Author + information
- Received March 3, 2016
- Revision received June 27, 2016
- Accepted June 29, 2016
- Published online September 7, 2016.
- Prabhat Kumar, MD,
- Matthew Baker, MD and
- Anil K. Gehi, MD∗ ()
- Department of Medicine, Division of Cardiology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
- ↵∗Reprint requests and correspondence:
Dr. Anil K. Gehi, Department of Medicine, Heart and Vascular Center, The University of North Carolina at Chapel Hill, 160 Dental Circle, CB 7075, Chapel Hill, North Carolina 27599.
Objectives The purpose of this investigation was to conduct a meta-analysis of studies comparing defibrillation threshold (DFT) and outcomes with single-coil and dual-coil implantable cardioverter-defibrillator (ICD) systems.
Background Use of dual-coil rather than single-coil defibrillator leads may lower the DFT with a transvenous ICD system; however, dual-coil ICDs may have higher lead-related complications.
Methods Sixteen studies, each with more than 10 human subjects, that compared single-coil and dual-coil ICD systems were included for the final analysis after a comprehensive publication search using predefined search terms and additional search from cross-references. A test of heterogeneity, pooling, and meta-analysis of the data from the studies were performed using R statistical software. A random effects model was used for meta-analysis.
Results Data pooled from 14 studies analyzed for difference in DFT showed an estimated difference in mean DFTs between single-coil and dual-coil ICDs of 0.81 J (95% confidence interval [CI]: 0.31 to 1.30 J), thus favoring dual-coil ICDs. However, pooled data from 5 studies revealed no difference in first-shock efficacy for dual-coil ICDs compared with single-coil ICDs (estimated overall odds ratio: 0.94; 95% CI: 0.49 to 1.78; p = 0.85). The all-cause mortality rate analyzed from 4 studies was lower in patients with single-coil ICDs (estimated hazard ratio: 0.91; 95% CI: 0.83 to 0.99).
Conclusions There was a marginal difference in the defibrillation threshold of transvenous ICDs between single-coil and dual-coil lead systems. However, first-shock efficacy was no different between the 2 groups, and patients with single-coil ICDs had favorable all-cause mortality rates on the basis of data from nonrandomized studies. Potential risks and benefits of single-coil and dual-coil ICD leads should be carefully weighed.
- all-cause mortality
- defibrillation threshold
- dual-coil implantable cardioverter-defibrillator
- implantable cardioverter-defibrillator
- single-coil implantable cardioverter-defibrillator
The implantable cardioverter-defibrillator (ICD) has become mainstay therapy for prevention of sudden cardiac death in high-risk patients (1–3). The dual-coil ICD lead, with one shocking coil in the right ventricle and the other in the superior vena cava (SVC), has become a standard configuration. Along with an active can, using pulse generator as one of the electrodes, a dual-coil ICD lead is used to provide an effective shocking vector. However, the additional shocking coil in the SVC adds to the amount of metal in the vascular system. Adhesions developing around the SVC coil add to difficulty in extraction of the ICD lead if lead removal is required (4,5). This concern has led to rethinking of the benefit of dual-coil ICD lead systems in improving defibrillation threshold (DFT) and preventing sudden death. In this meta-analysis we present comparative data from studies assessing the DFTs and long-term outcomes with single-coil and dual-coil ICD systems (6–21).
A publication search was performed on the PubMed (U.S. National Library of Medicine) database to identify reports published from 1990 to 2015 to compare single-coil and dual-coil ICD systems. The following Medical Subject Heading (MeSH) terms were used: coil AND (implantable defibrillator [MeSH Terms]) AND (cohort studies [MeSH Terms] OR clinical trial [publication type] OR controlled clinical trial [publication type]). Search terms single-coil AND dual-coil AND (ICD OR defibrillator) were used for additional search. The search was restricted to English language publications and human subjects older than 18 years of age. Additional reports were identified and included after a second search of papers published by investigators identified in the original search and a review of the bibliographies of the originally identified papers.
Study inclusion criteria were as follows: 1) study of single-coil versus dual-coil ICDs for primary prevention in congestive heart failure; 2) primary data; 3) human subjects; 4) more than 10 patients; and 5) published in the English language. Studies were excluded if: 1) the devices in the study were not transvenous; or 2) the devices had an alternate (nonpectoral) location.
