SEARCH

SEARCH BY CITATION

Keywords:

  • electrodes;
  • heart failure;
  • implantable cardioverter defibrillator;
  • subcutaneous ICD;
  • sudden cardiac death

Subcutaneous Defibrillation

  1. Top of page
  2. Subcutaneous Defibrillation
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Clinical Implications
  8. Study Limitations
  9. Acknowledgments
  10. References

Introduction

A purely subcutaneous implantable cardioverter defibrillator (ICD) requires higher energy but may be an effective alternative to transvenous ICDs to deliver lifesaving therapies.

Objective

To identify combinations of anteroposterior subcutaneous shock pathways and waveforms with defibrillation efficacy comparable to transvenous ICDs.

Methods

Defibrillation testing was performed in 141 patients temporarily implanted with an active can emulator and subcutaneous coil electrodes. The patients were subdivided into 5 groups within 2 study phases. In all groups, a posterior electrode was positioned with its tip close to the spine. In the first study phase, 2 different can locations were evaluated: (1) an inframammary pocket (IM-1–750), or (2) a conventional infraclavicular pocket (IC-1–750). In both cases, a 70 J biphasic shock was used (peak voltage 750 V; 270 μF capacitance). In the second phase, configuration IC-1–750 was enhanced by the addition of a second (parasternal) subcutaneous electrode (IC-2–750). Furthermore, the effects of a different 70 J shock waveform (1,000 V, 160 μF) were evaluated for configurations IM-1–750 and IC-2–750 (becoming IM-1–1000 and IC-2–1000).

Results

The proportion of patients satisfying a defibrillation safety margin test of 2 consecutive successes at ≤50 J was 74%, 11%, and 44%, respectively, for the IM-1–750, IC-1–750, and IC-2–750 configurations, and 93% and 86% for the IM-1–1000 and IC-2–1000 configurations.

Conclusions

Defibrillation efficacy comparable to that of a transvenous system was achieved with an anteroposterior subcutaneous ICD configuration, with 160 μF capacitance, 1,000 V, and 70 J output. An infraclavicular pocket location becomes feasible if a parasternal subcutaneous coil is added.


Introduction

  1. Top of page
  2. Subcutaneous Defibrillation
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Clinical Implications
  8. Study Limitations
  9. Acknowledgments
  10. References

A purely subcutaneous (SQ) implantable cardioverter defibrillator (ICD) may be an effective alternative to transvenous ICDs to deliver lifesaving therapy, but requires higher energy output to provide successful defibrillation.

Defibrillation efficacy for SQ defibrillation systems has been previously reported for both animals[1, 2] and humans[3-7] leading to the first permanent implants of a SQ-only system employing anterior and anterolateral electrodes.[8] For conventional ICD systems, posteriorly tunneled SQ electrodes have in the past been used to lower defibrillation thresholds by improving the shock vector.[9] Employing such a posterior SQ coil position, Lieberman et al. reported an anteroposterior SQ-only shock configuration using such an electrode with a can electrode in a low pectoral pocket.[10]

This study was intended to determine whether an anteroposterior shock pathway or alternative combinations using a posterior coil can provide adequate defibrillation to be considered for permanent implant of a system with twice the energy of conventional ICDs (i.e., 70 J). Defibrillation efficacy was examined for shock configurations using a posterior SQ coil, 2 anterior can positions (1 with and 1 without an auxiliary anterior SQ coil), and 2 different shock peak voltages.

Methods

  1. Top of page
  2. Subcutaneous Defibrillation
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Clinical Implications
  8. Study Limitations
  9. Acknowledgments
  10. References

Study Design Rationale

This study was divided into 2 phases designed to test the feasibility of a number of anteroposterior shock pathways and shock parameters in a series of exploratory investigations. This is a common approach in the early development of new defibrillation systems.[8] A summary on the different shock pathways and shock parameters used in both phases is provided in Figure 1.

image

Figure 1. Defibrillation configurations and shock parameters tested. Each parasternal SQ coil was electrically coupled to the respective can. E90 denotes the energy at which the likelihood of successful defibrillation is 90%.

