Author + information
- Received May 2, 2017
- Revision received July 3, 2018
- Accepted July 12, 2018
- Published online November 19, 2018.
- Valentina Kutyifa, MD, PhDa,b,∗∗ (, )
- Annamaria Kosztin, MD, PhDb,d,∗,
- Helmut U. Klein, MDa,
- Yitschak Biton, MDc,
- Vivien Klaudia Nagy, MDb,
- Scott D. Solomon, MDd,
- Scott McNitt, MSa,
- Wojciech Zareba, MD, PhDa,
- Ilan Goldenberg, MDa,
- Attila Roka, MDc,
- Arthur J. Moss, MDa,
- Bela Merkely, MD, PhDb,∗ and
- Jagmeet P. Singh, MD, DPhilc,∗
- aUniversity of Rochester Medical Center, Rochester, New York
- bSemmelweis University, Heart Center, Budapest, Hungary
- cMassachusetts General Hospital, Harvard Medical School, Boston, Massachusetts
- dBrigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts
- ↵∗Address for correspondence:
Dr. Valentina Kutyifa, Heart Research Follow-up Program, Cardiology Division, University of Rochester Medical Center, 265 Crittenden Boulevard, Box 653, Rochester, New York 14642.
Objectives The authors aimed to evaluate the association of left ventricular (LV) lead location and long-term outcomes in MADIT-CRT (Multicenter Automatic Defibrillator Implantation With Cardiac Resynchronization Therapy).
Background There is limited data on the association of lead location with long-term clinical outcomes in patients with cardiac resynchronization therapy with defibrillator (CRT-D).
Methods The LV lead location was classified in 797 patients with CRT-D, in 569 patients with left bundle branch block (LBBB), in 228 patients with non-LBBB, and in 505 patients with an implantable cardioverter-defibrillator (ICD) only. Leads were classified into apical (n = 83) and non-apical (n = 486); with the non-apical LV leads further categorized into anterior (n = 99) and posterior/lateral (n = 387) within LBBB. All-cause mortality and heart failure (HF) events were assessed using Kaplan-Meier and Cox analyses.
Results In CRT-D patients with LBBB and posterior/lateral LV lead location, there was an association with a significant reduction in long-term all-cause mortality (hazard ratio [HR]: 0.54, 95% confidence interval [CI]: 0.37 to 0.79; p = 0.001), and HF events (HR: 0.44, 95% CI: 0.33 to 0.60; p < 0.001) compared to an ICD only, accompanied with better LV reverse remodeling. CRT-D patients with LBBB and an anterior LV lead location were shown to be associated with a significant reduction in HF events compared to an ICD only (anterior HR: 0.50, 95% CI: 0.30 to 0.82; p = 0.006); however, no association with mortality reduction was observed from CRT-D versus an ICD only. CRT-D was not associated with improved outcomes in non-LBBB patients, regardless of LV lead location.
Conclusions In mild HF patients with LBBB and an implanted CRT-D, lateral/posterior, and anterior LV lead locations are similarly associated with reduction in the risk of HF or death events compared to ICD alone. Mortality benefit derived from CRT-D is associated only with patients with lateral/posterior LV lead location. An apical LV lead location should be avoided due to the early risk of death whenever possible. (Multicenter Automatic Defibrillator Implantation With Cardiac Resynchronization Therapy [MADIT-CRT], NCT00180271; Multicenter Automatic Defibrillator Implantation Trial With Cardiac Resynchronization Therapy Post Approval Registry [MADIT-CRT-PAR], NCT01294449; and MADIT-CRT Long-Term International Follow-Up Registry – Europe, NCT02060110)
Cardiac resynchronization therapy (CRT) improves heart failure (HF) symptoms, exercise capacity, and reduces the risk of HF events and all-cause mortality in patients with mild to severe HF and a prolonged QRS (1–3).
Our prior data from the MADIT-CRT (Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy) suggest that left ventricular (LV) apical lead position is associated with adverse clinical outcomes (4) and an anterior LV lead location is associated with higher risk of ventricular tachyarrhythmias (5) during short-term follow-up in CRT with defibrillator (CRT-D) patients. However, long-term follow-up data on the associations of LV lead locations with clinical outcomes are lacking.