Two of the authors (P.K. and M.B.) independently reviewed the titles and the abstracts of all the reports and excluded the reports that either did not meet inclusion criteria or had any exclusion criteria. The full paper was reviewed if there was lack of clarity with abstract review. Along with the third author (A.G.), a consensus was reached to finalize the study selection. Figure 1 shows study selection process in accordance with QUOROM (Quality of Reporting of Meta-analyses) guidelines (22).
The selected reports were reviewed in detail by 2 of the authors (P.K. and M.B.) independently; these authors were blinded to the author, journal, and institution. Data were abstracted in a standardized format for each paper. Any discrepancies were resolved by consensus with the third author (A.G.). Abstracted data from each report included the number of patients, demographic data (age and sex in each group), indications for ICD (primary vs. secondary prevention, ischemic vs. nonischemic cardiomyopathy, New York Heart Association functional class), device implantation and programming characteristics (left-sided implant, active can implant, shocking vector), use of antiarrhythmic therapy, and performance parameters for ICD (DFT, DFT threshold testing protocol, death, sudden cardiac death).
Meta-analysis was performed to estimate the mean difference in DFT and the hazard ratio (HR) for all-cause mortality data between single-coil and dual-coil ICDs. Mean and standard deviation of DFT in individual studies were recorded either as paired values for the 2 groups when the same patient was tested for DFT with single-coil and dual-coil configuration or as unpaired DFTs in different sets of patients. One of the studies had DFTs recorded for high and low SVC coil positions. Two separate analyses were done to account for the 2 different SVC coil positions. Two of the studies included in the meta-analysis of first-shock efficacy reported odds ratios (6,19); however, for the other 3 studies, odds ratios and standard errors were calculated from the numbers and proportions reported in the published papers (18,20,21). Summary estimates of mean difference in DFTs, HRs, and odds ratios were calculated using random-effects modeling of Der Simonian and Laird meta-analytic methods (23). Heterogeneity of the data across studies was assessed using Cochrane Q statistic and the I2 index. Publication bias was assessed using a funnel plot. Statistical tests were performed using open-source software R version 3.2.3 (The R Foundation for Statistical Computing).
Study identification and characteristics
The publication search resulted in 153 papers, which were reviewed for inclusion in the study as specified. After review of the title and the abstract of these papers, 127 papers were excluded. One paper was identified after reviewing the cross-references. After detailed review, a final selection of 14 papers was used for the meta-analysis to compare DFTs. Four studies were selected for analysis of all-cause mortality data; 3 of these studies were also included in the analysis of DFT, whereas 1 study did not have any data on DFT. Five studies were included in the meta-analysis of first-shock efficacy. Two of the studies were randomized comparisons of single-coil and dual-coil ICD systems, neither of which reported outcome data (14,16). Two large studies presented data from the SCD-HeFT (Sudden Cardiac Death in Heart Failure Trial) and MADIT-CRT (Multicenter Automatic Defibrillator Implantation With Cardiac Resynchronization Therapy) trials (6,18), and another large study published data from a Danish registry (19). Data from the ALTITUDE study were presented in another large study comparing long-term outcomes (20). Seven of the studies had paired data for DFT, with each patient tested for DFT in single-coil and dual-coil ICD configurations at the time of device implantation (7–11,15,17).
Characteristics of the study patients
In the 14 selected studies for analysis of DFT, the total number of subjects in the single-coil arm was 2,647, and in the dual-coil arm it was 5,407. Of these patients, 279 were from the studies with paired data on the same patients. Mean age of the patients ranged from 58 to 67 years. Most patients were male (range 62% to 92%). Three of these studies also reported mortality data. The study by Hsu et al. (20), containing ALTITUDE study data, was the fourth study included in the meta-analysis of all-cause mortality. These 4 studies together had 21,439 and 115,311 patients in the single-coil and dual-coil arms, respectively, for all-cause mortality analysis. Table 1 summarizes patient-related characteristics in the selected studies.
Most of the studies with unpaired data had a majority of patients receiving dual-coil ICDs, with the exception of 2 randomized studies (14,16) and 1 small study with a total of only 16 patients (12). In the 2 large, real-world studies and registry data by Hsu et al. (20) and Larsen et al. (19), respectively dual-coil ICDs were used in the majority (85.2% and 61.1%, respectively) of the patients.