Download figure to PowerPoint

Preceding the clinical investigation, the safety of delivering up to 1,000 V shocks in subcutaneous tissue was established in preclinical studies using a protocol similar to one published by Killingsworth et al.[11] Studies randomized animals to receive either five 35 J transvenous shocks or five 70 J SQ-only shocks into sinus rhythm. Measures of postshock tachyarrhythmic and bradyarrhythmic events, hemodynamics, ST-segment characteristics and blood chemistry analysis to assess cardiac damage were compared between groups. The results of these studies indicated that there was no additional damage caused by 70 J nontransvenous shocks compared to 35 J transvenous shocks (unpublished).

Phase 1 was based on the initial clinical experience of Lieberman et al.[10] and was intended to confirm the feasibility of an anteroposterior shock vector using the 750 V peak voltage of modern ICDs and 270 μF capacitance required to deliver 70 J. Configurations tested included a single posterior SQ lead and a can in either an inframammary pocket (configuration IM-1–750) or an infraclavicular pocket (IC-1–750). The choice of the pocket location was at the investigators’ discretion. This phase employed a group sequential design allowing up to 150 subjects to be tested with stopping for success or futility after each cohort of 25 subjects completed testing. This design provided >90% statistical power to identify a shock vector with at least 90% first shock efficacy at 70 J, while minimizing the number of subjects exposed to a poorly performing defibrillation pathway. Phase 1 was terminated after 75 subjects were tested with configurations IM-1–750 or IC-1–750, as the futility criterion was met in the prespecified pooled analysis.

In the light of the outcomes from the first phase, the design of phase 2 was adapted to utilize a more exploratory approach maximizing the likelihood of identifying at least 1 configuration with a true success rate of at least 90% while minimizing the number of subjects tested with poorly performing configurations. The feasibility of 3 additional combinations was tested in a nonrandomized fashion (see Fig. 1). First, the IC-1–750 shock vector was augmented by connecting a parasternal coil to the infraclavicular can. This and the inframammary configuration were then tested using an adapted shock waveform using a lower capacitance, shorter time-constant biphasic waveform previously shown to reduce defibrillation thresholds.[12-14] To maintain a 70 J maximum output with lower capacitance, the peak voltage had to be increased. Thus, the remaining 2 configurations used a 1,000 V/160 μF shock. The 3 new configurations were termed IC-2–750 (IC-1 shock vector with parasternal coil added), IM-1–1000 (inframammary configuration with a 1,000 V shock waveform), and IC-2–1000 (IC-2 vector with 1,000 V shock waveform).

Patient Selection

The study was approved by the institutional review board at each of the 15 participating centers and written informed consent was obtained from each patient.

Patients were eligible to participate in the study if they were scheduled for a left-sided ICD implantation for an approved indication. Exclusion criteria included age <18 years, pregnancy, medical conditions that might limit study participation, pacemaker dependence, surgical scarring to the left hemithorax, or preexisting defibrillation leads.

Implant Procedure

In each case, acute testing of a single SQ configuration was performed prior to implantation of the permanent transvenous ICD system to eliminate the possibility of distorting the defibrillation field with the defibrillator lead. An active can emulator (“can”) and 1 or 2 tunneled SQ coils were temporarily implanted under general anesthesia or conscious sedation.

The emulator can had a volume of 46 cm3 in anticipation of the volume of a device required to deliver 70 J shocks. It was implanted either in a low pocket (deep to pectoralis major, accessed via an inframammary incision), or in a high pocket (prepectoral, accessed via an infraclavicular incision). A dorsal lead (25 cm electrode, Medtronic 6996SQ) was tunneled posteriorly from an incision in the midaxillary line at the level of the xiphoid process; its tip was advanced as close as possible to the spine. In 2 groups of patients (IC-2–750 and IC-2–1000), a second SQ lead with a 15 cm coil was tunneled vertically along the left margin of the sternum until the tip was level to the xiphoid process. Finally, a diagnostic electrophysiology catheter was positioned in the right ventricle for induction of ventricular fibrillation and electrogram recording. Example radiographs of 2 configurations are shown in Figure 2.

image

Figure 2. Anteroposterior fluoroscopy images illustrating placement of can and SQ electrodes. Left: Can in inframammary pocket and (horizontally oriented) posterior SQ electrode (IM-1–750 and IM-1–1000 configuration). Right: Can in infraclavicular pocket with posterior and (vertically oriented) parasternal SQ electrodes (IC-2–750 and IC-2–1000 configuration). In both cases, a quadripolar EP catheter resides in the RV apex.

