase II): Prospective, nonrandomized, single-center, first-in-human study (NCT05352776) |
AtaCor EV-ICD lead system with commercially available DF-4 compatible TV-ICDs | November 2021 to June 2022 | PASS PULL EV ICD (Phase I): 7–10 days; SECURE EV (Phase II): 90 days | High | ****** |
| Doshi et al. [22] | STEP ICD Study: Prospective, single-center, early feasibility, nonrandomized study (NCT05791032) | AtaCor Atala™ EV-ICD lead with commercially available DF-4 compatible TV-ICDs | September 2023 to January 2024 | 82 days | High | ****** |
| van der Stuijt et al. [26] | ASCEND EV: Prospective, multicenter, single-arm nonrandomized study (NCT06592001) | AtaCor EV-ICD lead system with commercially available DF-4 compatible TV-ICDs (Model AC-700300) | February 2025, to July 2025 | 30 days | Medium | ****** |
Study information provided for each of the 14 included studies. Abbreviations are as follows: EV-ICD = extravascular implantable cardioverter-defibrillator; NR = not reported; PASS PULL EV ICD = Parasternal Access for Shocks and Pacing With an Acutely Placed Less-Invasive Lead for Extravascular-ICD; SECURE EV = Sensing and Defibrillation with a Commercially Available ICD Coupled with a Parasternal Extravascular Lead; STEP ICD = Sub Chronic Evaluation for ATP with an Extravascular Placed ICD Lead Study; TV-ICD = transvenous implantable cardioverter defibrillator.
*This article includes patients from the Continued Access study which reports data through discharge, Billakanty et al. reports outcomes for 23 of 33 Continued Access Study patients, some of which is not presented in the full cohort report (Crozier et al. 2025).
†A range of zero to six stars to indicate the highest and lowest risk of bias, respectively.
3.2. Patient demographics
In total, 806 patients were included with an implant attempt of either the substernal EV-ICD system (n = 731) or intercostal EV-ICD lead (n = 75), with a mean age that ranged from 13.1 to 58 years (weighted mean: 49.9 years; n = 785), a majority being male (n = 575/797, 72.1%), and mean left ventricular ejection fraction (LVEF) that ranged from 38% to 57% (weighted mean: 42.5%, n = 761) when reported. An average of 24.4% (n = 141/578) of patients had a secondary prevention indication for an ICD (Table 2).
Table 2.
Patient characteristics by study.
| Study reference | Cohort size | Mean patient age ± SD (years) | Mean BMI ± SD (kg/m2) | Male sex (n, %) |
Mean LVEF ± SD (%) | NYHA Class III or higher (n, %) |
ICD Indication – primary prevention (n, %) |
Disease indications (n, %) |
|---|---|---|---|---|---|---|---|---|
| Medtronic EV-ICD system (substernal EV-ICD system) | ||||||||
| Biffi et al. [18] | 7 | 13.3 ± 1.6 | 17.9 ± 1.6 | NR | 50.4 ± 12.2 | Mean class: 1.3 | 5 (71.4%) | ACM: 1 (14.3%) DCM: 3 (42.9%) HDC: 3 (42.9%) |
| Biffi et al. [19] | 65 | 44 ± 17 | 24 ± 5 | 43 (69%) | 47 ± 14 | 4 (6%) | 54 (83%) | ACM: 24 (35%) BrS: 3 (4%) DCM: 10 (15%) HCM: 11 (18%) Other: 9 (14%) |
| Billakanty et al. [13] | 23 | 58 ± 14 | 30.2 ± 6.7 | 18 (78.3%) | 40 [SD NR] |
NR | NR | NR |
| Coderre et al. [20] | 23 | 39 ± 16 | NR | 15 (65%) | 57 ± 12 | NR | 15 (65%) | ACM: 8 (35%) DCM: 2 (9%) HCM: 9 (39%) IVF: 3 (13%) LQTS: 1 (4%) |
| Crozier et al. [21] | 21 | 54.2 ± 13.8 | 30.9 ± 4.6 | 17 (81%) | 42.6 ± 18.2 | 1 (4.8%) | 18 (85.7%) | HCM: 6 (28.6%) ICM: 7 (33.3%) NICM: 4 (19.0%) |
| Crozier et al. (Enlighten Study) [14] | 228 | 49.4 ± 15.4 (n = 226) | 27.6 ± 6.2 (n = 226) | 163/226 (72.1%) | 43.5 ± 16.1 (n = 210) | 19/225 (8.4%) | NR | HCM: 36 (15.9%) ICM: 48 (21.2%) NICM: 71 (31.4%) |
| Crozier et al. (Continued Access Study) [14]* | 33 | 49.2 ± 13.9 | 31 ± 8 | 24 (72.7%) | NR | 3 (9.1%) | 17 (51.5%) | NR |
| Dragisic et al. [23] | 5 | 16.8 ± 2.3 | Mean: 30 [SD NR; range: 25.4–38.8] |
4 (80%) | NR | NR | 3 (60%) | BrS: 1 (20%) DCM†: 1 (20%) HCM: 2 (40%) LQTS: 1 (20%) |
| Falco et al. [24] | 10 | 52.0 [SD NR] |
26.0 [SD NR] |
6 (60%) | 49.2% [SD NR] | NR | 5 (50%) | AMVP: 2 (20%) ARVC: 3 (30%) HNDC:2 (20%) ICM: 2 (20%) HCM: 1 (10%) |
| Friedman et al. [6] | 316 | 53.8 ± 13.1 | 28.0 ± 5.6 | 236 (74.7%) | 38.9 ± 15.4 | 23 (7.3%) [NA: 34] |
258 (81.6%) | ARVC: 6 (1.9%) BrS: 2 (0.6%) HCM: 41 (13.0%) ICM: 128 (40.5%) LQTS: 5 (1.6%) NICM: 102 (32.2%) Other: 11 (3.4%) |
| Lousinha et al. [25] | 11 | 36.3 ± 11.3 | 22.8 ± 3.8 | 8 (72.7%) | 51.7 ± 11.7 | 0 (0%) | 9 (81.8%) | ARVC: 1 (9%) BrS: 1 (9%) DCM: 2 (18%) HCM: 3 (27%) LVNDC: 3 (27%) PVT: 1 (9%) |
| Walsh et al. [27] | 12 | 13.1 ± 3.1 | NR [Mean weight: 49.7 ± 14.9 kg] | 5 (41.7%) | NR | NR | 11 (91.7%) | ARVC: 1 (8.33%) BrS: 2 (16.7%) DCM: 3 (25%) HCM: 4 (33.3%) LQTS: 2 (16.7%) |
| AtaCor EV-ICD lead (intercostal EV-ICD lead) | ||||||||
| Burke et al. [8] | 36 | 58 ± 14 | 28 ± 4 | 28 (88%) | 38 ± 14 | 8 (22%) | 20 (56%) | ICM: 10 (28%) NICM: 21 (58%) Other: 3 (8%) |
| Doshi et al. [22] | 20 | 57 ± 15 | 29 ± 4 | 16 (80%) | 54 ± 14 | 1 (5%) | 7 (35%) | NICM: 58% [n not given] |
| van der Stuijt et al. [26] | 19 | NR [all were ≥18 years] |
29 ± 4 | 10 (53%) | 47 ± 18 | 1 (7%) | 15 (79%) | HCM: 5 (26.3%)‡ ICM: 10 (52.6%) NICM: 9 (47.4%) |
Baseline patient characteristics for each study where reported. Abbreviations are as follows: ACM: arrhythmogenic cardiomyopathy, AMVP: arrhythmic mitral valve prolapse, ARVC: arrhythmogenic right ventricular cardiomyopathy, BMI: body mass index, BrS: Brugada syndrome, DCM: dilated cardiomyopathy, HCM: hypertrophic cardiomyopathy, HDC: hypokinetic dilated cardiomyopathy, HNDC: hypokinetic non-dilated cardiomyopathy, ICM: ischemic cardiomyopathy, IVF: idiopathic ventricular fibrillation, LQTS: long QT syndrome, LVNDC: non-dilated left ventricular cardiomyopathy, NICM: non-ischemic cardiomyopathy, NR: not reported, PVT: polymorphic ventricular tachycardia, SD: standard deviation.