Seven of the studies reported whether the indication for ICD implantation was primary or secondary prevention (6–9,13,15,19). The largest study, by Larsen et al. (19), provided published results from a Danish nationwide registry, which had a mix of patients with primary (51%) and secondary prevention indications. Another large study, by Aoukar et al. (6), was an analysis of a primary prevention ICD trial. In 4 other studies, all patients had secondary prevention indications (7,9,13,15). In contrast, a study by Gold et al. (8) had a mixed population of patients, with 53% of the patients receiving an ICD for primary prevention. The majority of the patient population studied had ischemic heart disease; 51% to 91% of the patients in the 12 studies were reported to have cardiomyopathy.
Site of device implantation
One of the studies, by Varma et al. (17), evaluated patients undergoing right pectoral device implantation only. Of the remaining 13 studies, 4 had no clear documentation whether right or left pectoral device implantation was used (10–13). In the study by Manolis et al. (13) with a total of 94 patients, 12 patients had left paraumbilical implants. The other 82 patients in the study by Manolis et al. (13) had pectoral implants of unspecified laterality. In the Danish ICD registry report by Larsen et al. (19), 95% of the patients had left-sided implants. The remaining 7 studies reported patients with left pectoral device implantation (6–9,15,16,18).
DFT was estimated by binary search protocol in 6 studies and by step-down protocol in 3 studies. The study by Aoukar et al. (6) reported defibrillation testing by a protocol using 2 values of 20 J and 10 J. The remaining 4 studies did not report the protocol for DFT testing (12,13,18,19). Active can configuration was used in all the studies in all patients except for the study by Manolis et al. (13). In that study, only 79% and 84% of the patients in the single-coil and dual-coil ICD trial arms, respectively, had active can configuration.
Estimated mean difference in the DFT between the 2 groups of patients was 0.81 J (95% confidence interval [CI]: 0.31 to 1.30 J; p = 0.0014), thus favoring dual-coil ICDs. Figure 2 shows the Forest plot from meta-analysis to estimate mean difference in DFT. One of the studies by Gold et al. (8) compared the DFTs with single-coil and dual-coil configurations in 2 groups of patients with high and low SVC coil positions. Higher SVC coil position showed significant reduction in DFT compared with single-coil configuration. Analysis performed including and excluding the patients with a low SVC coil from this study did not change the overall result (Online Figures 1 to 3).
Subgroup analysis of studies with paired data alone showed an estimated mean difference in DFT of 1.14 J (95% CI: 0.24 to 2.05), whereas the estimate in the studies with unpaired data was 0.64 J (95% CI: 0.00 to 1.27) (Figure 2). Subgroup analysis for randomized trials alone showed an estimated mean difference in DFT of 0.28 J (95% CI: −0.99 to 1.55).
Mortality data were presented in 4 of the studies, 3 of which were included in the analysis for DFT (6,18,19), and another study by Hsu et al. (20) with data from the ALTITUDE study, which has the largest data on all-cause mortality comparing single-coil and dual-coil ICDs. Meta-analysis of all-cause mortality data showed lower mortality rates with single-coil ICD leads compared with dual-coil ICD (estimated overall HR: 0.91; 95% CI: 0.83 to 0.99; p < 0.031) (Figure 3). Mean follow-up duration in the study by Kutyifa et al. was 40 months, in the study by Aoukar et al. (6) it was 45.5 months, and in the studies of Hsu et al. (20) and Larsen et al. (19) it was 6 years.
In the studies by Kutyifa et al. (18) and Larsen et al. (19), patients receiving single-coil ICDs had a higher left ventricular ejection fraction, although marginally so. Both these studies reported HRs, which were adjusted for baseline covariates. The study by Hsu et al. (20) reported an adjusted HR; however, adjustment was made only for age, sex, device type, and year of device implantation. The study by Aoukar et al. (6) did not show any significant difference in prognostic variables including left ventricular ejection fraction between single-coil and dual-coil ICD groups, and the reported HR was adjusted for the prognostic variables.
Five studies presented first-shock efficacy data, 4 of which also presented mortality data. The study of Ellis et al. (21) presented first-shock efficacy data, but it did not include data on DFT or mortality rates. Combined data demonstrated no difference in first-shock efficacy of dual-coil ICD compared with the single-coil ICD (estimated overall odds ratio: 0.94; 95% CI: 0.49 to 1.78; p = 0.85) (Figure 4).
