Author + information
- Ian Crozier, MB ChBa,∗ (, )
- Haris Haqqani, MBBS, PhDb,
- Emily Kotschet, MBBSc,
- David Shaw, MB ChBa,
- Anil Prabhu, MChb,
- Nicholas Roubos, MBBS, BMedScid,
- Jeffrey Alison, MBBSc,
- Iain Melton, MB ChBa,
- Russell Denman, MBBSb,
- Tina Lin, MBBSd,
- Aubrey Almeida, MBBSc,
- Bridget Portway, BSe,
- Robert Sawchuk, BSEE, MBAe,
- Amy Thompson, MS, MBAe,
- Lou Sherfesee, PhDe,
- Samuel Liang, MbiomedEe,
- Linnea Lentz, DVM, PhDe,
- Paul DeGroot, MSe,
- Alan Cheng, MDe and
- David O’Donnell, MBBSd
- aDepartment of Cardiology, Christchurch Hospital, Christchurch, New Zealand
- bDepartment of Cardiology, The Prince Charles Hospital, Brisbane, Australia
- cDepartment of Cardiac Rhythm Services, Monash Medical Centre, Clayton, Australia
- dDepartment of Cardiology, The Austin Hospital, Heidelberg, Australia
- eMedtronic, Mounds View, Minnesota
- ↵∗Address for correspondence:
Dr. Ian Crozier, Christchurch Hospital, Department of Cardiology, 2 Riccarton Avenue, PO Box 4710, Christchurch 8140, New Zealand.
Objectives The aim of this study was to evaluate the safety and performance of an extravascular (EV) implantable cardioverter-defibrillator (ICD).
Background Limitations of existing transvenous and subcutaneous ICD systems include lead reliability and morbidity issues associated with ICD lead implantation in the vasculature or lack of pacing therapies (e.g., antitachycardia pacing) in subcutaneous systems. The EV defibrillator uses a novel substernal lead placement to address these limitations.
Methods This was a prospective, nonrandomized, chronic pilot study conducted at 4 centers in Australia and New Zealand. Participants were 21 patients referred for ICD implantation. Patients received EV ICD systems. Data collection included major systemic and procedural adverse events, defibrillation testing at implantation, and sensing and pacing thresholds.
Results Among 20 patients who underwent successful implantation, the median defibrillation threshold was 15 J, and 90% passed defibrillation testing with a ≥10-J safety margin. Mean R-wave amplitude was 3.4 ± 2.0 mV, mean ventricular fibrillation amplitude was 2.8 ± 1.7 mV, and pacing was successful in 95% at ≤10 V. There were no intraprocedural complications. Two patients have undergone elective chronic system removal since hospital discharge. In the 15 patients presently implanted, the systems are stable in long-term follow-up.
Conclusions This first-in-human evaluation of an EV ICD demonstrated the feasibility of substernal lead placement, defibrillation, and pacing with a chronically implanted system. There were no acute major complications, and pacing, defibrillation, and sensing performance at implantation were successful in most patients. (Extravascular ICD Pilot Study [EV ICD]; NCT03608670)
Implantable cardioverter-defibrillators (ICDs) have been the cornerstone in the management of patients at risk for sudden cardiac death for the past 30 years (1). With advancements in technology and improvements in implantation techniques, the ICD has evolved into a system that can be fully implanted with a transvenous lead in the right ventricle that is connected to a pulse generator located in the infraclavicular subcutaneous (SQ) pectoral region (2). ICD systems can be implanted safely and have demonstrated defibrillation efficacy of 88.8% to 91.2% when testing with a 10-J safety margin at the time of implantation (3). Although transvenous ICD systems have proven defibrillation efficacy, there are ongoing concerns related to defibrillation lead failure, vascular obstruction, and device-related systemic infections (4–6). Transvenous lead complications and resulting lead extraction have significant associated morbidity and mortality (7–9). Additionally, some patients have venous obstruction or cardiac anatomies that prevent transvenous lead implantation.