*Baseline characteristics for this cohort gathered from clinicaltrials.gov (NCT04060680) †TNNI3K mutation; ‡not mutually exclusive with ICM and NICM.
Broken down by type of retrosternal EV-ICD, patients with an implant attempt of the substernal EV-ICD system had a weighted mean age of 49.3 years (n = 729), 72.2% were male (n = 521/722), weighted mean LVEF of 42.3% (n = 686), 78.5% primary prevention (n = 395/503), and 77.0% (n = 536/752) with a documented cardiomyopathy. Patients with an implant attempt of the intercostal EV-ICD lead had a weighted mean age of 57.6 years (n = 56), 72.0% male (n = 54/75), weighted mean LVEF of 44.5% (n = 75), 56.0% primary prevention (n = 42/75), and 96.3% (n = 53/55) had a cardiomyopathy when reported.
3.3. Procedure outcomes
3.3.1. Substernal EV-ICD system
The lead placement success rate ranged from 95.2% to 100% and a weighted average of 99.6% (n = 728/731). Overall procedure success for the substernal EV-ICD system ranged from 81.0% to 100% with an average success of 95.6% (n = 689/721). The time to lead placement was reported in three substernal EV-ICD system studies and ranged from median of 25.0 to a mean of 42.6 minutes. Total procedure time, reported in nine studies, ranged from 62.0 to 115 minutes for the substernal EV-ICD system (weighted mean time of 79.8 minutes; n = 596). Of note, total procedure times for the substernal EV-ICD system include defibrillation testing when performed. Fluoroscopy time for the substernal EV-ICD system was reported in eight studies with mean times ranging from 2.7 to 6.2 total minutes (Table 3).
Table 3.
Procedure outcomes.
| Study reference | Lead placement success (%, n) | Procedure success (%, n) | Time to lead placement (min) (mean ± SD or median [IQR]) | Total procedure time (min) (mean ± SD or median [IQR]) | Fluoroscopy time (min) (mean ± SD or median [IQR]) |
|---|---|---|---|---|---|
| Medtronic EV-ICD system (substernal EV-ICD system) | |||||
| Biffi et al. [18] | 100% (n = 7/7) | 100% (n = 7/7) | NR | 97 ± 14 | Mean: 2.7 [SD NR] |
| Biffi et al. [19] | 100% (n = 65/65) | 98.5% (n = 64/65) | NR | NR | NR |
| Billakanty et al. (Subgroup of Continued Access Study) [13]* | NR | NR | 29 ± 13† | 62 ± 23* | 4 ± 4.5 |
| Coderre et al. [20] | 100% (n = 23/23) | 95.7% (n = 22/23) | NR | NR | NR |
| Crozier et al. [21] | 95.2% (n = 20/21) | 81% (n = 17/21) | 25 (21–35) | 85 (78–104) | Median: 4.2 [IQR NR] |
| Crozier et al. (Enlighten Study) [14] | 99.6% (n = 227/228) | 96.9% (n = 221/228) | NR | 84.9 ± 43.5 | NR |
| Crozier et al. (Continued Access Study) [14] | 100% (n = 33/33) | 93.9% (n = 31/33) | NR | 80.1 ± 32.9 | NR |
| Dragisic et al. [23] | 100% (n = 5/5) | 100% (n = 5/5) | NR | NR | 4.8 ± 1.8 |
| Falco et al. [24] | 100% (n = 10/10) | NR | NR | 99 (77.5–102.5) | Mean: 6.2 [SD NR] |
| Friedman et al. [6] | 99.7% (n = 315/316) | 94.6% (n = 299/316) | 42.6 ± 25.3 (n = 304) | 74.6 ± 33.2 (n = 305) | 5.8 ± 4.2 (n = 313) |
| Lousinha et al. [25] | 100% (n = 11/11) | 100% (n = 11/11) | NR | 68.2 ± 6.7 | 3.9 ± 1.4 |
| Walsh et al. [27] | 100% (n = 12/12) | 100% (n = 12/12) | NR | 115 ± 9.3 | 2.2 (1.6–3.4) |
| AtaCor EV-ICD lead (intercostal EV-ICD lead) | |||||
| Burke et al. [8] | 94% (n = 33/35) | 91.4% (n = 32/35) | 23.7 ± 9.4 | NR | NR |
| Doshi et al. [22] | 100% (n = 20/20) | 100% (n = 20/20) | 9 (7,15) [not including lead fixation] |
NR | NR |
| van der Stuijt et al. [26] | 100% (n = 19/19) | 89.5% (n = 17/19) | 39 ± 20 (n = 18) | 77 ± 24 (n = 18) | NR |
Listing of procedure outcomes including success rates and procedure times. Abbreviations are as follows: IQR: interquartile range, min: minutes, NR: not reported, SD: standard deviation.
*See Crozier et al. 2025 for the full cohort results; †reported in Billakanty et al. 2025 but not Crozier et al. 2026.
3.3.2. Intercostal EV-ICD lead
Lead placement success for the intercostal EV-ICD lead was reported in all three studies, ranging from 94.0% to 100%, with a weighted success rate of 97.3% (n = 74), and procedure success ranged from 89.4% to 100% (weighted success rate: 93.2%; n = 74). Lead placement time was reported in two studies (n = 53), with mean times ranging from 23.7 to 39.0 minutes (Table 3). Total procedure time was reported in one study with a mean of 77 minutes (n = 18). Fluoroscopy time was not reported for the intercostal EV-ICD lead, though fluoroscopy use was noted in all three studies.
3.4. Defibrillation testing outcomes
3.4.1. Substernal EV-ICD system
Defibrillation testing was performed in most patients (n = 675/731; 92.3%) with the substernal EV-ICD system and was successful in 90.0% to 100% of patients with a weighted average success rate of 98.5% (n = 665/675). The energy at which defibrillation testing was successful was ≤30J (minimum 10J safety margin) for most patients, with the lowest reported successful mean energy being 18.6J and the highest being 32.0J. For studies where a mean energy for success was reported or could be determined, the weighted minimum mean energy for a successful defibrillation test was 26.1J (n = 572). Shock impedance was measured in three studies with values ranging from 42.9 to 63.9 Ohms (Table 4).
