Author + information
- Anna C. Kydd, MD∗,
- Fakhar Z. Khan, MD∗,
- William D. Watson, MD∗,
- Peter J. Pugh, MD∗,
- Munmohan S. Virdee, MD† and
- David P. Dutka, DM∗∗ ()
- ∗Department of Cardiovascular Medicine, University of Cambridge, Addenbrooke’s Hospital, Cambridge, United Kingdom
- †Department of Cardiology, Papworth Hospital, Cambridge, United Kingdom
- ↵∗Reprint requests and correspondence:
Dr. David P. Dutka, Department of Cardiovascular Medicine, University of Cambridge, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 0QQ, United Kingdom.
Objectives This study was conducted to assess the impact of left ventricular (LV) lead position on longer-term survival after cardiac resynchronization therapy (CRT).
Background An optimal LV lead position in CRT is associated with improved clinical outcome. A strategy of speckle-tracking echocardiography can be used to guide the implanter to the site of latest activation and away from segments of low strain amplitude (scar). Long-term, prospective survival data according to LV lead position in CRT are limited.
Methods Data from a follow-up registry of 250 consecutive patients receiving CRT between June 2008 and July 2010 were studied. The study population comprised patients recruited to the derivation group and the subsequent TARGET (Targeted Left Ventricular Lead Placement to guide Cardiac Resynchronization Therapy) randomized, controlled trial. Final LV lead position was described, in relation to the pacing site determined by pre-procedure speckle-tracking echocardiography, as optimal (concordant/adjacent) or suboptimal (remote). All-cause mortality was recorded at follow-up.
Results An optimal LV lead position (n = 202) conferred LV remodeling response superior to that of a suboptimal lead position (change in LV end-systolic volume: –24 ± 15% vs. –12 ± 17% [p < 0.001]; change in ejection fraction: +7 ± 8% vs. +4 ± 7% [p = 0.02]). During long-term follow-up (median: 39 months; range: <1 to 61 months), an optimal LV lead position was associated with improved survival (log-rank p = 0.003). A suboptimal LV lead placement independently predicted all-cause mortality (hazard ratio: 1.8; p = 0.024).
Conclusions An optimal LV lead position at the site of latest mechanical activation, avoiding low strain amplitude (scar), was associated with superior CRT response and improved survival that persisted during follow-up.
Cardiac resynchronization therapy (CRT) has been an important advance in the treatment of patients with heart failure, leading to improvements in symptoms, left ventricular (LV) function, and survival (1–3). However, a significant proportion of recipients, up to 30%, still do not achieve a favorable response. In addition to patient selection and device optimization, LV lead position is of fundamental importance in delivering effective CRT, and as such was included in recent international CRT guidelines (4).
Early retrospective studies indicated that an LV lead placed at the site of latest mechanical activation, away from scar, and in a nonapical location conferred greatest benefit (5–7). Characterization of both myocardial substrate (8) and mechanical activation pattern (9) in the LV using speckle-tracking echocardiography (STE) can be applied to determine the optimal LV pacing site in individual patients (10). The utility of this approach to guide the implanter has been demonstrated in 2 prospective, randomized trials: TARGET (Targeted Left Ventricular Lead Placement to guide Cardiac Resynchronization Therapy) (11) and STARTER (Speckle Tracking Assisted Resynchronization Therapy for Electrode Region) (12). Benefits observed in the echo-guided groups in both studies included a larger proportion of patients with a favorable LV lead position, greater LV remodeling response, and improved survival at up to 2 years' follow-up.
On the basis of previous studies, it was observed that the initial beneficial effects of CRT are maintained, and that the impact on survival persists over a longer duration of follow-up (13). Therefore, we examined the effects of LV lead position on all-cause mortality following CRT in a prospective cohort of patients.
