Author + information
- Received August 22, 2016
- Revision received November 29, 2016
- Accepted December 8, 2016
- Published online July 17, 2017.
- Ammar M. Killu, MBBSa,
- Niyada Naksuk, MDa,
- Zdeněk Stárek, MD, PhD, MScb,
- Christopher V. DeSimone, MD, PhDa,
- Faisal F. Syed, MBChBa,
- Prakriti Gaba, BSc,
- Jiří Wolf, Ingb,
- Frantisek Lehar, MD, PhDb,
- Martin Pesl, MD, PhD, MScb,
- Pavel Leinveber, Ing, PhDb,
- Michal Crha, MVDr, PhD, MScd,
- Dorothy Ladewig, BSe,
- Joanne Powersf,
- Scott Suddendorff,
- David O. Hodge, MSg,
- Gaurav Satam, MSe,
- Miroslav Novák, MD, PhDb,
- Tomas Kara, MD, PhDb,
- Charles J. Bruce, MDa,
- Paul A. Friedman, MDa and
- Samuel J. Asirvatham, MDa,h,∗ ()
- aDepartment of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota
- bDepartment of Cardiovascular Diseases, St Anne's University Hospital, Brno, Czech Republic
- cMayo Medical School, Rochester, Minnesota
- dUniversity of Veterinary and Pharmaceutical Sciences Brno, Brno, Czech Republic
- eMayo Clinic Ventures, Rochester, Minnesota
- fDepartment of Cardiovascular Surgery, Mayo Clinic, Rochester, Minnesota
- gDepartment of Health Science Research, Mayo Clinic, Jacksonville, Florida
- hDepartment of Pediatrics and Adolescent Medicine, Mayo Clinic, Rochester, Minnesota
- ↵∗Address for correspondence:
Dr. Samuel J. Asirvatham, Department of Cardiovascular Diseases, Mayo Clinic, 200 1st Street SW, Mary Brigh 4-523, Rochester, Minnesota 55905.
Objectives This study aimed to develop a percutaneous defibrillation system with partially insulated epicardial coils to focus electrical energy on the myocardium and prevent or minimize extracardiac stimulation.
Background Epicardial defibrillation systems currently require surgical access.
Methods We tested 2 prototypes created for percutaneous introduction into the pericardial space via a steerable sheath. This testing included a partially insulated defibrillation coil and a defibrillation mesh with a urethane balloon acting as an insulator to the face of the mesh not in contact with the epicardium. The average energy associated with a chance of successful defibrillation 75% of the time was calculated for each experiment.
Results Of 16 animal experiments, 3 pig experiments had malfunctioning mesh prototypes such that results were unreliable; these were excluded. Therefore, 13 animal experiments were analyzed, 6 in canines (29.8 ± 4.0 kg) and 7 in pigs (41.1 ± 4.4 kg). The overall chance of successful defibrillation 75% of the time was 12.8 ± 6.7 J (10.9 ± 9.1 J for canines and 14.4 ± 3.9 J in pigs; p = 0.37). The lowest chance of successful defibrillation 75% of the time obtained in canines was 2.5 J, whereas in pigs it was 9.5 J. The lowest energy resulting in successful defibrillation was 2 J in canines and 5 J in pigs. There was no evidence of coronary vessel injury or trauma to extrapericardial structures.
Conclusions Percutaneous, epicardial defibrillation using a partially insulated coil is feasible and seems to be associated with low defibrillation thresholds. Focusing insulation may limit extracardiac stimulation and potentially lower energy requirements for efficient defibrillation.
The developments of the transvenous pacemaker and implantable cardioverter-defibrillator (ICD) stand as 2 of the greatest innovations in cardiology. Despite their life-saving capabilities, the transvenous approach for implantation is associated with numerous limitations and complications. This includes lead-associated tricuspid valve regurgitation, venous thrombosis, cardiac/vascular perforation, lead thrombus, and cardiac implantable electronic device–associated infection (1–7).
Although some of these complications can be avoided with leadless techniques (8) or subcutaneous placement in the case of ICDs (9), problems remain. Leadless techniques are limited to pacing function only and subcutaneous ICDs are devoid of chronic pacing ability (10,11). Defibrillation coils can be placed epicardially via surgical techniques (12). The surgical approach may be useful in patients with contraindications to transvenous device placement such as venous thrombosis, congenital heart disease–associated anatomic variance and mechanical tricuspid valve. However, surgical placement is associated with significant morbidity, longer duration of hospital stay, and added costs. Therefore, the ability to place epicardial leads without the need for thoracotomy would be a big advance.