Table 4.
Defibrillation testing outcomes and electrical measurements at implant.
| Study reference | Defibrillation testing success definition/ protocol | Defibrillation testing success (%) | Minimum energy successful (J) | R-wave amplitude at implant (mV) | Pacing capture threshold at implant (V) | Impedance at implant (Ω) | Shock impedance (Ω) |
|---|---|---|---|---|---|---|---|
| Medtronic EV-ICD system (substernal EV-ICD system) | |||||||
| Biffi et al. [18] | Success at any energy | 100% (n = 7/7) | 18 ± 5 | 5.3 ± 2 (n = 7) [R1-R2] | NR | NR | NR |
| Biffi et al. [19] | Success at any energy | 100% (n = 64/64) | NR | 3.2 ± 1.9 (n = 64) | NR | NR | NR |
| Coderre et al. [20] | Success at any energy | 100% (n = 22/22) | 30.5 ± 2.1 | NR | NR | NR | NR |
| Crozier et al. [21] | Single 20J success or two consecutive 30J successes | 90% (n = 18/20) | 18.6 ± 5.6 (n = 18) | 3.4 ± 2.0 (n = 20) | 5.4 ± 2.2 (n = 19) | NR | NR |
| Crozier et al. (Enlighten Study) [14] | Success at any energy | 99% (n = 191/193) | 31 ± 4.2 | 2.9 ± 1.7 (n = 205) [R1-R2] |
5.4 ± 2.1 (n = 137) [R1-C2] |
429.4 ± 190.6 (n = 202) [R1-R2] | NR |
| Crozier et al. (Continued Access Study) [14] | Single 20J success or two consecutive 30J successes | 94% (n = 30/32) | NR | 2.4 ± 0.9 (n = 22) [R1-R2] |
5.7 ± 2.1 (n = 31) [R1-C2] |
403.9 ± 189.6 (n = 31) [R1-R2] |
NR |
| Dragisic et al. [23] | First shock success at 25J or 30J | 100% (n = 5/5) | 26 ± 2.2 | 2.8 ± 1.1 [R1-R2] |
NR | 307.6 ± 68.8 [R1-R2] |
59.0 ± 6.2 |
| Falco et al. [24] | Success at any energy | 100% (n = 10/10) [2/10 successful 3 days post-implant] |
32 ± 4 | Mean: 1.7 (n = 8) [SD NR] |
NR | NR | NR |
| Friedman et al. [6] | Single 20J success or two consecutive 30J successes | 98.7% (n = 298/302) | 22.8 ± 4.5 (n = 299) | 2.9 ± 1.6 (n = 312) [R1-R2] |
4.9 ± 2.0 (n = 239) [R1-C2] |
NR | NR |
| Lousinha et al. [25] | Success at any energy | 100% (n = 11/11) | 30.9 ± 3.0 | 2.7 ± 1.0 [R1-R2] |
NR | NR | 63.9 ± 11.3 |
| Walsh et al. [27] | Success at any energy | 100% (n = 9/9) [3/12 patients couldn’t be induced] |
30.0 ± 0 (n = 9) |
4.7 ± 3.1 | NR | 306.6 ± 45.0 [R1-R2] |
42.9 ± 5.0 (n = 9) |
| AtaCor EV-ICD lead (intercostal EV-ICD lead) | |||||||
| Burke et al. [8] | 30J start; step down to a minimum of 10J | LMA: 100% (n = 27/27) LP: 83% (n = 5/6) |
LMA: 16.3 ± 8.6 (n = 27) LP: 22.2 ± 8.3 (n = 5) |
3.9 ± 2.0 (n = 33) [dedicated bipolar] | ≤10 V: 18.2% (n = 6/33) | 1183 ± 614 (n = 33) [dedicated bipolar] | LMA: 74 ± 19 (79 shocks) LP: 70 ± 5 (19 shocks) |
| Doshi et al. [22] | Step down threshold testing | 100% (n = 20/20) [LP: 19, LMA: 1] |
LMA: 14.5 (n = 17) [SD NR] |
3.4 ± 1.7 (n = 20) [dedicated bipolar] 2.8 ± 1.7 (n = 20) [integrated bipolar] |
8.5 ± 4.3 (n = 17) [dedicated bipolar] 5.7 ± 3.3 (n = 19) [integrated bipolar] |
1008 ± 340 (n = 20) [dedicated bipolar] 497 ± 265 (n = 20) [integrated bipolar] |
LMA: 76 ± 15 (42 shocks) LP: 73 ± 13 (35 shocks) |
| van der Stuijt et al. [26] | Single 20J success or 2 consecutive 10J successes | 94% (n = 17/18) [LP: 2, LMA: 15] |
94% success at ≤20J (n = 16/17) | 3.6 ± 1.9 (n = 17) [dedicated bipolar] 2.5 ± 1.2 (n = 17) [integrated bipolar] |
6.9 ± 3.7 (n = 17) [dedicated bipolar] 4.3 ± 2.4 (n = 17) [integrated bipolar] |
679 ± 237 (n = 17) [dedicated bipolar] 349 ± 129 (n = 17) [integrated bipolar] |
70.8 ± 17.9 (n = 17) |
Listing of defibrillation testing outcomes and electrical measurements taken at implant, where reported. Abbreviations and symbols are as follows: Ω: Ohms, C2: coil 2, J: joules, LMA: left mid-axillary pulse generator position, LP: left pectoral pulse generator position, NR: not reported, R1-R2: Ring 1/Ring 2, V: volts.
3.4.2. Intercostal EV-ICD lead
All patients with successful lead placement of the intercostal EV-ICD lead underwent defibrillation testing at implant with a success rate of 94% to 100% (weighted success rate: 97.2%, n = 69/71). The mean energy at which defibrillation testing was successful was reported in two studies with a weighted mean energy of 16.2J (n = 49 patients) at a range of 14.5J (n = 17; 15 left mid-axillary) to 22.2J (n = 5; left pectoral generator placement). In the third intercostal EV-ICD lead study, 94.4% (17/18) were successful at ≤20J, with no mean energy reported. Of note, defibrillation testing was evaluated with two different pulse generator positions with the intercostal EV-ICD lead, a left mid-axillary and a left pectoral location in all studies (see Table 3). The mean shock impedance during defibrillation testing was reported in all studies with minimum and maximum values of 70.8 and 76.0 Ohms (most in the left mid-axillary position), respectively (Table 4).
3.5. Electrical measurements
3.5.1. Substernal EV-ICD system
Mean R-wave amplitude measured during the implant procedure, most commonly from the ring 1 to ring 2 vector, ranged from 1.7 mV to 5.3 mV, with a weighted mean R-wave amplitude of 3.0 mV across studies where reported (n = 666 patients). Mean lead impedance values at implant from ring 1 to ring 2 ranged from 302 to 429.4 Ohms (reported in 4 studies). Pacing capture threshold ranged from 4.9 to 5.7 volts across four studies, where reported (Table 4). When documented, patients noted sensation of pacing during threshold testing, with 4.6% of patients in the Pivotal study having ATP programmed off due to in-office testing and 1.8% having post-shock pacing turned off for the same reason. In the Pilot study pacing sensation was described as mild to moderate but tolerable in all patients to allow programming of pacing therapies.