Recent interest in alternative solutions has resulted in the development of the SQ ICD, which allows the delivery of defibrillation and cardioversion with the lead and device fully implanted in the SQ plane (10,11). However, the SQ ICD system has limitations: it does not provide antitachycardia pacing (ATP) or pacing support aside from post-shock pacing (12–14), and it is a high-energy system, resulting in a large and often uncomfortable device with less longevity than transvenous defibrillators (15).
The extravascular (EV) ICD was developed to avoid the long-term lead risks of transvenous defibrillators and to address the limitations of SQ defibrillators. Prior case reports of leads placed in the substernal space (16–20), coupled with more recent feasibility studies (21–24), have shown that a lead in the substernal space can achieve satisfactory sensing and pacing as well as provide defibrillation with lower energies than required by the SQ system (10). This report describes the first chronic human experience of the EV ICD.
This EV ICD pilot study was a prospective, nonrandomized, chronic, first-in-human study conducted at 4 centers in Australia and New Zealand. The study complied with the Declaration of Helsinki (25) and was approved by the regulatory authorities and/or the ethics committees of the participating centers: Austin Health Human Research Ethics Committee (Australia) and Northern B Health and Disability Ethics Committee (New Zealand). All patients gave informed consent prior to study enrollment.
The EV ICD lead and defibrillator
The EV ICD is a purpose-built system (Figure 1) designed to use the substernal space. The device is the size of a transvenous defibrillator (33 cm3), delivering up to 40 J and with sensing and pacing circuits designed for substernal therapy. The lead has an epsilon shape with 2 pace/sense electrodes and 2 defibrillation coil segments (4 cm each) that are tied together for defibrillation purposes to form an overall 8-cm defibrillation coil. There are 3 sensing and 3 pacing vectors available in the EV ICD system (Figure 2). Sensing vectors include a near-field vector between the 2 ring electrodes and 2 far-field sensing vectors from each ring electrode to the ICD device. Pacing vectors include the ring-to-ring vector, the coil-to-coil vector, and a vector from the distal ring to the proximal coil segment.
Eligible patients were those undergoing ICD implantation with a Class I or IIa indication on the basis of current clinical practice guidelines (26,27). Patients were excluded if there was an indication for bradycardia pacing or cardiac resynchronization therapy; existing implanted device or prior cardiac implanted device; any condition that would compromise access to the anterior mediastinum, including previous sternotomy, prior chest radiotherapy, mediastinitis, or severe pericarditis; any condition that would prevent defibrillation testing, including left ventricular thrombus, severe aortic stenosis, severe chronic obstructive lung disease, decompensated heart failure, or left ventricular ejection fraction <20%; any abnormality that would increase implantation risk, including severe obesity, marked right ventricular dilation, marked sternal abnormality, gastrostomy tube, or hiatal hernia that distorted mediastinal anatomy; anticoagulation that could not be interrupted; and active infection or renal dialysis.
All investigators underwent training in EV ICD implantation using simulators and both live animal and human cadaver models to establish skills in safe access and tunneling within the substernal space. Implantations were performed by electrophysiologists, and a cardiac surgeon provided mentoring through the first 5 implantation procedures. All implantations were performed in a cardiac catheterization laboratory or hybrid laboratory with general anesthesia. External defibrillation pads were placed outside the surgical field for rescue defibrillation if required. Patients’ computed tomographic scans were reviewed pre-procedurally to evaluate relevant anatomy. A small incision (approximately 3 cm) was made between the inferior point on the xiphoid and the left costal margin to access the substernal space and was extended to the rectus fascia, which was divided. Blunt dissection was performed to the rectus fascia on the posterior border of the left xiphoid rib junction and then advanced through the diaphragmatic attachments to enter the anterior mediastinum. The tunneling tool (Figure 1) with peel-away sheath was introduced into the substernal space and advanced under lateral fluoroscopy to ensure that the tip of the tunneling tool was adjacent to the posterior face of the sternum to avoid cardiac injury. A tunneling path was created to the upper border of the heart silhouette using anteroposterior fluoroscopy, as marked by the lower margin of the carina, and to the junction between the right and left pleural reflections. A defibrillation lead was inserted into the substernal space via the peel-away introducer sheath once the tunneling tool had been removed. Acute sensing measurements were collected with the expectation of R-wave amplitudes ≥1 mV. The lead was then secured to the rectus sheath using an anchoring sleeve and sutures. The proximal portion of the lead was then tunneled to an SQ device pocket on the left side of the chest close to the left midaxillary line. The device was placed either over the serratus muscle or between serratus and latissimus dorsi, and the device was sutured within the pocket. Following hemostasis, the wounds were closed using standard closure techniques. Electrical parameters, including sensing, impedance, and pacing in multiple vectors, were measured through the device by telemetry.