Even if these limitations are circumvented, ICD shocks are associated with considerable morbidity secondary to shock-associated pain. Numerous studies have investigated shock pain, although these have predominantly used transvenous defibrillator systems (13–15).
We have previously shown feasibility of partially insulated epicardial pacing leads (16). Thus, we sought to develop an entirely percutaneous defibrillator system coupled with partially insulated epicardial defibrillation leads designed to focus energy on the heart. We hypothesized that this would improve the defibrillation threshold (DFT), thus permitting use of lower shock energy.
Four animals were used for prototype development and improvement. Because only 1 energy level was tested during prototype refinement, these were not included in the overall analysis. We designed and tested 2 distinct prototypes (50 cm long, 12-F size) with partial insulation. These prototypes were built for percutaneous introduction into the pericardial space via a custom, steerable epicardial sheath (patent number: WO 2015/143327 A1; 7,620,458 B2; US 8,315,716 B2) (Figure 1):
• Partially insulated forked lead: The defibrillation coil was made from SS 0.070OD 3-filar coil and partially insulated with a cover of polyether block amide insulation (Pebax, Arkema, King-of-Prussia, Pennsylvania). A radio-opaque marker at the catheter tip helped to define the position with respect to insulation. A modification included a fixation screw made of 0.02-inch-thick stainless steel wire coil wrapped to a 0.075-inch diameter and elongated to 0.15-inch length that may be screwed into the epicardial surface of the heart to improve stability after positioning and defibrillation.
• Partially insulated mesh: The defibrillation mesh was constructed from 0.008-inch stranded nitinol mesh wire. A urethane balloon was attached to the frame and mesh, acting as an insulator to the face of the mesh not in contact with the epicardium.
Electrophysiology laboratory setup
Experiments were performed in dedicated large-animal translational electrophysiology laboratories at Mayo Clinic, Rochester, Minnesota, and at St Anne's University Hospital, Brno, Czech Republic. Electrocardiography recordings were performed using the Cardiolab System (version 6.8.1 release 2, GE Healthcare, Wauwatosa, Wisconsin). High- and low-pass filters were set at 0.05 Hz and 100 Hz, respectively. Sensed signals were gained at 2,500 times as standard.
The study was approved by the Mayo Clinic Institutional Animal Care and Use Committee and the Ethics Committee of the University of Veterinary and Pharmaceutical Sciences, Brno, Czech Republic. Male mongrel dogs were used at Mayo Clinic, and female swine were used at Brno, Czech Republic. Animals were in the fasting state. All animals were instrumented in an identical fashion. Intravenous ketamine (10 mg/kg) and diazepam (0.5 mg/kg) were used for sedation in canines. Intramuscular telazol (5 mg/kg) and xylazine (2 mg/kg) were used for sedation in swine. After sedation, animals were shaved in the groin, neck, and chest areas, where applicable. Isoflurane (1% to 3%) inhalation anesthetic was used for maintenance of general anesthesia. Continuous surface electrocardiography was performed. Sheaths were placed using Seldinger technique. A 9-F sheath was placed in the left femoral artery for continuous blood pressure monitoring. 9-F sheaths were placed in the femoral veins; in dogs, a 12-F external jugular vein sheath was also placed. An intracardiac echocardiography catheter (AcuNav, Siemens, Washington, DC) was used. Animal body temperature was maintained at 38°C using a dorsal water flow heating pad. Anterior percutaneous epicardial access was obtained using the technique of Sosa et al. (17). Two 18-F sheaths were inserted to permit 2 prototype defibrillator leads to be positioned (Figure 2). Prototype leads were introduced into the epicardial space and steered over the right and left ventricles to provide an adequate defibrillation vector using fluoroscopy and intracardiac echocardiographic guidance. The same leads were used across experiments in each individual species.
Ventricular fibrillation (VF) was induced using direct current energy delivered through a catheter placed in the right ventricle. Once VF was induced, it was allowed to persist for 10 s before defibrillation was attempted. Defibrillation was delivered manually at the pre-specified energy using a Maximo II DR (Medtronic, Minneapolis, Minnesota) or ICS 3000 (Biotronic, Berlin, Germany) ICD using a standard biphasic waveform (18). The device was charged pre-emptively during the 10-s VF period so that defibrillation could be delivered exactly at 10 s. If VF persisted despite the first shock, either an internal maximal output shock or an external rescue shock using a standard Zoll defibrillator (Zoll, Chelmsford, Massachusetts) with manual pads was delivered. After successful defibrillation, a 2-min period was observed between to allow the animal to stabilize and recover after therapy. Hemodynamic parameters were monitored continuously. However, laboratory parameters such as pH or electrolytes levels were not monitored routinely. Air in the pericardial space, if present, was documented given that this could adversely affect DFTs.