Heterogeneity and publication bias
There was no significant heterogeneity in the data from different studies (tau = 0.4925; I2 < 39.43%; H2 = 1.65; Q = 21.46; p = 0.064) for meta-analysis of DFT. A funnel plot asymmetry analysis was used to assess publication bias by using the regression test described by Egger et al. (24). There was no statistical evidence of publication bias (p = 0.244).
No significant heterogeneity among the studies was seen for meta-analysis of all-cause mortality data (I2 = 0%; tau-squared = 0; p = 0.9414) or first-shock efficacy data (I2 = 0%; tau-squared = 0; p = 0.9041). There was no statistically significant publication bias in the data for mortality (p = 0.997).
Analysis of heterogeneity of data among studies included in shock efficacy also did not reveal any significant heterogeneity (tau = 0; I2 = 0%; H = 0; Q = 1.04; p = 0.9041). There was no significant publication bias in the data for first-shock efficacy (p = 0.43).
In this meta-analysis we compared the DFTs, first-shock efficacy, and all-cause mortality between single-coil and dual-coil ICDs. Fourteen studies were included for the analysis of DFTs, 4 studies were included in the mortality analysis, and 5 studies were included in the first-shock efficacy analysis. The estimated mean difference in the DFT was found to be 0.81 J (95% CI: 0.31 to 1.30 J) for all the studies. Analysis of studies with paired comparison alone showed the estimated mean difference of 1.14 J (95% CI: 0.24 to 2.05 J). This difference is unlikely to be of any significant clinical importance. The all-cause mortality rate was lower in patients receiving a single-coil ICD as compared with a dual-coil ICD, with estimated overall HR of 0.91 (95% CI: 0.83 to 0.99), and first-shock efficacy was no different between the 2 groups (estimated overall odds ratio for dual-coil compared with single-coil ICD: 0.94; 95% CI: 0.49 to 1.78; p = 0.85). The data did not have significant heterogeneity, and there was no significant publication bias.
The benefit of dual-coil leads in current ICD systems has been questioned, yet nearly one-half of the patients receiving an ICD currently in the United States receive a dual-coil lead (20). Although the addition of an SVC coil decreases impedance, alters the vector of energy delivery, and thus may improve the DFT, our study demonstrates that the introduction of an active can defibrillator system to single-coil defibrillator systems has efficacy that is nearly equal to that of dual-coil systems in DFT. Moreover, with the high safety margin of current defibrillators as a result of high-energy output and optimization of defibrillation waveforms, defibrillation efficacy may not be affected by small differences in DFTs. Studies have demonstrated that DFT testing may not be necessary and may, in fact, be harmful (25–27), even in patients with single-coil leads. In light of these findings, there is a current trend toward eliminating the need for DFT testing on a routine basis at the time of device implantation (25–27). In the absence of any benefit in terms of defibrillation efficacy, potential problems with durability and risks associated with extraction of dual-coil ICD leads may lead to an increase in adverse events including death, as suggested by this meta-analysis. Indeed, the trend toward single-coil leads evident from large, real-world device implantation data in the ALTITUDE study (1.9% single-coil in 2004 to 55.2% single-coil in 2014) suggests that clinicians are gradually changing practice patterns accordingly (20).
An active can configuration has been thought to be instrumental in diminishing the importance of an SVC coil. Arguably, in right pectoral devices with a less favorable right ventricular coil-to-can configuration, an SVC coil may be expected to have more effect on DFT. There has been concern about elevated DFT with right pectoral implants (28,29). One of the studies included in this meta-analysis, however, did not show any significant difference in DFT between single-coil and dual-coil configurations, even with a right pectoral device implantation site.
This study showed that all-cause mortality rates were lower in the group with single-coil ICD systems than with dual-coil ICDs. Although the all-cause mortality analysis in these studies was performed with adjustment for possible patient-related covariates, not all the important covariates were available for analysis. Because these studies were not randomized between single-coil ICDs and dual-coil ICDs, it is conceivable that patients with dual-coil ICDs were sicker and hence more likely to have worse outcomes. It is also possible, however, that the increased mortality rate reflected lead-related complications. This possibility is supported by the finding that differences in survival occurred several years after the initial implantation procedure (19,20). It is possible that further analysis from ALTITUDE study data by Hsu et al. (20) comparing patients who received single-coil ICD early versus late in the study period may help clarify this issue. However, such analysis was not presented in their published report.