3.5.2. Intercostal EV-ICD lead
The mean R-wave amplitude at implant ranged from 3.6 to 3.9 mV via the dedicated bipolar vector (weighted mean: 3.7 mV; n = 70) and 2.5 to 2.8 mV via the integrated bipolar vector (weighted mean: 2.7 mV; n = 37). Mean impedance values during implant at the integrated bipolar vector ranged from 349 Ohms to 497 Ohms across all three studies. Pacing capture threshold with the integrated bipolar vector was reported in two studies with mean values of 4.3 V to 5.7 V (Table 4). Doshi et al. reported 5% of subjects reporting intolerable pacing sensation which diminished to sensation only by the time of lead removal, while Van der Stuijt et al. reported mild to no pacing sensation in all patients, with none finding it intolerable.
3.6. Acute safety
3.6.1. Substernal EV-ICD system
Freedom from major system- or procedure-related complications was reported in two studies of the substernal EV-ICD system with a rate of 92.6% (one side CI: 89%) and 94.1% (95% CI: 83.6%-100%) through 182 days and 90 days, respectively (Supplemental Table S4). The rate of periprocedural complications (occurring before discharge) ranged from 0% to 8.3% (n = 1/12 patients), with a weighted average of 1.5% (n = 10/731). None of the events required cardiothoracic surgeon intervention or resulted in any lasting effects for the patient and were resolved with lead repositioning or medication. The most commonly reported complications that occurred from procedure through all follow-up included lead dislodgment (weighted average: 1.2%, range: 0%-4.3%, n = 9/731), infection (weighted average: 1.4%, range: 0%-4.8%, n = 7/503), pneumothorax (weighted average: 0.3%, range: 0%-0.9%, n = 2/731), and inadvertent lead placement in the pericardial (n = 2) or pleural space (n = 2) (weighted average: 0.5%, range: 0%-8.7%, n = 4/654) (Ta b le 5). Three of the four inadvertent lead placements were considered minor and required no additional intervention.
3.6.2. Intercostal EV-ICD lead
Freedom from major complications through 30-days was reported in the study by van der Stuijt et al. with a rate of 95% (CI: 74%-99%) (Supplemental Table S4). The rate of major device-related periprocedural complications for the intercostal EV-ICD lead ranged from 10.0% (n = 2/20) to 11.4% (n = 4/35), with a weighted average of 8.2% (n = 6/73). Through all of follow-up, the intercostal EV-ICD lead reported complications such as pneumothorax (2.8%, n = 2/72), lead migration resulting in lead revision (1.4%, n = 1/72), and infection of the incision site requiring complete system removal (1.4%, n = 1/72) (Table 5).
Table 5.
Overview of periprocedural complications and select complications.
| Study reference | Periprocedural complications (n, %) | Device dislocation (n, %) |
Hemorrhage (n, %) | Infection (n, %) | Inadvertent lead placement (n, %) | Lead dislodgment (n, %) | Pneumo-peritoneum (n, %) | Pneumo-thorax (n, %) |
|---|---|---|---|---|---|---|---|---|
| Medtronic EV-ICD System (substernal EV-ICD system) | ||||||||
| Biffi et al. [18] | 0% (n = 0/7) |
0% (n = 0/7) |
0% (n = 0/7) |
0% (n = 0/7) |
0% (n = 0/7) |
0% (n = 0/7) |
0% (n = 0/7) |
0% (n = 0/7) |
| Biffi et al. [19] | 0% (n = 0/65) | 0% (n = 0/65) |
0% (n = 0/65) |
0% (n = 0/65) |
0% (n = 0/65) |
1.5% (n = 1/65) | 0% (n = 0/65) |
0% (n = 0/65) |
| Coderre et al. [20] | 4.3% (n = 1/23)* |
0% (n = 0/23) |
0% (n = 0/23) |
0% (n = 0/23) |
8.7% (n = 2/23)† |
4.3% (n = 1/23) | 0% (n = 0/23) |
0% (n = 0/23) |
| Crozier et al. [21] | 0% (n = 0/21) | 0% (n = 0/21) | 0% (n = 0/21) | 4.8% (n = 1/21) | 4.8% (n = 1/21)† |
0% (n = 0/21) | 0% (n = 0/21) |
0% (n = 0/21) |
| Crozier et al. (Enlighten Study) [14] | 3.9% (n = 9/228) | 0.4% (n = 1/228) |
0% (n = 0/228) | NR | 0% (n = 0/228) | 0.9% (n = 2/228) | 0.44% (n = 1/228) | 0.9% (n = 2/228) |
| Crozier et al. (Continued Access Study) [14] | 0% (n = 0/33) | 0% (n = 0/33) | 0% (n = 0/33) | 0% (n = 0/33) | 0% (n = 0/33) | 0% (n = 0/33) | 0% (n = 0/33) |
0% (n = 0/33) |
| Dragisic et al. [23] | 0% (n = 0/5) |
0% (n = 0/5) |
0% (n = 0/5) |
0% (n = 0/5) |
0% (n = 0/5) |
0% (n = 0/5) |
0% (n = 0/5) |
0% (n = 0/5) |
| Falco et al. [24] | 0% (n = 0/10) | 0% (n = 0/10) | 0% (n = 0/10) | 0% (n = 0/10) | 0% (n = 0/10) | 0% (n = 0/10) | 0% (n = 0/10) |
0% (n = 0/10) |
| Friedman et al. [6] | 0% (n = 0/316) | 0% (n = 0/316) | 0.3% (n = 1/316) | 1.9% (n = 6/316) | 0% (n = 0/316) | 1.6% (n = 5/316) | 0% (n = 0/316) | 0% (n = 0/316) |
| Lousinha et al. [25] | 0% (n = 0/11) | 0% (n = 0/11) | 0% (n = 0/11) | 0% (n = 0/11) | 0% (n = 0/11) | 0% (n = 0/11) | 0% (n = 0/11) | 0% (n = 0/11) |
| Walsh et al. [27] | 8.3% (n = 1/12) |
0% (n = 0/12) | 0% (n = 0/12) | 0% (n = 0/12) | 8.3% (n = 1/12) | 0% (n = 0/12) | 0% (n = 0/12) | 0% (n = 0/12) |
| Weighted Average | 1.5% (n = 10/731) | 0.1% (n = 1/731) | 0.1% (n = 1/731) | 1.4% (n = 7/503) | 0.3% (n = 4/731) | 1.2% (n = 9/731) | 0.1% (n = 1/731) | 0.3% (n = 2/731) |
| AtaCor EV-ICD Lead (intercostal EV-ICD lead) | ||||||||
| Burke et al. [8] | 11.4% (n = 4/36) | 0% (n = 0/35) | 0% (n = 0/35) | NR | 0% (n = 0/35) | 0% (n = 0/35) | 0% (n = 0/35) | 2.9% (n = 1/35) |
| Doshi et al. [22] | 5.0% (n = 1/20) | 0% (n = 0/20) | 0% (n = 0/20) | 5% (n = 1/20) | 0% (n = 0/20) | 5% (n = 1/20) | 0% (n = 0/20) | 0% (n = 0/20) |
| van der Stuijt et al. [26] | 5.9% (n = 1/17) | 0% (n = 0/17) | 0% (n = 0/17) | 0% (n = 0/17) | 0% (n = 0/17) | 0% (n = 0/17) | 0% (n = 0/17) | 5.9% (n = 1/17) |
| Weighted Average | 8.2% (n = 6/73) |
0% (n = 0/72) |
0% (n = 0/72) |
2.7% (n = 1/37) |
0% (n = 0/72) |
1.4% (n = 1/72) |
0% (n = 0/72) |
2.8% (n = 2/72) |
Listing of select major complications as defined by the study through varying follow-up times; note that a rate of 0% means that the study did not report the complication to have occurred. Abbreviations are as follows: NR: not reported, but insufficient follow-up time for event to occur.