We have also shown that in MADIT-CRT–only patients with left bundle branch block (LBBB) electrocardiogram (ECG) pattern derive benefit from CRT-D (6). However, it has not been studied whether there are differences in long-term clinical outcomes by LV lead location and by the presence of baseline LBBB ECG pattern. MADIT-CRT is 1 of the largest randomized controlled studies on cardiac resynchronization therapy with adjudicated LV lead location data available; therefore, it provides a unique opportunity to conduct this analysis.
In the current sub-study of the long-term follow-up of MADIT-CRT, we aimed: 1) to investigate long-term clinical outcomes of all-cause mortality and HF in CRT-D patients by anterior, posterior, or lateral, and by apical or non-apical LV lead location, performing separate analyses in LBBB and non-LBBB patients; and 2) to assess LVC reverse remodeling by LV lead location in patients with LBBB ECG pattern.
The design, protocol, and results of the MADIT-CRT study have been published previously (1,7). Briefly, 1,820 patients with ischemic cardiomyopathy (New York Heart Association [NYHA] functional class I or II) or nonischemic cardiomyopathy (NHYA functional class II only), LV ejection fraction (LVEF) of <30%, and a prolonged QRS duration >130 ms were randomized to receive CRT-D or ICD therapy in a 3:2 ratio. All eligible patients met the guideline criteria for implantable cardioverter-defibrillator (ICD) (8). A total of 110 hospital centers from North America and Europe participated in this international multicenter trial. The study was in compliance with the Declaration of Helsinki and all enrolling sites had the protocol approved by local institutional review boards. All patients provided informed consent before enrollment.
In this study, patients with implanted CRT-D and LV lead location information were included. A total of 1,089 patients were randomized to CRT-D, 967 of them had a successful CRT-D implantation. An additional 90 patients were excluded because of epicardial LV lead placement or LV lead revision during follow-up, 78 patients had incomplete data to determine the LV lead location, and 2 patients were excluded due to crossover to a CRT-D following randomization to an ICD. Therefore, a total of 797 patients were evaluated for final LV lead location; 569 of them had LBBB, and 228 had non-LBBB ECG patterns at baseline. ICD patients with LBBB (n = 505) served as a control group.
Data acquisition and patient follow-up
The MADIT-CRT trial was performed from December 22, 2004, through September 2010. After September 10, 2010, there was a long-term follow-up conducted at 48 of 88 U.S. centers requested by the Food and Drug Administration for patients enrolled in the United States (coordinated by the Heart Research Follow-Up Program at the University of Rochester Medical Center, Rochester, New York), and at 23 of the 24 non-U.S. centers (coordinated by the Israeli Association for Cardiovascular Trials at Sheba Medical Center, Tel Hashomer, Israel), involving a total of 854 patients (9). The median follow-up of the enrolled patients was 5.6 years. All patients had a clinical evaluation at each follow-up visit or at any meaningful clinical event.
Evaluation of LV lead position
During the device implantation, implanting physicians in the study were requested to perform a pre-implantation coronary venous angiogram in at least 2 orthogonal views (left anterior oblique, 20° to 40°, and right anterior oblique, 20° to 40°). We also asked the physicians to store post-implantation fluoroscopic images in the same views, and obtain post-procedural chest x-rays (anteroposterior and lateral views) before discharge if possible. These records were copied onto a CD-ROM and sent to the core laboratory at the University of Rochester Medical Center. Physicians were recommended to implant the LV lead along the lateral or posterolateral region of the LV; however, the final LV lead position was left to the physician’s discretion. Final placement of the LV lead was adjudicated by the core laboratory based on the pre-implantation venogram and post-implantation images. The LV epicardial surface was divided into 15 different segments using the 2-view approach (10,11). Using the right anterior oblique view images, the long axis positions were classified such as basal, mid-ventricular, and apical positions (12). When assessing the left anterior oblique view images, we divided the LV wall into 5 equal parts: anterior, anterolateral, lateral, posterolateral, and posterior segments. For this specific analysis, the anterolateral, lateral, and posterolateral segments were grouped together as lateral LV lead location (13).