Patients were observed in the hospital overnight and usually discharged 24 h following the procedure. Patients were followed up at 2 weeks, 4 to 6 weeks, and 3 months and remain under follow-up. At follow-up, devices were interrogated, sensing and pacing tolerability testing performed, and chest radiography (day 1, week 2, weeks 4 to 6, and 3 months) and chest computed tomography (3 months) performed.
The primary efficacy endpoint was defibrillation testing success at implantation. Ventricular fibrillation (VF) was induced via the device. Defibrillation efficacy was characterized at implantation by inducing, detecting, and converting VF episodes. Implantation required termination of VF with either a single 20-J shock or on 2 consecutive episodes with a 30-J shock. If the patient was successfully defibrillated at 20 J, defibrillation efficacy was assessed at 15 J.
The primary safety endpoint was to characterize any complication related to the EV ICD system or procedure that resulted in death, system revision, hospitalization, prolongation of a hospitalization, or permanent loss of defibrillation function due to device dysfunction. Freedom from such complications was evaluated through 90 days.
Descriptive statistics were used for demographics and medical history and summary of most endpoints. The 90-day rate of freedom from major complications was generated using the Kaplan-Meier method, with 95% confidence bounds also generated.
Twenty-six patients (Figure 3 and Central Illustration) were enrolled, with 21 proceeding to attempted implantation of an EV ICD. (Reasons for not proceeding to implantation included withdraw of consent in 2, unsuitable anatomy on computed tomography in 1, resolution of implantation indication in 1, and implantation contraindication in 1.)
Among the 21 patients undergoing attempted implantation, 81% were men (Table 1), the age range was 22 to 77 years, and 86% had primary ICD indications. The underlying cardiac conditions were ischemic cardiomyopathy (n = 7), nonischemic cardiomyopathy (n = 5), hypertrophic cardiomyopathy (n = 6), and other conditions (n = 3).
Total time for substernal lead placement was measured from the time of first incision to the time of final lead placement, and the median was 25 min (interquartile range: 21 to 35 min). The median total procedure time from first incision to final suture was 85 min (interquartile range: 78 to 104 min). Median total fluoroscopy time was 4.2 min.
In all 21 patients, defibrillation testing was attempted at implantation. However, 1 patient could be induced into VF only one time and had an R-wave amplitude <1 mV on the following day, and the device was explanted. This patient was not included in the defibrillation data but was included in the sensing data group of 20 patients.
In 1 additional patient, the lead could not be advanced fully into the mediastinal space because of unexpected mediastinal fibrosis. Although this patient underwent defibrillation testing and was successful, all shocks were delivered manually because of poor sensing performance, and the device was explanted. However, this patient successfully completed the defibrillation testing protocol and contributed to the defibrillation dataset (n = 20).
Among the 20 patients who completed defibrillation testing, 18 (90%) were able to be converted to sinus rhythm with 15 J (n = 11), 20 J (n = 4), or 30 J in 2 consecutive terminations (n = 3) as required per protocol (Figure 4). Additionally, 2 patients who successfully defibrillated at 15 J were tested at 10 J, and both were successful at 10 J. The 2 patients who did not pass defibrillation testing underwent explantation, with subsequent implantation of transvenous defibrillators prior to discharge; transvenous defibrillation succeeded at 25 J in 1 patient, while the other did not have the transvenous system tested.