At the end of the procedure, the animal was humanely killed by VF induction using direct current energy. Necropsy was performed immediately upon death to assess the pericardial access and document injury to cardiac and extracardiac structures.
Individual estimates of the energy associated with a chance of successful defibrillation 75% of the time (ED75) was calculated by fitting a logistic regression model to each individual animal. Namely, the combination of the energy and success/failure values on each animal was used to estimate that individual estimate of the ED75 for that animal. The overall ED75 values were calculated from the estimates from those individual models. In 9 animals there were not enough failures to get a fit from those models to converge to 1 estimate. In those cases, we estimated the ED75 as the energy level where the number of successful shocks begins to decrease. The energies in each of these cases were stepped down in sequence, making the value fairly easy to estimate. Overall ED75 values were summarized using means and standard deviations. A comparison of those values between groups was completed with 2-sample Student t tests. A 2-tailed p value of <0.05 was considered significant. Analysis was performed using SAS version 9.4 (SAS Institute Inc., Cary, North Carolina).
Of 16 animal experiments, 3 pig experiments had malfunctioning mesh prototypes such that the results were highly variable and unreliable, and these were excluded. Therefore, 13 animal experiments were analyzed. This included 6 canine (mean weight: 29.8 ± 4.0 kg) and 7 pig (mean weight: 41.1 ± 4.1 kg) experiments. Successful dual pericardial access was obtained in all 13 experiments with no acute complications.
The average ED75 by species was 10.9 ± 9.1 J for canines and 14.4 ± 3.9 J in pigs (p = 0.37). Overall, the ED75 was 12.8 ± 6.7 J. Noteworthy is that the lowest ED75 obtained in canines was 2.5 J whereas in pigs it was 9.5 J. Meanwhile, the lowest energy resulting in successful defibrillation was 2 J in canines and 5 J in pigs. Tables 1 and 2⇓⇓ show the individual and overall ED75 according to species.
In total, 118 shocks were delivered using the 2-finger lead and 12 were delivered using the mesh lead. The 2-finger lead was used in 11 experiments (4 in canines and 7 in pigs) and the mesh was used in 5 experiments (2 canine and 3 pig). In the 3 pig experiments excluded, the mesh lead was used exclusively. Therefore, after exclusion, the mesh lead was only used in 2 canine experiments. With this, there was no significant difference between lead performance overall. In the 11 animals with a 2-finger lead, the mean ED75 was 13.7 ± 6.6 J; meanwhile, the ED75 using the mesh lead (in the 2 canines) was 7.7 ± 7.3 J (p = 0.26). In addition, there was no difference in lead performance in canine experiments. The mean ED75 in the 4 dogs with the 2-finger lead and 2 dogs with the mesh lead was 12.5 ± 10.5 J and 7.7 ± 7.3 J (p = 0.60). The defibrillation lead with the active-fixation screw was used in 2 experiments. When deployed, the lead did not dislodge despite multiple defibrillation shocks (Figure 3). However, retraction and removal of the lead was challenging when attempted.
At necropsy, there was occasional evidence of epicardial damage especially with repetitive defibrillation therapy as would be expected (Figure 4). However, there was no evidence of coronary vessel injury, hemorrhage, or trauma to extrapericardial structures. The animals did twitch during energy delivery.
ICD system–associated complications are caused mainly by the implanted transvenous lead and include thrombus formation, thromboembolism, venous occlusion, endovascular infection necessitating extraction, and tricuspid regurgitation (1–6). The adverse consequences and limitations associated with endocardial defibrillator systems have directed investigation into alternative methods of defibrillation. Whereas epicardial lead placement originally required open surgical techniques for implantation, less invasive methods have become increasingly attractive and include subxiphoid pericardial windows and more recently percutaneous implantation (19–21). Furthermore, a recent report highlighted experience with a thoracoscopic method in a piglet model (22).
Our 2-center study demonstrates that defibrillation using our partially insulated defibrillation leads is feasible and results in successful defibrillation despite extremely low energy (as low as 2 J in this study). Furthermore, successful defibrillation despite repeated shocks using the functioning leads suggests that the prototypes are durable.
These data support the basis for movement toward a minimally invasive, extravascular defibrillation system, increasing therapeutic options in patients with structural heart disease that require defibrillator therapy.