Echocardiography images were obtained according to a study-specific protocol at baseline before device implantation and at 1 year after CRT-D placement. Echocardiography investigators and sonographers were qualified to perform echocardiography according to the pre-approved echocardiography protocol. Recordings were analyzed off-line at the Brigham and Women's Hospital, Boston, Massachusetts, as an independent echocardiography core laboratory. Echocardiography investigators analyzing the images were blinded to treatment assignment or clinical outcome.
LV volumes were measured by Simpson’s disk method in the apical 4- and 2-chamber views and LVEF was calculated according to the established American Society of Echocardiography protocols (14). Coefficients of variation for end-diastolic volume, end-systolic volume and LVEF were 5.2%, 6.2%, and 5.5%, respectively (15).
Definitions and endpoints
The primary endpoint of the current study was the first occurrence of an HF episode or death from any cause. The secondary endpoint included all-cause mortality.
In our analysis, posterior and lateral LV lead locations were combined because we found similar outcomes of HF or death in these 2 groups.
When analyzing LV reverse remodeling response to CRT, percent reduction (change from baseline to 12-month follow-up) in LV end-diastolic volume (LVEDV), LV end-systolic volume (LVESV), and left atrial volume (LAV) were assessed by LV lead location subgroups.
Continuous variables are expressed as mean ± SD. Categorical data are summarized as frequencies and percentages. Baseline clinical characteristics were compared between the ICD group and apical or non-apical LV lead positions, and between the ICD group and anterior or posterior/lateral LV lead positions using the Kruskal-Wallis test for continuous variables and the chi-square test or Fisher exact test for dichotomous variables.
Cumulative probabilities of all-cause mortality and HF events in the CRT-D LBBB patient subset by ICD versus non-apical LV lead position, and by ICD versus anterior versus posterior/lateral LV lead position was displayed according to the Kaplan-Meier method, with comparisons of cumulative event rates by the log-rank test. Multivariate Cox proportional hazards regression analysis was used to evaluate the impact of different LV lead locations (apical vs. non-apical and anterior vs. posterior/lateral) on the endpoint of long-term all-cause mortality or HF events only. Multivariate Cox proportional hazards regression models were adjusted for relevant clinical covariates using best subset regression modeling. An apical LV lead location was not directly compared to the ICD group due to the extremely low number of events following the third year of follow-up that was deemed insufficient for analysis by the co-authors of this paper.
LV reverse remodeling response was characterized by analyzing the extent of left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV), right ventricular diameter, and dyssynchrony reduction at 12 months as compared to baseline as an efficacy, on-treatment analysis, with comparisons using the nonparametric Kruskal-Wallis test, as appropriate.
All statistical tests were 2-sided; a p value of <0.05 was considered statistically significant. Analyses were performed with SAS software (version 9.4, SAS institute, Cary, North Carolina).
There were a total of 797 patients with CRT-D evaluated for final LV lead location, 569 (71.4%) of them had LBBB at baseline. A total of 505 patients with LBBB and an implanted ICD served as the control group in the current analysis. Of the 569 patients with LBBB, CRT-D, and LV lead location data available, 83 (15%) had an apical LV lead location. The remaining 486 patients (85%) with non-apical LV lead location included 387 patients (68%) with posterior/lateral LV lead location and 99 patients (17%) with anterior LV leads.
Baseline clinical characteristics
Patients with apical LV lead location had a higher rate of short-term device-related complications, better baseline renal function, and higher systolic and diastolic blood pressure as compared to those with non-apical LV leads or an ICD only. Patients with lateral/posterior LV lead location also had a higher rate of device-related complications within 3 months, lower baseline blood urea nitrogen levels, higher ejection fraction, and lower LVEDV compared to those with anterior LV lead location or an ICD only (Table 1).