In the 20 patients with completed lead implantation, the mean R-wave amplitude was 3.4 ± 2.0 mV, with all R-wave amplitudes exceeding 1 mV, while the mean amplitude of induced VF was 2.8 ± 1.7 mV. All 20 patients detected VF at sensitivity ≥0.3 mV (Figure 5). An example of induction, detection, and termination of VF recorded during implant testing is shown (Figure 6).
Pacing threshold at implantation was assessed at up to 10 V and 8-ms pulse width in all but 1 patient. Pacing capture was achieved in 19 of 20 patients (5.4 ± 2.2 V), while in the remaining patient, testing was performed only to 8 V, without capture.
No major complications occurred during the implantation procedures. Within 90 days post-operatively, 6 EV ICD adverse events had occurred. One patient experienced an inappropriate shock 78 days post-implantation because of P-wave oversensing that occurred when the lead tip deflected toward the right atrial appendage. The system was subsequently explanted at 85 days post-implantation without complication using only simple traction for lead removal; the patient then received a SQ defibrillator. The 90-day rate of freedom from systemic or procedural major complication was 94.1% (95% confidence interval: 83.6% to 100%). In addition to the single instance of inappropriate shock, 2 patients reported inspiratory discomfort post-operatively, and 3 had minor wound issues (2 with swelling or impaired healing and 1 with superficial wound infection at the xiphoid incision site with minor purulent discharge, which resolved with an antibiotic course and a change of dressing), all of which resolved without sequelae.
Beyond 90 days, 1 patient with arrhythmogenic right ventricular dysplasia underwent investigator-initiated elective chronic defibrillation testing at 3 months and had the system replaced by an SQ defibrillator. At implantation, this patient passed testing by defibrillating in 2 consecutive episodes at 30 J; however, at elective retesting, there was successful defibrillation at 40 J, but not 30 J, using a more limited defibrillation testing protocol. The system was removed at 114 days post-implantation without complication using simple traction for lead removal.
All actively implanted patients remain in follow-up. The results reported herein are limited to follow-up through 3 months.
Observations from chronic follow-up that were not associated with adverse events included the following. There was 1 patient with spontaneous ventricular tachycardia. At the programmed output, ATP did not consistently capture the ventricle, and ventricular tachycardia was terminated by appropriate shock. The pacing output for ATP was thereafter increased.
In 1 patient, the tip of the lead was discovered post-procedure to be implanted in the left pleural space. Electrical performance and lead position have been stable and have not caused discomfort; no interventions have occurred or are expected.
During pacing threshold testing at follow-up visits, all patients reported awareness of cardiac pacing, with sensations variably described as being of mild to moderate intensity. However, the pacing was sufficiently tolerable to allow programming of ATP therapy.
Fifteen patients remain under follow-up to date. Their systems are active and leads stable, and their devices are very well tolerated. They have not experienced any further events related to the system.
This EV ICD first-in-human chronic study demonstrates the ability to position a substernal defibrillation lead and to achieve high effectiveness for acute defibrillation, pacing, and sensing.
The EV ICD is a purpose-built device designed to deliver defibrillation and pacing therapies from the EV substernal space and to avoid the limitations of currently available transvenous and SQ defibrillators (4–15). As demonstrated through this first-in-human experience and previous feasibility evaluations, the EV ICD lead proximity to the heart results in a lower energy requirement for cardiac defibrillation and provides potential for ATP and asystole support pacing features not available in the SQ ICD, while offering a truly EV system placement (21–24).
Several key observations were made from this first-in-human experience. First, despite the novel technique and approach to lead implantation into the substernal space, no acute procedure-related major complications were observed. Although implantation of transvenous defibrillators is associated with significant risks of implantation of up to 10% (6), potential risks related to the EV ICD procedure were mitigated by a number of measures. Initially, a series of proof-of-concept studies were performed in which leads were tunneled into the substernal space with surgical backup (21–24). On the basis of the observations from these studies, techniques were developed to minimize cardiac risk from lead placement, particularly the use of blunt dissection to safely access the substernal space and the use of lateral and anteroposterior fluoroscopy during the procedure. All implanting electrophysiologists underwent training using simulators and live animal and human cadaver models. Cardiac surgeons initially proctored the electrophysiologists in accessing the substernal space. Although 1 lead was inadvertently implanted within the pleural cavity, this did not result in pneumothorax or other complications, and functionality of the system was not compromised.