Download figure to PowerPoint

Defibrillation Testing Protocol

For defibrillation testing, the electrodes were connected to an external investigational instrument mimicking the envisioned ICD. VF was induced and maintained for a duration of 20 seconds prior to shock delivery based on a “worst-case” scenario in which arrhythmia detection is not straightforward, a maximum energy shock is needed, and battery performance is reduced due to aging. All shocks had biphasic truncated-exponential waveforms with a 50/50% tilt and used the can as the first phase cathode. Three defibrillation tests were performed using a binary search protocol (Fig.3) starting at 70 J in phase 1 or 50 J in Phase 2. The protocol in phase 1 was designed to characterize success of shocks from a 70 J device. In contrast, the protocol in phase 2 was designed to more directly measure the ability to pass a criterion of 2 consecutive (2/2) successful defibrillations at a 20 J safety margin. (In view of the higher energies needed for SQ defibrillation, the clinically accepted safety margin requirement of 10 J was doubled to 20 J below maximum output, i.e., 50 J test shocks.) In the event of a failed shock, for patient safety changes of shock polarity were not allowed. Failed test shocks were immediately followed by transthoracic rescue shocks. If more than 1 rescue shock was required, defibrillation testing was stopped to minimize patient risk. After testing, the anterior lead was removed if present and the prescribed transvenous ICD system was implanted.

image

Figure 3. Defibrillation testing protocols.

Download figure to PowerPoint

Statistical Methods

The safety margin success rate was defined as the percentage of patients with 2 consecutive successful shocks at programmed energies ≤50 J. For patients in phase 2, this was directly assessed by the defibrillation testing protocol. An exact 95% binomial confidence interval was computed. Patients in whom a safety margin could not be determined (e.g., a successful 50 J shock followed by a failed 35 J shock with no further inductions) were excluded from analysis per the investigational protocol, but were included in a sensitivity analysis using the Bayesian logistic energy response model described by Lieberman et al.[10] For patients in phase 1, the protocol did not permit direct calculation of the safety margin, but the latter was estimated based on the Bayesian model. For all groups, the model was also used to estimate the relationship between probability of defibrillation success and programmed energy.

Summary statistics and graphics were computed using R v2.4.1 (www.R-project.org) or SAS v9.1 (SAS Institute, Cary, NC, USA). Bayesian computations were performed using WinBUGs v1.4 (Imperial College, UK). The sponsor was responsible for data management and all authors had full access to the study data.

Results

  1. Top of page
  2. Subcutaneous Defibrillation
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Clinical Implications
  8. Study Limitations
  9. Acknowledgments
  10. References

Patients

A total of 141 patients were initially enrolled in the study. Data were excluded from 5 patients, 3 of whom exited prior to defibrillation testing, and 2 because of pneumothorax, which affects defibrillation performance.[15] Table 1 details the baseline characteristics of the 136 patients analyzed.

Table 1. Clinical Characteristics of Patients Analyzed
 AllIM-1–750IC-1–750IC-2–750IC-2–1000IM-1–1000
 (n = 136)(n = 58)(n = 15)(n = 25)(n = 24)(n = 14)
  1. Continuous variables are expressed as mean (standard deviation), categorical variables as percentage of each group.

  2. n = 135; n = 133.

  3. CABG = coronary artery bypass graft; CAD = coronary artery disease; CHF = congestive heart failure; LVEF = left ventricular ejection fraction; MI = myocardial infarction.