*One of the inadvertent lead placements occurred during the procedure but was recognized immediately and lead repositioned with no negative outcomes for the patient.
†None of the events resulted in any adverse outcomes for the patients and were not considered major complication.
4. Discussion
This systematic review demonstrates acute results from studies that evaluated outcomes with retrosternal EV-ICDs (the substernal EV-ICD system and intercostal EV-ICD lead). Patients across all studies were young with a relatively preserved LVEF, and nearly a quarter with a secondary prevention ICD indication. Lead placement (99.6% for substernal and 97.3% for intercostal) and procedure success rates were high across studies (95.6% and 93.2% for substernal and intercostal, respectively). Defibrillation testing was performed in most patients, and although definitions for success varied across studies, testing was highly successful when performed (98.5% for substernal EV-ICD system and 97.2% for intercostal EV-ICD lead). Additionally, the energy at which defibrillation testing was successful was ≤30 J for most patients. Periprocedural complications were low with an overall weighted average for retrosternal EV-ICDs of 2.0% (n = 16/804). The rate of periprocedural complications with the substernal EV-ICD system was low (1.5%, n = 11/731). The rate of periprocedural complications for the intercostal EV-ICD lead was higher (8.2%, n = 6/73), however, this may be attributed to the small sample sizes across studies.
Patients with an implant attempt of the substernal EV-ICD system included a wide range of age groups from pediatric populations (three studies) to an average age of up to 54 years old, while those limited number of patients with the intercostal EV-ICD lead were slightly older (57–58 years old; two studies reported) [8,18,21–23,27]. This age is in line with what has also been reported for the S-ICD system, which is also extravascular but with a lead that sits outside of the ribcage [4]. However, traditional TV-ICD patients are generally older (60–64 years old) with more progressive heart failure as evidenced by an LVEF of around 30–34% in recent TV-ICD studies [28,29]. The younger age of EV-ICD patients likely represents a strategy by implanters to utilize extravascular ICD systems in younger patients who project to be implanted for many years to preserve the vasculature and avoid complications associated with chronic lead extraction. A preliminary analysis of the substernal EV-ICD demonstrated that it is feasible with simple traction in most cases, although more data is needed regarding chronic lead extraction of retrosternal EV-ICDs with dwell times greater than 5 years [30]. Another factor that may affect patient selection and general adoption of extravascular ICD technology is the inability to provide chronic bradycardia pacing. Across studies of S-ICD and EV-ICD, typically less than 2% of patients developed the need for chronic pacing, and in the PRAETORIAN trial, with an older patient group, only six patients (0.7%) developed a pacing indication across both study arms [28]. An analysis of single chamber TV-ICDs showed that 5% of patients required an upgrade to a CRT-D system at five years [31]. These rates suggest that considering a future need for chronic pacing may not be a major factor for patients when determining the correct ICD system.
Procedure success was around 96% for the retrosternal EV-ICD systems, with time to lead fixation and total procedure time being similar between EV-ICD types. Procedure success may be lower than what is expected for TV-ICD systems, with poor sensing being the main reason for failed implantation with retrosternal EV-ICDs, though electrical testing may not be as rigorous for TV-ICDs as they are for more novel ICD systems. Procedure time is longer for EV-ICDs than what is reported for TV-ICD [28,29]. Novelty of the procedure and anatomical location, performing defibrillation testing, and use of general anesthesia likely contribute to the longer time, but this could improve as implanters become more comfortable with the procedure/system and move away from defibrillation testing or use of general anesthesia in every patient [32].
Both the substernal EV-ICD system (weighted success rate: 98.5%) and intercostal EV-ICD lead (weighted success rate: 97.2%) demonstrated a high defibrillation testing success rate, which is in line with recently reported success rates for TV-ICDs (97.9%) and the S-ICD (97.1%) [28,29]. While defibrillation testing is still recommended for the S-ICD as it is for the substernal EV-ICD system, they are rarely performed for TV-ICD implants in clinical practice due to the cumulative confidence in the system over time [33]. This represents an advantage of TV-ICD to more confidently forego defibrillation testing particularly for hemodynamically unstable patients at higher risk for complications during testing. Many substernal EV-ICD studies did not perform step-down defibrillation threshold testing but instead opted to test at one energy, likely to minimize the number of inductions and shocks. Across successful tests, the minimum energy needed to defibrillate for retrosternal EV-ICD systems was around 30J, which is slightly higher, though still in agreement, with energies reported for TV-ICDs [29,33]. In comparison, the S-ICD has a higher defibrillation energy requirement relative to other ICD types, with tests normally performed up to 65J [34]. Collectively, these data provide confidence that retrosternal lead placement is an effective strategy for providing defibrillation therapy for ventricular arrhythmias at energies similar to TV-ICD. However, consideration should be given to conducting a randomized clinical trial to confirm the safety of no defibrillation test during implantation of a substernal EV-ICD system.
An additional advantage of EV-ICD lead placement under the sternum is the ability to provide both defibrillation and ATP therapy. However, pacing in the retrosternal space does necessitate a higher voltage (>4 V versus ~0.6 V for TV-ICD) to achieve capture which can be perceived by some patients, if only for a short time [29]. Higher pacing thresholds with retrosternal ICDs can activate skeletal muscle tissue resulting in sensation that can be uncomfortable for some. However, higher pacing thresholds were generally tolerated across patients, with a small number not being able tolerate pacing sensation, and in-office testing may not be totally representative of ambulatory pacing therapy [6,21,22]. The option for ATP therapy provides the potential to avoid high-energy shocks which can be painful and reduce patient quality of life [35]; so, ATP, even if perceptible in some patients, should be preferable over receiving a shock. Additionally, Sears et al. showed improved quality of life with the substernal EV-ICD system compared to a historical cohort of TV-ICD patients using the Florida patient acceptance survey [36]. Given that only a small subset of subjects reported intolerable sensation, more work will be needed to identify factors that lead to pacing sensation, whether it be patient-specific anatomy, location of the electrode, or the type of pacing waveform. As such, the availability of ATP with retrosternal EV-ICDs is an improvement upon the S-ICD and may result in greater adoption of extravascular ICD technology over time.