First, only 2 of the studies among those with unpaired samples were randomized studies. It is possible that patients receiving dual-coil ICDs were sicker and hence had higher DFTs, thus masking the differences in the DFTs. However, even the studies with paired data showed similar DFTs in the 2 groups. Second, there was a heterogeneous population of patients included in these studies. Subgroup analysis and metaregression with factors such as left ventricular dimensions, patients’ body sizes, and use of various drugs were not possible because of the unavailability of that information in these studies. Moreover, there was no uniformity in the methodologies for DFT testing across these studies. Third, data on single-coil versus dual-coil implants with right-sided implants are limited, and hence no conclusion can be drawn for right-sided implants.
Single-coil defibrillator systems have a higher DFT compared with dual-coil ICD systems but no difference in first-shock efficacy. The small difference in DFT in the current era of high-energy devices with optimized defibrillation waveforms is unlikely to be of any clinical significance. Moreover, data on mortality in the present analysis, although largely consisting of nonrandomized data, favor single-coil ICDs. In view of inherent risks with dual-coil ICD leads in terms of structural integrity and lead extraction, any potential advantage of a dual-coil device should be carefully weighed against its inherent problems. Further investigation to determine which patient subgroups are likely to have high DFTs and may warrant dual-coil ICD systems is required.
COMPETENCY IN MEDICAL KNOWLEDGE: This meta-analysis suggests that although DFTs are marginally lower with dual-coil ICDs, there is no significant difference in first-shock efficacy, and all-cause mortality rates may be higher in the dual-coil ICD group.
TRANSLATIONAL OUTLOOK: It appears from this study that dual-coil ICDs may be associated with worsened long-term outcomes, although the major studies analyzed in this meta-analysis were not randomized studies, and the unfavorable outcomes with dual-coil ICDs may be related to differences in baseline comorbidities. Randomized studies will clarify this question. Although in this analysis, DFTs were only marginally different in the 2 groups, comparative study in subgroup of patients likely to have higher DFTs would be helpful in assessing the benefit for dual-coil ICD in selected patients.
For supplemental figures, please see the online version of this article.
Dr. Gehi has received speaker honoraria from Medtronic, Biotronik, Boston Scientific, St. Jude Medical, and Zoll. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- defibrillation threshold
- implantable cardioverter-defibrillator
- superior vena cava
- Received March 3, 2016.
- Revision received June 27, 2016.
- Accepted June 29, 2016.
- American College of Cardiology Foundation
- Epstein A.E.,
- DiMarco J.P.,
- Ellenbogen K.A.,
- et al.
- Connolly S.J.,
- Hallstrom A.P.,
- Cappato R.,
- et al.
- Nanthakumar K.,
- Epstein A.E.,
- Kay G.N.,
- Plumb V.J.,
- Lee D.S.
- Epstein L.M.,
- Love C.J.,
- Wilkoff B.L.,
- et al.
- Bardy G.H.,
- Dolack G.L.,
- Kudenchuk P.J.,
- Poole J.E.,
- Mehra R.,
- Johnson G.
- Gold M.R.,
- Olsovsky M.R.,
- Pelini M.A.,
- Peters R.W.,
- Shorofsky S.R.
- Schulte B.,
- Sperzel J.,
- Carlsson J.,
- et al.
- ↵Ellis CR, Hurt JT. Single-coil versus dual-coil ICD lead shock efficacy in a large ICD registry Journal of Innovations in Cardiac Rhythm Management 2012;3:953–6. Also available at: http://www.innovationsincrm.com/cardiac-rhythm-management/2012/october/348-single-coil-versus-dual-coil-icd-lead-shock-efficacy#sthash.M4kVm5ag.dpuf. Accessed July 14, 2016.
- Egger M.,
- Davey Smith G.,
- Schneider M.,
- Minder C.
- Blatt J.A.,
- Poole J.E.,
- Johnson G.W.,
- et al.
- Phan K.,
- Ha H.,
- Kabunga P.,
- Kilborn M.J.,
- Toal E.,
- Sy R.W.