Second, the EV ICD was able to sense normal rhythm with an R-wave amplitude >1 mV in all patients at implantation, though in the patient in whom VF was induced only once, the R-wave amplitude decreased to <1 mV in next-day testing. However, all devices in which the lead was successfully implanted in the desired location, including the patient with the small R wave, detected VF at 0.3 mV, allowing a 2-fold sensing safety margin over the nominal sensitivity of 0.15 mV and a margin of 4-fold for the maximum sensitivity of 0.075 mV. Although T-wave oversensing has been an issue with the SQ defibrillator (28), especially for patients with hypertrophic cardiomyopathy (29), significant T-wave oversensing was not observed in this first-in-human study, nor was screening required to exclude patients with large T waves or hypertrophic cardiomyopathy. Although P-wave oversensing was observed in a single patient subsequent to lead deflection toward the right atrial appendage, a more inferior lead placement may reduce the risk for P-wave oversensing in future implantations.
Third, the defibrillation performance was within expectations and similar to previous reports on transvenous ICD systems (3). A median defibrillation threshold of 15 J was achieved, which is similar to transvenous defibrillation energies and compares favorably with the mean defibrillation threshold of 36.6 J in the SQ defibrillator (10). The EV ICD first-in-human study required successful termination of induced VF with a minimum of a 10 J defibrillation safety margin, a target that was achieved in 90% of patients and is comparable with conventional devices. The recent UNTOUCHED (Understanding Outcomes With the S-ICD in Primary Prevention Patients With Low EF) study of the SQ defibrillator demonstrated that 93.5% of patients were able to defibrillate with a 15-J safety margin on a single test (30), while with transvenous defibrillators, the initial efficacy of a 10-J safety margin on 2 consecutive inductions prior to lead revision was 88.8% to 92.1% (3). It is not universally appreciated that all patients do not pass defibrillation testing with conventional defibrillators. Often the device is implanted with the presumption that 6 full-energy shocks will terminate ambulatory VF despite not passing the defibrillation efficacy target at implantation. In this study, devices were explanted in patients who did not pass implant testing.
Fourth, far-field pacing from the substernal space was successfully achieved in all but 1 patient. Although all patients were aware of pacing and described the sensation as mild to moderate in intensity, pacing was sufficiently tolerable to allow the programming of ATP.
Fifth, although the stability and performance of the system were adequate in most patients, 2 patients underwent system explantation, including 1 patient with relatively minor lead displacement that resulted in P-wave oversensing and inappropriate shock and 1 patient with arrhythmogenic right ventricular dysplasia whose chronic defibrillation testing could not ensure that there was a 10-J safety margin available. Longer term follow-up in a larger cohort will be required to determine if these are isolated issues; all remaining patients remain stable. Both explanted leads were easily removed at 85 and 114 days using simple traction for lead removal and without complications.
Overall, 15 of the 21 patients in whom implantation was attempted retained the systems at 3 months. Reasons for explantation included limited inducibility and poor sensing in 1 patient, 1 failed implantation due to mediastinal fibrosis, failed defibrillation testing in 2 patients, and 2 elective explantations due to lead dislodgement and failed repeat defibrillation testing. Although this retention rate is considerably lower than with conventional devices, it should be considered in the light of rigorous requirements in place for implantation as required per the study investigative protocol and the fact that the implantation procedure and device are novel and still in evolution.
At present, only short-term follow-up data in a small cohort of predominantly male patients from a single geographic region are available. Larger, longer term evaluation will be needed to address, for example, the long-term sensing performance of the system and detection algorithms, whether predictors exist to ascertain probable defibrillation efficacy prior to implantation, how effectively ATP from a lead in this configuration performs relative to transvenous systems, and the extractability of the EV ICD system.