The substernal EV-ICD system, with multiple large cohorts with extended follow-up durations, reported a low rate of periprocedural complications (1.5%) including several studies with no intraprocedural or periprocedural complications. This is important given that the implant technique and location of the lead are novel and necessitate placement in anatomical spaces not typically accessed by electrophysiologists. The substernal EV-ICD system requires that physicians complete a dedicated implanter training program before implanting, and their first five cases must be performed with a cardiothoracic surgeon present [37]. Importantly, none of the substernal EV-ICD studies required surgical intervention and no complications resulted in irreversible damage to the heart or lungs. Major complication rates for the substernal EV-ICD system were in line with those reported for both TV-ICD and S-ICD systems [3,28,38]. There were four inadvertent lead placements into the pericardial (n = 2) or pleural (n = 2) space reported with the substernal EV-ICD system. Three of four events were considered minor by the study investigators with no lasting impacts or symptoms reported, while one required an additional invasive procedure to revise the lead. Though the focus of this analysis was to report on shorter-term outcomes, the substernal EV-ICD system Pilot and Pivotal studies have demonstrated outcomes through three years, pointing to the long-term safety of the substernal EV-ICD system [7,39].
Periprocedural complications with the intercostal EV-ICD were higher compared to the substernal EV-ICD system, though this may likely be due in part to the small number of patients in studies with the intercostal EV-ICD lead. Therapy was turned off during short follow-up durations of the intercostal EV-ICD lead, but a 30-day Kaplan-Meier rate of major complications of 5% was reported in one study [26]. The most common complication reported for both retrosternal EV-ICDs was lead dislodgement (n = 10 events across both EV-ICD types), and this is also one of the most common complications associated with TV-ICDs, though less commonly seen with the S-ICD [40]. The S-ICD with the electrode being located outside the chest wall naturally mitigates certain complications that, although rare, may occur with retrosternal lead placement near the heart and lungs. However, this comes with the tradeoff of inability to provide ATP and a higher energy requirement necessitating a larger generator, which may lead to higher risk of erosion in older patients and discomfort in smaller patients [5]. Additionally, S-ICD has a shorter projected battery life potentially resulting in more frequent device change outs and associated complications over time [41].
One of the limitations of extravascular ICD systems to date is the higher rate of inappropriate shock compared to TV-ICD. Inappropriate shocks have yet to be investigated with the intercostal EV-ICD given the nature of the studies. The substernal EV-ICD reported a one-year inappropriate shock rate of 9.8% in the pre-market Pivotal trial, with P-wave oversensing being the most common culprit [7] and Crozier et al. reported one inappropriate shock due to P-wave oversensing in 19 patients in the Pilot study with both studies using the pre-market substernal EV-ICD model [19,21]. To address this, the post-market Aurora EV-ICD system was updated to include a novel P-wave discriminator (Smart Sense) which showed a 29% decrease in inappropriate shock rate during modeling of Pivotal episodes [42]. None of the substernal EV-ICD studies included in this review that implanted the post-market Aurora EV-ICD reported inappropriate shocks due to P-wave oversensing [14,19,20,23–25,27], but more real-world data is needed in larger cohorts to assess the true impact of the Smart Sense algorithm on reducing P-wave oversensing and inappropriate shock. For context, the S-ICD has reported one-year rates of inappropriate shock ranging from 6.7–8.6% [5,34] with the main culprit being T-wave oversensing, and, with the latest S-ICD model, an inappropriate shock rate of 3.1%-8.5% [38]. One-year inappropriate shock rates with TV-ICDs using modern programming have been shown to be around 5% [43,44] and as low as 2% [45,46].
4.1. Limitations
There are several limitations of this systematic review that must be considered. This was a retrospective analysis of the literature with no prospective or matched comparisons to S-ICD, TV-ICD, or guideline-directed medical therapy. Additionally, there were no p-value comparisons to assess outcomes between studies or between both types of retrosternal EV-ICD. Experience with the intercostal EV-ICD lead is limited with a small number of patients, and so results are preliminary compared to those with the substernal EV-ICD system. Definitions for some endpoints and for what constitutes a major complication were not reported in all studies and thus may have differed slightly between studies. Patients who were implanted with a retrosternal EV-ICD were younger with less severe heart failure compared to typical TV-ICD patients, more so for the substernal EV-ICD system than for the intercostal EV-ICD lead. Weighted rates and averages give a high-level view of outcomes from the studies herein but may not be representative of rates for the total retrosternal EV-ICD population (specifically the substernal EV-ICD system which has an expanding number of users globally). Some of the studies were reports from pre-market clinical trials, which may not be representative of the real-world experience.
5. Conclusion
The literature outlined in this report shows a high rate of implant and defibrillation testing success with a low rate of acute complications with retrosternal EV-ICD systems, particularly with the substernal EV-ICD system which includes experience from several large global studies, whereas intercostal EV-ICD lead experience is limited. The available evidence suggests that periprocedural outcomes and acute safety of retrosternal EV-ICD systems is an effective strategy to provide potentially lifesaving therapy while preserving the vasculature.
Supplementary Material
Acknowledgments
The authors would like to thank Beatrice Ludovica Ritondo, PhD, from Medtronic Inc for assistance with data extraction and Felipe Atienza, MD, PhD, from Hospital General Universitario Gregorio Marañón for critical review of the manuscript.
Correction Statement
This article has been corrected with minor changes. These changes do not impact the academic content of the article.
Funding Statement
This analysis was supported by Medtronic Inc. The roles of Medtronic authors and those in the acknowledgments are described in the CRediT and acknowledgments sections, respectively.
Article highlights
Retrosternal extravascular implantable cardioverter-defibrillators (EV-ICD) place a lead in the mediastinal space to provide both shock and antitachycardia pacing therapy without the need for venous access.
A systematic review was performed to identify studies involving use of two known retrosternal EV-ICDs, the market-approved Medtronic substernal EV-ICD system and investigational AtaCor intercostal EV-ICD lead.
In total, 14 articles were included that reported on 806 patients with an implant attempt of either the substernal EV-ICD system (n = 731) or intercostal EV-ICD lead (n = 75), with a mean age of 49.9 years, a majority being male (72.1%), and mean left ventricular ejection fraction (LVEF) of 42.5%, when reported.
Procedure success was favorable across all studies with high rates of lead placement (99.6% for substernal and 97.3% for intercostal) and procedure success (95.6% and 93.2% for substernal and intercostal, respectively). Defibrillation testing was also highly successful when performed (98.5% for substernal EV-ICD system and 97.2% for intercostal EV-ICD lead) at energies ≤30 J for most patients.
Periprocedural complications were low with an overall weighted average for retrosternal EV-ICDs of 2.0% (n = 16/804). The most common complication reported was lead dislodgement.
Electrical measurements including R-wave amplitudes and impedance remained consistent across cohorts while pacing capture was achieved at thresholds higher than what is typically observed for transvenous ICDs.