Associations of LV lead location with long-term HF and death
When we investigated the cumulative probability of HF or death events in patients with LBBB, there was a significantly lower rate of HF or death in patients with non-apical LV lead locations when compared to those with an ICD alone (p < 0.001) (Figures 1A and 1B). After adjustment for relevant clinical covariates, multivariate regression analysis showed that anterior (hazard ratio [HR]: 0.52; p = 0.002) or posterior/lateral LV lead locations (HR: 0.42; p < 0.001) are both associated with significant risk reduction of HF or death when compared to an ICD only. However, there was no difference in the associations with HF or death reduction in CRT-D LBBB patients when we compared anterior or posterior/lateral LV lead locations (Table 2).
Associations of LV lead location with long-term HF events or death
Similar to HF or death reduction, there were significantly lower cumulative probabilities for HF in patients with non-apical, posterior/lateral, or anterior LV lead locations when compared to the ICD-only group (p < 0.001) (Figures 2A and 2B). Multivariate models confirmed significant risk reduction in HF with HRs ranging from 0.36 to 0.39 associated with anterior or posterior/lateral LV lead locations compared to an ICD alone (p < 0.05 for all) (Table 2).
Only LBBB patients with non-apical, posterior/lateral LV lead positions have shown an association with significant risk reduction in mortality from CRT-D (HR: 0.48; p < 0.001) as compared to ICD only (Figures 3A and 3B, Table 2).
LV reverse remodeling to CRT-D by LV lead locations
When evaluating 1-year LV reverse remodeling to CRT-D by LV lead locations, patients with non-apical LV lead locations, especially those with posterior/lateral LV leads, show the greatest LV reverse remodeling with reductions in LVEDV, LAV, and right ventricular diameter (Table 3).
Associations of LV lead location with long-term HF events or death in non-LBBB patients
When assessing clinical outcomes in the non-LBBB cohort by assigned treatment and various LV lead locations, there was no association with significant benefit from CRT-D with non-apical, anterior, or posterior/lateral LV lead locations when compared to an ICD only. However, in patients with CRT-D and an apical LV lead location, there was a trend toward a higher risk of death when compared to ICD only (HR: 2.18; p = 0.068) (Table 4).
In the current MADIT-CRT sub-study, we report that lateral or posterior LV lead locations are associated with long-term all-cause mortality reduction in mild HF patients with CRT-D and LBBB. Furthermore, we found that non-apical short axis positions (anterior and posterior/lateral) were associated with reductions of the combined endpoint of HF or death, or HF alone, when compared to the ICD-only group. Patients with posterior/lateral LV leads showed the greatest improvement in LV, RV, and left atrial volumes, and LV ejection fraction (LVEF), whereas in patients with LV anterior lead locations, less reverse remodeling was observed. Apical LV lead locations should continually be avoided.
Prior studies described LV activation patterns in HF patients with LBBB QRS morphology using conventional activation mapping. In most patients, a U-shaped pattern was observed terminating at the basal region located between the septum and lateral wall (16). Based on these findings it is conceivable that pacing at the lateral-posterolateral region could be the most beneficial for both acute and long-term clinical response in CRT candidates. This hypothesis has been proven by Butter et al. (17), showing stimulation of the LV free wall to be associated with better acute hemodynamic response compared to the anterior sites using both biventricular or univentricular pacing and different atrioventricular delays.
However, to date, data on the association of long-term clinical outcomes by LV lead location in CRT recipients are scarce. Some of the prior investigations showed LV reverse remodeling or functional improvement associated with various LV lead locations in CRT patients, but they failed to show an association with mortality benefit. In the COMPANION (Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure) trial, anterior, lateral, and posterior LV lead locations were investigated separately (4). There were no differences in functional outcomes such as 6-min walk test, quality of life, and New York Heart Association (NYHA) functional class, or in clinical outcomes such as all-cause mortality or HF hospitalizations between the different pacing sites (4).