Age (years)62 (12)60 (11)60 (12)60 (11)66 (12)66 (11)
Male (%)909780967193
Height (cm)174 (8)174 (7)170 (7)178 (8)170 (7)172 (9)
Weight (kg)87 (20)87 (20)100 (27)88 (22)78 (13)84 (14)
Primary prevention (%)827887808886
LVEF mean (SD)33 (14)32 (13)32 (14)35 (16)37 (16)31 (9)
NYHA (%)      
Not reported477400
None437440
Class I18247161717
Class II434773402929
Class III31197365050
Class IV000000
CAD (%)575260645857
MI (%)474760484243
CHF (%)111427840
CABG (%)252227281743
Amiodarone (%)152402047
Beta-blocker (%)858393888386

Defibrillation Efficacy

A total of 400 shocks were included in the analysis. Numerical estimates of safety margin success, maximum output success, and energies with 50% and 90% probability of success (derived directly and from the Bayesian logistic energy response model) are given in Table 2, along with numbers of shocks and impedances for each configuration. Shock impedance was not associated with defibrillation performance. The two 1,000 V configurations (IM-1–1000, IC-2–1000) performed best with the highest success both at maximum output (97% predicted first shock success at 70 J) and with a safety margin test, whereas configuration IC-1–750 performed the worst. Figure 4 displays an estimated probability of success curve for each configuration, calculated from the Bayesian model.

Table 2. Subcutaneous Defibrillator Performance by Configuration
   Shock% PassingE50E90 
 SubjectsSQ ShocksImpedanceSafety MarginJoulesJoulesP70
ConfigurationTestedDeliveredΩ (± SD)(95% CI)(± SD)(± SD)(± SD)§
  1. Safety margin defined as 2 successful shocks with a programmed energy below 50 J. CI is the exact 95% binomial confidence interval for configurations IC-2–750, IM-1–1000, and IC-2–1000.

  2. E50, E90 are the energies at which the likelihood of successful defibrillation is 50% and 90%, respectively.

  3. §P70 is the average defibrillation efficacy at 70 J.

  4. Two subjects were excluded from analysis, as they were not retested at 50 J, after a successful first 50 J shock and a failed 35 J shock. A sensitivity analysis using the Bayesian model that included shocks from these 2 subjects yielded a similar estimated safety margin rate of 87% (71–100%).

IM-1–7505817266 ± 1174% (62–84%)31 (2)56 (4)0.91 (0.02)
IC-1–750154478 ± 1311% (0–27%)80 (10)138 (20)0.44 (0.07)
IC-2–750257554 ± 844% (24–65%)41 (3)75 (8)0.85 (0.04)
IC-2–1000246855 ± 886% (65–97%)25 (3)46 (5)0.97 (0.01)
IM-1–1000144162 ± 993% (66–100%)21 (4)36 (7)0.97 (0.02)
image

Figure 4. Likelihood of first shock success by configuration, calculated based on the Bayesian method described by Lieberman.[10]

Download figure to PowerPoint

Adverse Events and Complications

There were no deaths or adverse events directly related to defibrillation testing. Among 89 patients requiring transthoracic shocks, 12 patients required more than 1 rescue shock for an episode. One patient awoke from anesthesia with unilateral lower limb paresis. No abnormality was found on CT imaging of the spine but the paresis may have been caused by trauma during dorsal tunneling. The 2 pneumothoraces were caused by subclavian puncture during placement of the electrophysiology catheter used for VF induction.

Discussion

  1. Top of page
  2. Subcutaneous Defibrillation
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Clinical Implications
  8. Study Limitations
  9. Acknowledgments
  10. References

This investigation examined a variety of shock vectors and waveforms for a SQ defibrillator utilizing a posteriorly tunneled electrode. Two configurations were identified that demonstrated reliable defibrillation with a 1,000 V shock. In one, the can was implanted in an inframammary pocket using a dorsal SQ defibrillation lead (IM-1–1000), while in the other the can was in a conventional infraclavicular pocket but an additional parasternal SQ coil was added (IC-2–1000). Both configurations successfully defibrillated 93% and 86% of patients, respectively, in 2/2 tests at energies ≤50 J. The corresponding energies at which a first shock was predicted to have a 50% chance of success (E50) were 21 ± 4 J and 25 ± 3 J. Furthermore, with both of these configurations the predicted probability of a successful first shock at 70 J was greater than 95%—the level generally considered acceptable in clinical practice.