The available evidence suggests that retrosternal EV-ICD systems are a feasible extravascular option for patients who are at risk for sudden cardiac death.
Disclosure statement
G. Crossley has served as a speaker for Medtronic and Philips and as a consultant for Medtronic and Boston Scientific; A. Siehr and T. Holmes are employees of Medtronic Inc. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
Reviewer disclosures
Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.
Supplementary material
Supplemental data for this article can be accessed online at https://doi.org/10.1080/14796678.2026.2676272
References
- 1.Al-Khatib SM, Stevenson WG, Ackerman MJ, et al. 2017 AHA/ACC/HRS guideline for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: executive summary: a report of the American college of cardiology/American heart association task force on clinical practice guidelines and the heart Rhythm society. Heart Rhythm. 2018;15(10):e190–e252. doi: 10.1016/j.hrthm.2017.10.035 [DOI] [PubMed] [Google Scholar]
- 2.Maron BJ, Estes NAM, Rowin EJ, et al. Development of the implantable cardioverter-defibrillator: JACC historical breakthroughs in perspective. J Am Coll Cardiol. 2023;82(4):353–373. doi: 10.1016/j.jacc.2023.04.056 [DOI] [PubMed] [Google Scholar]
- 3.Ezzat VA, Lee V, Ahsan S, et al. A systematic review of ICD complications in randomised controlled trials versus registries: is our ‘real-world’ data an underestimation? Open Heart. 2015;2(1):e000198. doi: 10.1136/openhrt-2014-000198 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Weiss R, Knight BP, Gold MR, et al. Safety and efficacy of a totally subcutaneous implantable-cardioverter defibrillator. Circulation. 2013;128(9):944–953. doi: 10.1161/CIRCULATIONAHA.113.003042 [DOI] [PubMed] [Google Scholar]
- 5.Gold MR, El-Chami MF, Burke MC, et al. Postapproval study of a subcutaneous implantable cardioverter-defibrillator system. J Am Coll Cardiol. 2023;82(5):383–397. doi: 10.1016/j.jacc.2023.05.034 [DOI] [PubMed] [Google Scholar]
- 6.Friedman P, Murgatroyd F, Boersma LVA, et al. Efficacy and safety of an extravascular implantable cardioverter-defibrillator. N Engl J Med. 2022;387(14):1292–1302. doi: 10.1056/NEJMoa2206485 [DOI] [PubMed] [Google Scholar]
- 7.Friedman P, Murgatroyd F, Boersma LVA, et al. Performance and safety of the extravascular implantable cardioverter-defibrillator through long-term follow-up: final results from the pivotal study. Circulation. 2024;151(4):322–332. doi: 10.1161/CIRCULATIONAHA.124.071795 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Burke MC, Knops RE, Reddy V, et al. Initial experience with intercostal insertion of an extravascular ICD lead compatible with existing pulse generators. Circ Arrhythm Electrophysiol. 2023;16(8):421–432. doi: 10.1161/CIRCEP.123.011922 [DOI] [PubMed] [Google Scholar]
- 9.Chen CF, Jin CL, Liu MJ, et al. Efficacy, safety, and in-hospital outcomes of subcutaneous versus transvenous implantable defibrillator therapy: a meta-analysis and systematic review. Medicine (Baltim). 2019;98(19):e15490. doi: 10.1097/MD.0000000000015490 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Chue CD, Kwok CS, Wong CW, et al. Efficacy and safety of the subcutaneous implantable cardioverter defibrillator: a systematic review. Heart. 2017;103(17):1315–1322. doi: 10.1136/heartjnl-2016-310852 [DOI] [PubMed] [Google Scholar]
- 11.Nso N, Nassar M, Lakhdar S, et al. Comparative assessment of transvenous versus subcutaneous implantable cardioverter-defibrillator therapy outcomes: an updated systematic review and meta-analysis. Int J Cardiol. 2022;349:62–78. doi: 10.1016/j.ijcard.2021.11.029 [DOI] [PubMed] [Google Scholar]
- 12.Wolf S, Gotz G, Wernly B, et al. Subcutaneous implantable cardioverter-defibrillator: a systematic review of comparative effectiveness and safety. ESC Heart Fail. 2023;10(2):808–823. doi: 10.1002/ehf2.14249 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Billakanty SR, Badin AG, Cline MJ, et al. Characteristics of extravascular implantable cardioverter-defibrillator implantation. JACC Clin Electrophysiol. 2025;11(8):1877–1879. doi: 10.1016/j.jacep.2025.05.004 [DOI] [PubMed] [Google Scholar]
- 14.Crozier I, Murgatroyd F, Amin A, et al. Periprocedural outcomes from the postmarket study of the extravascular implantable cardioverter-defibrillator: preliminary Enlighten study results and meta-analysis. Heart Rhythm. 2026;23(1):180–185. doi: 10.1016/j.hrthm.2025.02.012 [DOI] [PubMed] [Google Scholar]
- 15.Oxford Uo . Oxford centre for evidence-based medicine: levels of evidence. 2009. [cited 2026]. Available from: https://www.cebm.ox.ac.uk/resources/levels-of-evidence/oxford-centre-for-evidence-based-medicine-levels-of-evidence-march-2009 [Google Scholar]
- 16.Carra MC, Romandini P, Romandini M.. Risk of bias evaluation of cross-sectional studies: adaptation of the Newcastle-Ottawa scale. J Periodontal Res. 2025. doi: 10.1111/jre.13405 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Barker TH, Hasanoff S, Aromataris E, et al. The revised JBI critical appraisal tool for the assessment of risk of bias for cohort studies. JBI Evid Synth. 2025;23(3):441–453. doi: 10.11124/JBIES-24-00103 [DOI] [PubMed] [Google Scholar]
- 18.Biffi M, Quaranta A, Carecci A, et al. The extravascular implantable defibrillator: a novel opportunity for the paediatric population. Europace. 2025;27(7). doi: 10.1093/europace/euaf128 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Biffi M, Quaranta A, Martignani C, et al. Tips and tricks for extravascular implantable cardioverter-defibrillator implantation: a single-centre experience. Europace. 2026;28(2). doi: 10.1093/europace/euag007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Coderre R, Castrichini M, Haq I, et al. Extravascular implantable cardioverter-defibrillator therapy in patients with genetic heart disease: an early single-center experience. Heart Rhythm O2. 2025;6(10):1673–1675. doi: 10.1016/j.hroo.2025.07.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Crozier I, Haqqani H, Kotschet E, et al. First-in-human chronic implant experience of the substernal extravascular implantable cardioverter-defibrillator. JACC Clin Electrophysiol. 2020;6(12):1525–1536. doi: 10.1016/j.jacep.2020.05.029 [DOI] [PubMed] [Google Scholar]