In prior analyses from the MADIT-CRT in-trial experience, better clinical outcomes were reported to be associated with non-apical LV lead locations; however, an increased ventricular tachyarrhythmia risk was observed in patients with an anterior LV lead position (5,18). Lateral or posterior LV lead locations did not emerge superior to anterior LV lead location in those studies in terms of reduction in HF or death. However, these initial studies were conducted in both patients with LBBB and non-LBBB ECG morphologies. In LBBB patients, lateral or posterior LV pacing seems to be more often in alignment with the latest activated areas within the LV as described above. Such a hypothesis is further supported by our study, showing significantly better improvement in LV parameters and an association with significant reduction in long-term all-cause mortality in patients with posterior or lateral LV lead locations. This is also in alignment with a small prior study by Rossillo et al. (19) showing LV reverse remodeling with posterolateral LV lead positions. However, in their particular study, mortality was similar between anterolateral and posterolateral LV lead positions during a relatively short follow-up of 546 days.
Other studies reported conflicting results on short-term outcomes by LV lead location in CRT recipients. In the REVERSE (REsynchronization reVErses Remodeling in Systolic left vEntricular dysfunction) study, enrolling patients with mild HF, LV dysfunction (≤40% LVEF), and wide QRS (≥120 ms) Gold et al. (20) found that there was a large reduction in LV end-systolic volume index and LVEF by LBBB morphology. By investigating the location of LV leads, Thebault et al. (21) reported that a higher amount of patients improved >15% reduction in LV end-systolic volume index with non-apical positions compared to apical positions (apical position: 120 patients [58%], non-apical position: 12 patients [35%]; p = 0.016), but not between lateral and nonlateral positions (lateral position: 107 patients [57%], nonlateral position: 24 patients [46%]; p = 0.21). The incidence of HF events and all-cause mortality was significantly lower in patients with lateral LV leads (HR: 0.44; 95% confidence interval [CI]: 0.19 to 0.99; p = 0.04), and in patients with LV non-apical leads (HR: 0.27; 95% CI: 0.11 to 0.63; p = 0.001), primarily driven by HF events (21). In the RAFT (Resynchronization-Defibrillation for Ambulatory Heart Failure) trial, Wilton et al. (22) did not find an association of LV lead position with the primary composite endpoint of HF events and all-cause mortality during the mean follow-up time of 39 ± 20 months. However, apical LV lead position was associated with a higher risk of HF hospitalization (HR: 1.99; 95% CI: 1.24 to 3.18; p = 0.004) (22). Our study shows an association of better outcomes of HF and death in LBBB patients with lateral/posterior LV lead location during long-term follow-up.
Our present analysis provides further interesting insights into long-term clinical outcomes by LV lead location in CRT-D patients. However, our findings must be taken with caution. Although we show an association with better outcomes in patients with lateral/posterior LV leads, we acknowledge that an anterior or apical LV lead placement might be due to the inability of implanting the LV lead into a lateral or posterior position. Therefore, an anterior or apical LV lead placement might be simply the marker for lateral and/or anterior scar and patients with scars in these regions have been shown to have worse outcomes (23).
Supporting our prior findings (24), in the present analysis, non-LBBB patients did not experience clinical benefit from CRT-D with any LV lead locations. The current role of CRT-D in mild HF patients with non-LBBB is not well understood and needs further investigation. In non-LBBB patients, LV electrical activation and therefore optimal LV lead location are more heterogeneous (16). In such patients, targeted LV lead location strategies might be warranted (CRT Implant Strategy Using the Longest Electrical Delay for Non–Left Bundle Branch Block Patients [ENHANCE CRT] clinical trial; NCT01983293).
In mild HF patients with CRT-D and LBBB ECG morphology, posterior or lateral LV lead locations were associated with greater LV reverse remodeling and lower risk of all-cause mortality. Therefore, in CRT candidates with LBBB, posterior/lateral LV lead positioning should be considered whenever possible. Quadripolar leads are preferable in most cases to achieve a basal or mid-ventricular pacing site.