Access options in patients with structural heart disease
Epicardial defibrillator systems are highly desirable, particularly in individuals with structural heart disease. This includes those with congenital heart anomalies (e.g., single-ventricle defects and intracardiac shunts) and situations in which anatomy prohibits using a transvenous approach. Furthermore, up to 2% and 20% of patients implanted with an endocardial ICD may develop cardiac implantable electronic device infections or have worsening of tricuspid valve regurgitation, respectively (23,24). Although a subcutaneous ICD can help to avoid the problems associated with transvenous implantation, lack of pacing capability due to the remote location from the heart remains a problem especially in such populations who have an high rate of conduction system disease. Furthermore, the DFT values in the subcutaneous system are high, requiring a larger pulse generator, and if implanted in a child, may increase as the child grows. Current epicardial systems require surgical implantation, which limits their desirability. A percutaneous epicardial implantation procedure would help considerably reduce procedural-related morbidity. However, it deserves mention that, although still feasible, percutaneous pericardial access may be challenging and catheter positioning limited in those with previous cardiac surgery (25).
A mandatory requirement of ICDs is to have adequate safety margins for defibrillation. The energy level for successful defibrillation was as low as 2 J in our experiments. In addition, the overall ED75 was 12.8 J, suggesting the potential for having an adequate safety threshold using our devices. This would have important clinical implications in individuals with high DFTs, which may occur secondary to antiarrhythmic drug therapy, right-sided device implants and hypertrophic cardiomyopathy and may mitigate the need for a subcutaneous array (26). The ED75 attained might be higher than with a final commercial product because our prototype sheaths were not designed to preclude air from entering the pericardium at this stage effectively. Our prototype required 2 large sheaths, which permit more air entry than typically observed with conventional epicardial based procedures. Although measures to alleviate this were deployed, it is still possible that this affected DFTs (27). Future work using a smaller, universal sheath for deployment and lower French lead would help to remedy this problem.
We found no difference in performance between the 2 prototypes after excluding 3 experiments in which the mesh lead provided unreliable results. This likely represented faulty leads for those 3 experiments, although species-specific differences in effect cannot be excluded. Taken together, this suggests that further refinements in the mesh lead are necessary. In addition, given that we did not monitor serum electrolytes or blood gas analysis, derangement in these parameters (especially with repeated shocks) could also affect the results. The 2-min recovery period was intended to assuage this, however. Although DFTs may have still been adversely affected, our results remained satisfactory.
Less painful defibrillation system
Although current defibrillation systems reduce mortality, a major limitation is shock-associated pain (28). Although ICD systems have adopted sophisticated features such as antitachycardia pacing and enhanced detection algorithms to minimize the need for unnecessary shock delivery, shocks remain painful when delivered. In part, this is due to the shock vector directly traversing the chest wall and subcutaneous tissue. Other groups have tried to minimize chest wall stimulation; for example, one used a shielding device protecting the musculature and another delivered tetanizing pre-pulses to gradually stimulate muscle (13–15). Although intriguing, these novel methods also had some shortcomings, namely the need for thoracotomy in one; both were used in conjunction with an endocardial system. By comparison, our leads are introduced percutaneously into the pericardial space, precluding the need for thoracotomy, thereby negating some of the shortcomings of these investigational techniques. Additionally, our device provides a direct vector occupied only by the heart that may result in a lower DFT. However, because lower DFT is only a minor factor in mitigating shock-related pain, further refinements are required in the quest to reduce chest discomfort with shocks, if possible (15,29).
Our results are best appreciated in the context of a proof-of-concept study, while recognizing several important limitations. The numbers of shock deliveries between animal studies was not uniform; therefore, estimates may be skewed. Because defibrillation was performed on anesthetized animals, DFT may have been affected. Because our aim was to determine the feasibility of a novel defibrillation system, routine measurement of pH and serum electrolytes was not performed. Although care was taken to minimize the presence of intrapericardial air, smaller amounts not detected by fluoroscopy may still have been present. Given the acute nature of our study, we do not have information on chronic lead stability, long-term efficacy, or complications such as pericarditis or pericardial constriction. Although the percutaneous approach precludes development of complications associated with transvenous lead placement, it may itself predispose to serious complications, including cardiac tamponade from the access procedure. However, complication rates associated with percutaneous epicardial access are improving with increased procedural experience (30). In addition, data from surgically implanted epicardial leads suggests that later development of this complication is rare (31,32). In our experiments, there were no instances of significant pericardial effusion or cardiac tamponade, coronary vessel injury, or trauma to extrapericardial structures. Furthermore, although we demonstrated the feasibility of implantation, leads may be difficult to extract, if required, given their large surface area and possible adhesion formation. Formal measures of muscle contraction force were not used and a control arm comparing a device without insulation was not used. Chronic animal studies, in which subcutaneous implantation of the generator (in subpectoral, pre-pectoral, or subaxillary folds depending on clinical and patient factors) and epicardial screw fixation are implemented to test lead stability and performance, potential complications, and formal measures of muscle contraction as a surrogate for pain, are required. Finally, differences in structure and position of the heart between the animal models and humans may cause different results.