- 22.Doshi SK, Knops RE, Ebner A, et al. Evaluation of a second-generation intercostal extravascular implantable cardioverter defibrillator lead with a pectoral pulse generator for sensing, defibrillation, and anti-tachycardia pacing. Europace. 2025;27(3). doi: 10.1093/europace/euaf044 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Dragisic N, Backes ER, Malloy-Walton L, et al. Short- and mid-term outcomes of the extravascular implantable cardioverter-defibrillator in pediatric and adolescent patients: a single-center experience. J Innov Card Rhythm Manag. 2025;16(11):6488–6495. doi: 10.19102/icrm.2025.16113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Falco R, Baroni M, Gigli L, et al. Protection from sudden cardiac death using a new extravascular implantable cardioverter-defibrillator device: a literature review and a case series from a high-volume single center. G Ital Cardiol (Rome). 2025;26(2):100–105. [DOI] [PubMed] [Google Scholar]
- 25.Lousinha A, Kahlbau H, Coelho S, et al. Initial experience with the extravascular implantable cardioverter-defibrillator: workflow, feasibility and safety. Rev Port Cardiol. 2026;45(3):107–115. doi: 10.1016/j.repc.2025.10.006 [DOI] [PubMed] [Google Scholar]
- 26.van der Stuijt W, Friedman PA, Deshmukh AJ, et al. Acute results of the ascend EV study: a new modality of parasternal extravascular implantable cardioverter-defibrillator therapy including antitachycardia pacing. Heart Rhythm. 2026;23(2):e247–e253. doi: 10.1016/j.hrthm.2025.10.035 [DOI] [PubMed] [Google Scholar]
- 27.Walsh M, Ryan C, Prendiville T, et al. Initial experience with the extravascular implantable cardioverter-defibrillator in pediatric patients. Heart Rhythm O2. 2026;7(1):103–108. doi: 10.1016/j.hroo.2025.10.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Knops RE, Olde Nordkamp LRA, Delnoy PHM, et al. Subcutaneous or transvenous defibrillator therapy. N Engl J Med. 2020;383(6):526–536. doi: 10.1056/NEJMoa1915932 [DOI] [PubMed] [Google Scholar]
- 29.Crossley GH, Sanders P, Hansky B, et al. Safety, efficacy, and reliability evaluation ofa novel small-diameter defibrillation lead: global LEADR pivotal trial results. Heart Rhythm. 2024;21(10):1914–1922. doi: 10.1016/j.hrthm.2024.04.067 [DOI] [PubMed] [Google Scholar]
- 30.Sagi V, Murgatroyd F, Boersma LVA, et al. Comprehensive analysis of substernal lead removal: experience from EV ICD Pilot, Pivotal, and Continued Access studies. Europace. 2024;26(9). doi: 10.1093/europace/euae225 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Scott PA, Whittaker A, Zeb M, et al. Rates of upgrade of ICD recipients to CRT in clinical practice and the potential impact of the more liberal use of CRT at initial implant. Pacing Clin Electrophysiol. 2012;35(1):73–80. doi: 10.1111/j.1540-8159.2011.03255.x [DOI] [PubMed] [Google Scholar]
- 32.Briosa EGA, Wardak F, Evans J, et al. Evolution of extravascular ICD implantation: shift toward local anesthesia and nurse-led sedation without propofol. Pacing Clin Electrophysiol. 2025;48(10):1203–1207. [DOI] [PubMed] [Google Scholar]
- 33.Russo AM, Chung MK. Is defibrillation testing necessary for implantable transvenous defibrillators?: defibrillation testing is necessary at the time of implantable cardioverter defibrillator implantation. Circ Arrhythm Electrophysiol. 2014;7(2):337–346. doi: 10.1161/CIRCEP.113.000371 [DOI] [PubMed] [Google Scholar]
- 34.Boersma L, Barr C, Knops R, et al. Implant and midterm outcomes of the subcutaneous implantable cardioverter-defibrillator registry: the effortless study. J Am Coll Cardiol. 2017;70(7):830–841. doi: 10.1016/j.jacc.2017.06.040 [DOI] [PubMed] [Google Scholar]
- 35.Sears SF, Conti JB. Quality of life and psychological functioning of icd patients. Heart. 2002;87(5):488–493. doi: 10.1136/heart.87.5.488 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Sears SF, Harrell R, Crozier I, et al. Patient-reported quality of life and acceptance of the extravascular implantable cardioverter-defibrillator: results from pivotal study. J Cardiovasc Electrophysiol. 2024;35(2):240–246. doi: 10.1111/jce.16151 [DOI] [PubMed] [Google Scholar]
- 37.Boersma LVA, Amin A, Clémenty N, et al. Design of a post-market registry for the extravascular implantable cardioverter-defibrillator: the Enlighten study. Heart Rhythm O2. 2024;6(1):64–69. doi: 10.1016/j.hroo.2024.09.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Gold MR, Lambiase PD, El-Chami MF, et al. Primary results from the understanding outcomes with the S-ICD in primary prevention patients with low ejection fraction (untouched) trial. Circulation. 2021;143(1):7–17. doi: 10.1161/CIRCULATIONAHA.120.048728 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Crozier I, Haqqani H, Kotschet E, et al. Three-year chronic follow-up from the pilot study of a substernal extravascular implantable cardioverter-defibrillator. Europace. 2023;25(10). doi: 10.1093/europace/euad301 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Qin D, Filippaios A, Murphy J, et al. Short- and long-term risk of lead dislodgement events: real-world experience from product surveillance registry. Circ Arrhythm Electrophysiol. 2022;15(8):e011029. doi: 10.1161/CIRCEP.122.011029 [DOI] [PubMed] [Google Scholar]
- 41.Knight BP, Clementy N, Amin A, et al. The clinical and economic impact of extended battery longevity of a substernal extravascular implantable cardioverter defibrillator. J Cardiovasc Electrophysiol. 2024;35(2):230–237. doi: 10.1111/jce.16150 [DOI] [PubMed] [Google Scholar]
- 42.Swerdlow C, Gillberg J, Boersma LVA, et al. Extravascular implantable cardioverter-defibrillator sensing and detection in a large global population. JACC Clin Electrophysiol. 2024;10(8):1896–1912. doi: 10.1016/j.jacep.2024.02.033 [DOI] [PubMed] [Google Scholar]
- 43.Moss AJ, Schuger C, Beck CA, et al. Reduction in inappropriate therapy and mortality through ICD programming. N Engl J Med. 2012;367(24):2275–2283. doi: 10.1056/NEJMoa1211107 [DOI] [PubMed] [Google Scholar]
- 44.Arenal A, Proclemer A, Kloppe A, et al. Different impact of long-detection interval and anti-tachycardia pacing in reducing unnecessary shocks: data from the advance iii trial. Europace. 2016;18(11):1719–1725. doi: 10.1093/europace/euw032 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Auricchio A, Schloss EJ, Kurita T, et al. Low inappropriate shock rates in patients with single- and dual/triple-chamber implantable cardioverter-defibrillators using a novel suite of detection algorithms: PainFree SST trial primary results. Heart Rhythm. 2015;12(5):926–936. doi: 10.1016/j.hrthm.2015.01.017 [DOI] [PubMed] [Google Scholar]
- 46.Auricchio A, Hudnall JH, Schloss EJ, et al. Inappropriate shocks in single-chamber and subcutaneous implantable cardioverter-defibrillators: a systematic review and meta-analysis. Europace. 2017;19(12):1973–1980. doi: 10.1093/europace/euw415 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.