Two device locations were studied. The inframammary pocket, although less familiar to implanters, provides some advantages over the conventional infraclavicular pocket. First, the device and posterior lead can be implanted along a common transverse plane achieved using a single inframammary incision. Second, the IM-1 configuration may be more cosmetically acceptable and comfortable. Finally, the conventional infraclavicular ICD pocket is still available for future implants or upgrades requiring transvenous leads.

As the results achieved with conventional peak voltage (750 V) and high capacitance shocks observed in phase 1 were not satisfactory, we investigated the effect of increased peak voltage and reduced capacitance via configurations IM-1–1000 and IC-2–1000 in phase 2 of our study. For the given energy setting of 70 J, the corresponding peak voltage increased by ∼30% while capacitance and thus waveform time constant were reduced by ∼40%. These waveform changes improved defibrillation efficacy for both pocket locations. The estimated E50 was reduced from 31 to 21 J (32%) with the IM-1–1000 configuration, and from 41 to 25 J (39%) with the IC-2–1000 configuration. The corresponding safety margin success rates were increased from 74% to 93% and from 44% to 86%, respectively.

The final efficacies achieved with the higher voltage configurations in this study compare well to those associated with transvenous ICD studies testing 10 J safety margins such as the CREDIT registry, in which the initial system yielded success in 213/229 patients (93%).[16]

Our results are comparable to those in a study by Lieberman et al., which also used an anteroposterior SQ electrode configuration, where 80% of the patients were defibrillated with 35 J or less.[10] In a recent analysis by Jolley et al., using finite element modeling of 150 potential shock vectors, anteroposterior configurations of this type were predicted to be the most efficient.[17]

A completely SQ ICD system was reported recently by Bardy et al., using an 80 J defibrillator placed over the sixth rib in the anteroaxillary line and a parasternal SQ lead.[8] Defibrillation success in 53 candidates for permanent implantation was 94% for 2 consecutive tests with a 15 J safety margin (i.e., ≤65 J), after a mean VF duration of 14 seconds. The final success rate increased to 98% after reversing polarity in 2 patients. In this study, a more stringent level was set, with a 20 J safety margin for a maximum 70 J shock, VF duration of 20 seconds, and no reversal of polarity permitted.

Clinical Implications

  1. Top of page
  2. Subcutaneous Defibrillation
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Clinical Implications
  8. Study Limitations
  9. Acknowledgments
  10. References

This study has demonstrated the feasibility of an entirely SQ ICD using an anteroposterior shock pathway. Assuming that no other issues arise (including sensing, lead stability, and reliability) with completely SQ systems arise, such a system could provide an alternative to transvenous ICDs for delivering lifesaving therapy. Due to increased energy requirements, the device would be larger than a transvenous ICD system. The system would be reserved for patients without pacing indications and in situations where the need for antitachycardia pacing was not anticipated.

Our results also have implications for conventional ICDs in patients where placing coils in the heart is difficult or impossible. Although the experimental device used in the second phase of the study had a higher peak voltage (1,000 V) than conventional ICDs, most patients in these groups were defibrillated with ≤35 J (peak voltage ≤703 V). A conventional ICD delivers shocks with a maximum voltage of 750–800 V but has a smaller capacitance (105–125 μF vs 160 μF), so the waveform should be at least as effective as with the experimental system. Thus, similar to the results of Lieberman et al.,[10] a 35 J ICD designed for transvenous use but configured with SQ electrode(s) as in our study would be expected to defibrillate a high percentage of patients. Conventional defibrillators with nontransvenous shock pathways have been reported anecdotally,[3-6] but this study provides comparative data that could help determine the most effective electrode positions.

Study Limitations

  1. Top of page
  2. Subcutaneous Defibrillation
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Clinical Implications
  8. Study Limitations
  9. Acknowledgments
  10. References

The purpose of the study was to identify subcutaneous configurations with defibrillation performance comparable to those of transvenous ICDs. It was not designed as a randomized trial comparing configurations, nor was it powered as such. There were differences in patient characteristics between configurations (Table 1) suggesting body habitus may have been an important component in the selection of a configuration to test. Thus, a confounding effect of these baseline characteristics cannot be excluded, and comparisons between configurations must be drawn cautiously.