Percutaneous, epicardial defibrillation using a partially insulated coil is feasible and associated with low DFTs. However, in addition to chronic studies, further refinements are required to maximize efficacy and reliability.
COMPETENCY IN MEDICAL KNOWLEDGE: Transvenous ICD implantation is associated with several limitations. Although some of these limitations can be circumvented with newer techniques, problems remain. Leadless techniques are limited to pacing function only and subcutaneous ICDs are devoid of chronic pacing ability. Although defibrillation coils can be placed epicardially via surgical techniques, this is associated with significant morbidity. Therefore, the ability to place epicardial defibrillation leads percutaneously minimizes surgical morbidity and bypasses the limitations associated with transvenous and subcutaneous lead placement.
TRANSLATIONAL OUTLOOK: Our novel, partially insulated defibrillator leads are placed in the pericardial space percutaneously by a standard subxiphoid access. The present experiments have shown that using this system, we can successfully defibrillate the heart with relatively low DFTs. This is seen as a critical step toward further testing to assess whether such an insulated directional epicardial lead system can minimize or eliminate pain during defibrillation shocks.
The Mayo Clinic owns intellectual property related to the investigated technology reported in the manuscript under WO2015/143327A1; 7,620,458B2; US8,315,716B2. Neither Mayo Clinic nor the inventors will receive compensation for the use of this product with Mayo Clinic patients.
Funding support provided by the project no. LQ1605 from the National Program of Sustainability II (MEYS CR) and by project FNUSA-ICRC no. CZ.1.05/1.1.00/02.0123 (OP VaVpI).
Dr. DeSimone has received a U.S. National Institutes of Health training grant, HL007111. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Drs. Killu and Naksuk contributed equally to this work.
All 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
- defibrillation threshold
- chance of successful defibrillation 75% of the time
- implantable cardioverter defibrillator
- ventricular fibrillation
- Received August 22, 2016.
- Revision received November 29, 2016.
- Accepted December 8, 2016.
- 2017 American College of Cardiology Foundation
- Lin G.,
- Nishimura R.A.,
- Connolly H.M.,
- Dearani J.A.,
- Sundt T.M. 3rd.,
- Hayes D.L.
- Haghjoo M.,
- Nikoo M.H.,
- Fazelifar A.F.,
- Alizadeh A.,
- Emkanjoo Z.,
- Sadr-Ameli M.A.
- Sohail M.R.,
- Uslan D.Z.,
- Khan A.H.,
- et al.
- Lee R.C.,
- Friedman S.E.,
- Kono A.T.,
- Greenberg M.L.,
- Palac R.T.
- DeSimone C.V.,
- Friedman P.A.,
- Noheria A.,
- et al.
- Novák M.,
- Kamarýt P.,
- Dvořák I.,
- Mach P.,
- Vykypěl T.,
- Müllerová J.
- Lewis G.F.,
- Gold M.R.
- Eberhardt F.,
- Bode F.,
- Bonnemeier H.,
- et al.
- Hunter D.W.,
- Tandri H.,
- Halperin H.,
- Tung L.,
- Berger R.D.
- Syed F.F.,
- DeSimone C.V.,
- Ebrille E.,
- et al.
- Gliner B.E.,
- Lyster T.E.,
- Dillion S.M.,
- Bardy G.H.
- Jacob S.,
- Lieberman R.A.
- Clark B.C.,
- Davis T.D.,
- El-Sayed Ahmed M.M.,
- et al.
- Prutkin J.M.,
- Reynolds M.R.,
- Bao H.,
- et al.
- Delling F.N.,
- Hassan Z.K.,
- Piatkowski G.,
- et al.
- Killu A.M.,
- Ebrille E.,
- Asirvatham S.J.,
- et al.
- Hsu J.C.,
- Marcus G.M.,
- Al-Khatib S.M.,
- et al.
- Killu A.M.,
- Sugrue A.M.,
- Mulpuru S.K.,
- et al.
- Rezq A.,
- Basavarajaiah S.,
- Latib A.,
- et al.
- McAloon C.J.,
- Anderson B.M.,
- Dimitri W.,
- et al.