This first-in-human evaluation of the EV ICD system demonstrated the feasibility of substernal lead placement, defibrillation, and pacing with a chronically implanted system. There were no acute major complications, and pacing, defibrillation, and sensing performance at implantation were successful in the majority of patients. On the basis of this favorable first-in-human experience, a worldwide pivotal study has already begun, with results forthcoming.
COMPETENCY IN MEDICAL KNOWLEDGE: The EV ICD system with novel lead placement in the substernal space provides an implantable defibrillator option for patients at risk for sudden cardiac death who could benefit from EV system placement and pacing therapies such as ATP in a single implantable device.
TRANSLATIONAL OUTLOOK: Current SQ defibrillator technology does not allow pacing and requires high-energy shock with a large device. The future novel EV ICD platform may enable physicians to provide bradycardia and ATP options and to defibrillate the heart with energy levels similar to those of current transvenous ICD platforms. A larger scale clinical study is required to more fully characterize the EV ICD system.
The authors thank all the investigators of the first-in-human EV ICD study, as well as the patients who consented to be enrolled.
This work was sponsored in its entirety by Medtronic. Dr. Crozier is a consultant for and has received research support and fellowship support from Medtronic; and has received grants from Boston Scientific. Dr. O’Donnell is a consultant for Medtronic and Abbott. Dr. Haqqani is a consultant for Medtronic, Abbott, and Boston Scientific. Drs. Kotschet and Shaw are consultants for Medtronic. Dr. Denman has received grants from Medtronic and Boston Scientific. Ms. Portway, Mr. Sawchuk, Ms. Thompson, Dr. Sherfesee, Mr. Liang, Dr. Lentz, Mr. DeGroot, and Dr. Cheng are employees of Medtronic. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- antitachycardia pacing
- Implantable Cardioverter Defibrillator
- Sudden Cardiac Death
- ventricular fibrillation
- Received February 11, 2020.
- Revision received May 15, 2020.
- Accepted May 21, 2020.
- 2020 The Authors
- Bardy G.H.,
- Johnson G.,
- Poole J.E.,
- et al.
- Neuzner J.,
- Hohnloser S.H.,
- Kutyifa V.,
- et al.
- Kleemann T.,
- Becker T.,
- Doenges K.,
- et al.
- Bongiorni M.G.,
- Kennergren C.,
- Butter C.,
- et al.
- Weiss R.,
- Knight B.P.,
- Gold M.R.,
- et al.
- Boersma L.,
- Barr C.,
- Knops R.,
- et al.
- Burke M.C.,
- Gold M.R.,
- Knight B.P.,
- et al.
- Brouwer T.F.,
- Yilmaz D.,
- Lindeboom R.,
- et al.
- Theuns D.A.M.J.,
- Crozier I.G.,
- Barr C.S.,
- et al.
- Hata H.,
- Sumitomo N.,
- Nakai T.,
- Amano A.
- Do K.,
- Lee C.,
- Kiankhooy A.,
- Chang P.,
- Doshi R.
- Boyle T.A.,
- Cohen J.,
- Carrillo R.
- Bhagwandien R.E.,
- Kik C.,
- Yap S.-C.,
- Szili-Torok T.
- Boersma L.V.A.,
- Merkely B.,
- Neuzil P.,
- et al.
- Chan J.Y.S.,
- Lelakowski J.,
- Murgatroyd F.D.,
- et al.
- Sholevar D.P.,
- Tung S.,
- Kuriachan V.,
- et al.
- Brouwer T.F.,
- Smeding L.,
- Berger W.R.,
- et al.
- Priori S.G.,
- Blomstrom-Lundqvist C.,
- Mazzanti A.,
- et al.
- Al-Khatib S.M.,
- Stevenson W.G.,
- Ackerman M.J.,
- et al.
- Lambiase P.D.,
- Gold M.R.,
- Hood M.,
- et al.
- Boersma L.V.,
- El-Chami M.F.,
- Bongiorni M.G.,
- et al.