The study used 2 different defibrillation-testing protocols so a statistical model was used to estimate the safety margin rates for configurations used in phase 1. However, in phase 2 (which showed the most effective configurations), the protocol allowed direct estimation of the safety margin rate.

As with the overwhelming majority of ICD studies, there is no guarantee that the performance for induced VF episodes is consistent with that for spontaneous episodes, and data from chronic implants will be required to evaluate ambulatory performance. For example, changes in posture may alter the shock electrode position relative to the heart and may impact defibrillation performance as seen in transvenous configurations.[18] Finally, the study did not examine other important aspects of a nontransvenous system including the sensing and detection performance, long-term lead stability, or the implications of the absence of bradycardia pacing.

Acknowledgments

  1. Top of page
  2. Subcutaneous Defibrillation
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Clinical Implications
  8. Study Limitations
  9. Acknowledgments
  10. References

 We thank the SOLO investigators and their institutions for their participation (subjects recruited in parentheses): Steven Compton, Alaska Heart, Anchorage, Alaska (1); Derek Connelly, Glasgow Royal Infirmary, Glasgow, United Kingdom (1); Charles Haffajee, Caritas St. Elizabeth's Medical Center, Boston, Massachusetts; (5) Volker Kühlkamp, Herz-Neuro-Zentrum Bodensee, Konstanz, Germany (14); Jürgen Kuschyk, University Hospital, Mannheim, Germany (46); Goran Milasinovic, Clinical Center of Serbia, Belgrade, Serbia (21); Franck Molin, Hôpital Laval, Sainte-Foy, Quebec (9); Francis Murgatroyd and Nicholas Gall, King's College Hospital, London, United Kingdom (12); Brigitte Osswald, University of Essen, Essen, Germany; (5, while at University Hospital, Mannheim) Paul Roberts and John Morgan, Southampton General Hospital, Southampton, United Kingdom (11); Lisa Schiller, St. Louis University, St. Louis, Missouri (1); Stephen Shorofsky, University of Maryland Medical Center, Baltimore, Maryland (3); Jay Wright, Cardiothoracic Center Liverpool, Liverpool, United Kingdom (1); Raymond Yee, London Health Sciences Center, London, Ontario (4); Markus Zabel and Dieter Zenker, University of Göttingen, Göttingen, Germany (7). We also thank David Kress, Randy Lieberman, and Robert Wiechmann for their help in developing the implant procedure and with subsequent investigator training. Finally, we thank the Medtronic research, field, and clinical support staff for their support and management of the trial.