Our study is a post hoc analysis of a randomized clinical study, LV lead location was not randomized, and the study was not designed to assess outcomes by LV lead locations. Therefore, causal relationship between LV lead location and long-term outcomes cannot be established; only associations are described. The investigated groups had different sample sizes and differences in baseline clinical characteristics. Because of the small sample size in the apical LV group and low long-term event rates, we were limited for analyses of long-term outcomes. Furthermore, final LV lead locations might have been influenced by available coronary sinus branches for lead placement in our study. Therefore, anterior or apical LV lead location might be simply a marker for extensive lateral and/or anterior scar. In some patients, we could not reliably assess the final LV lead locations; however, we still have 1 of the largest cohort of patients with centrally adjudicated LV lead location data.
In mild HF patients with LBBB ECG morphology implanted with CRT-D, posterior or lateral LV lead location was associated with significant long-term mortality reduction, accompanied by LV reverse remodeling. Furthermore, there was a significant reduction in long-term HF or death and in HF alone in CRT-D patients with anterior and posterior/lateral LV lead locations. An apical LV lead location should nevertheless be avoided due to the high early mortality risk shown in prior studies.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients with LBBB ECG morphology implanted with cardiac resynchronization therapy benefit from posterior or lateral LV lead location, accompanied by LV reverse remodeling. An apical LV lead location should be avoided.
TRANSLATIONAL OUTLOOK: Further studies are needed to better understand the role of LV lead location in cardiac resynchronization therapy patients with non-LBBB at baseline.
↵∗ Drs. Kutyifa, Kosztin, Merkely, and Singh contributed equally to this sub-study.
The MADIT-CRT study was supported by a research grant from Boston Scientific, St. Paul, Minnesota, to the University of Rochester School of Medicine and Dentistry. Dr. Kutyifa has received grants from Boston Scientific and ZOLL Inc.; and has received honoraria from Biotronik and ZOLL Inc. Dr. Klein has received grants from Boston Scientific and Zoll; and has received personal fees from Zoll. Dr. Solomon has received grants from Boston Scientific. Dr. Zareba has received grants from Boston Scientific. Dr. Moss has received grants from Boston Scientific. Dr. Merkeley has received grants from Boston Scientific; and has received personal fees from Medtronic, Biotronik, and St. Jude Medical. Dr. Singh has received personal fees from St. Jude Medical, Boston Scientific, Medtronic, and Sorin. All other authors have reported that they have no relationships relevant to the content of this paper to report.
Angelo Auricchio, MD, served as Guest Editor for this paper.
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
- cardiac resynchronization therapy
- cardiac resynchronization therapy with defibrillator
- heart failure
- implantable cardioverter-defibrillator
- intraventricular conduction disturbances
- left bundle branch block
- left ventricular
- left ventricular end-diastolic volume
- left ventricular ejection fraction
- left ventricular end-systolic volume
- New York Heart Association
- right bundle branch block
- Received May 2, 2017.
- Revision received July 3, 2018.
- Accepted July 12, 2018.
- 2018 American College of Cardiology Foundation
- Zareba W.,
- Klein H.,
- Cygankiewicz I.,
- et al.
- Epstein A.E.,
- DiMarco J.P.,
- Ellenbogen K.A.,
- et al.
- Vardas P.E.,
- Auricchio A.,
- Blanc J.J.,
- et al.
- Fantoni C.,
- Raffa S.,
- Regoli F.,
- et al.
- Singh J.P.,
- Houser S.,
- Heist E.K.,
- Ruskin J.N.
- Lang R.M.,
- Bierig M.,
- Devereux R.B.,
- et al.
- Solomon S.D.,
- Foster E.,
- Bourgoun M.,
- et al.
- Auricchio A.,
- Fantoni C.,
- Regoli F.,
- et al.
- Butter C.,
- Auricchio A.,
- Stellbrink C.,
- et al.
- Singh J.P.,
- Klein H.U.,
- Huang D.T.,
- et al.
- Gold M.R.,
- Thebault C.,
- Linde C.,
- et al.
- Wilton S.B.,
- Exner D.V.,
- Healey J.S.,
- et al.
- Tompkins C.,
- Kutyifa V.,
- McNitt S.,
- et al.