References

  1. Top of page
  2. Subcutaneous Defibrillation
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Clinical Implications
  8. Study Limitations
  9. Acknowledgments
  10. References
  • 1
    Schuder JC, Stoeckle H, Gold JH, West JA, Keskar PY: Experimental ventricular defibrillation with an automatic and completely implanted system. Trans Am Soc Artif Intern Organs 1970;16:207-212.
  • 2
    Cappato R, Castelvecchio S, Erlinger P, Sanghera R, Scheck D, Ostroff A, Rissmann W, Gropper C, Bardy GH: Feasibility of defibrillation and automatic arrhythmia detection using an exclusively subcutaneous defibrillator system in canines. J Cardiovasc Electrophysiol 2013;24:77-82.
  • 3
    Madan N, Gaynor JW, Tanel R, Cohen M, Nicholson S, Vetter V, Rhodes L: Single-finger subcutaneous defibrillation lead and “active can”: A novel minimally invasive defibrillation configuration for implantable cardioverter-defibrillator implantation in a young child. J Thorac Cardiovasc Surg 2003;126:1657-1659.
  • 4
    Berul CI, Triedman JK, Forbess J, Bevilacqua LM, Alexander ME, Dahlby D, Gilkerson JO, Walsh EP: Minimally invasive cardioverter defibrillator implantation for children: An animal model and pediatric case report. Pacing Clin Electrophysiol 2001;24:1789-1794.
  • 5
    Osswald BR, Tochtermann U, Tanzeem A, Hagl S: Minimal touch surgical implantable cardioverter defibrillator implantations. Interact Cardiovasc Thorac Surg 2006;5:343-344.
  • 6
    Stephenson EA, Batra AS, Knilans TK, Gow RM, Gradaus R, Balaji S, Dubin AM, Rhee EK, Ro PS, Thogersen AM, Cecchin F, Triedman JK, Walsh EP, Berul CI: A multicenter experience with novel implantable cardioverter defibrillator configurations in the pediatric and congenital heart disease population. J Cardiovasc Electrophysiol 2006;17:41-46.
  • 7
    Burke MC, Coman JA, Cates AW, Lindstrom CC, Sandler DA, Kim SS, Knight BP: Defibrillation energy requirements using a left anterior chest cutaneous to subcutaneous shocking vector: Implications for a total subcutaneous implantable defibrillator. Heart Rhythm 2005;2:1332-1338.
  • 8
    Bardy GH, Smith WM, Hood MA, Crozier IG, Melton IC, Jordaens L, Theuns D, Park RE, Wright DJ, Connelly DT, Fynn SP, Murgatroyd FD, Sperzel J, Neuzner J, Spitzer SG, Ardashev AV, Oduro A, Boersma L, Maass AH, Van Gelder IC, Wilde AA, van Dessel PF, Knops RE, Barr CS, Lupo P, Cappato R, Grace AA: An entirely subcutaneous implantable cardioverter-defibrillator. N Engl J Med 2010;363:36-44.
  • 9
    Kuhlkamp V, Dornberger V, Khalighi K, Mewis C, Suchalla R, Ziemer G, Seipel L: Effect of a single element subcutaneous array electrode added to a transvenous electrode configuration on the defibrillation field and the defibrillation threshold. Pacing Clin Electrophysiol 1998;21:2596-2605.
  • 10
    Lieberman R, Havel WJ, Rashba E, DeGroot PJ, Stromberg K, Shorofsky SR: Acute defibrillation performance of a novel, non-transvenous shock pathway in adult ICD indicated patients. Heart Rhythm 2008;5:28-34.
  • 11
    Killingsworth CR, Melnick SB, Litovsky SH, Ideker RE, Walcott GP: Evaluation of acute cardiac and chest wall damage after shocks with a subcutaneous implantable cardioverter defibrillator in swine. Pacing Clin Electrophysiol 2013 May 28. doi: 10.1111/pace.12173. [Epub ahead of print] PMID:23713608
  • 12
    Kroll MW: A minimal model of the monophasic defibrillation pulse. Pacing Clin Electrophysiol 1993;16:769-777.
  • 13
    Swerdlow CD, Brewer JE, Kass RM, Kroll MW: Application of models of defibrillation to human defibrillation data: Implications for optimizing implantable defibrillator capacitance. Circulation 1997;96:2813-2822.
  • 14
    Irnich W: Optimal truncation of defibrillation pulses. Pacing Clin Electrophysiol 1995;18:673-688.
  • 15
    Luria D, Stanton MS, Eldar M, Glikson M: Pneumothorax: An unusual cause of ICD defibrillation failure. Pacing Clin Electrophysiol 1998;21:474-475.
  • 16
    Healy JS, Dorian P, Mitchell LB, Talajic M, Philippon F, Simpson C, Yee R, Morillo CA, Lamy A, Basta M, Birnie DH, Wang X, Nair GM, Crystal E, Kerr CR, Connolly SJ; Canadian Registry of ICD Implant Testing Procedures (CREDIT): Current practice, risks, and costs of intraoperative defibrillation testing. J Cardiovasc Electrophysiol 2010;21:177-182.
  • 17
    Jolley M, Stinstra J, Tate J, Pieper S, Macleod R, Chu L, Wang P, Triedman JK: Finite element modeling of subcutaneous implantable defibrillator electrodes in an adult torso. Heart Rhythm 2010;7:692-698.
  • 18
    Schauerte P, Diem B, Ziegert K, Franke A, Hanrath P, Stellbrink C: Influence of body position on defibrillation thresholds of nonthoracotomy implantable defibrillators: A prospective randomized evaluation. J Cardiovasc Electrophysiol 1998;9:696-702.