We studied data from a total of 250 patients who underwent CRT implantation between June 2008 and July 2010. An initial group of 50 patients participated in an open-label study that assessed the feasibility of using STE to guide LV lead placement to achieve the greatest increase in LV performance. A randomized trial was then undertaken that compared conventional LV lead placement to a targeted approach in which the LV lead was placed in a segment defined on echocardiography (TARGET ). This registry comprises data from the 50 patients in the derivation group studied prospectively from June 2008 to March 2009, and 200 patients from the TARGET randomized, controlled trial, which enrolled from April 2009 to July 2010 (Fig. 1). There were no differences in the baseline characteristics between the 2 groups, including age, etiology, ejection fraction (EF), and dyssynchrony. Patient characteristics, study methods, and results from the TARGET study (11) have been published previously in detail; although 247 patients were initially screened, 27 were not eligible for inclusion in the trial due to suboptimal echocardiographic imaging. Echocardiographic image quality was satisfactory for STE in all patients in the derivation group. Patients who died before CRT implantation or in whom there was failure to implant an LV lead were not included in this cohort follow-up study.
All patients met standard indications for CRT implantation, with impaired LV function (EF ≤35%), left branch bundle block or intraventricular conduction delay (QRS duration ≥120 ms), and New York Heart Association (NYHA) functional class III/IV symptoms despite optimal medical therapy, and had satisfactory echocardiographic image quality for STE analysis. Ischemic etiology was defined as the presence of coronary stenosis >50% by invasive coronary angiography. All patients underwent detailed clinical assessment, including a 6-min walk test, a quality-of-life questionnaire (Minnesota Living with Heart Failure Questionnaire), and echocardiography before and 6 months after CRT device implantation. LV remodeling response was pre-defined as a ≥15% reduction in left ventricular end-systolic volume (LVESV) at 6 months. Ongoing data on all-cause mortality were collected, with a maximal follow-up duration of 5 years. The study protocol was approved by the local ethics committee.
Standard 2-dimensional (2D) echocardiography was performed using a commercially available ultrasound system (Vivid 7, GE Healthcare Vingmed, Wauwatosa, Wisconsin) and a 3.5-MHz transducer. Standard grayscale images and color Doppler data were acquired in a cine-loop format and digitally stored for off-line analysis (EchoPac version 7.0, GE Medical Systems, Horten, Norway). LVESV, left ventricular end-diastolic (LVEDV), and EF according to the biplane Simpson method, were derived from standard 2- and 4-chamber apical views. Intraventricular dyssynchrony was assessed by speckle-tracking radial strain as the time delay between the anteroseptal and posterior segments. A previously defined cutoff ≥130 ms was considered significant (14). The echocardiographic data were analyzed by assessors blinded to all other patient-related data.
Speckle-tracking radial strain analysis was performed on the pre-implantation basal and mid-LV short-axis images (frame rate ≥40 Hz) using a standardized approach, as previously described (15). The endocardial border was traced in end-systole using a point-and-click technique, with care taken to adjust tracking of all segments. A second, larger, concentric circle was then automatically generated and manually adjusted near the epicardium so that the area of interest included the entire myocardial wall. Tracking was fine-tuned by visual assessment of the cine image to ensure appropriate tracking of all segments throughout the cardiac cycle. Time-strain curves for each of the 12 nonapical segments were analyzed for time to peak strain and strain amplitude. Time to peak strain was recorded to identify the latest segment (most delayed peak) from the onset of the QRS. If >1 latest segment was identified, all late segments were reported to the implanter. A segmental strain amplitude of <10% was considered to represent scar, as previously described (11).
Optimal LV pacing site was defined, on the basis of both timing and amplitude of segmental radial strain, as the segment with the longest time to peak strain, with preserved strain amplitude ≥10%.
CRT implantation and final LV lead position
CRT implantation was undertaken via a transvenous approach in accordance with the TARGET study protocol. In the group randomized to echo-guided CRT, an attempt was made to position the LV lead at the optimal site, defined by 2D speckle-tracking radial strain. The remaining patients underwent standard CRT implantation, without echocardiographic guidance, with the LV lead placed preferentially in a posterolateral or lateral vein. Optimization of atrioventricular and interventricular delays using echocardiography was performed on the day after implantation. Further optimization was performed subsequently according to clinical need.
Final LV lead position was determined by an independent assessor using biplane fluoroscopy and post-implantation chest radiography, as previously reported (16). LV lead position was described as basal, mid, or apical on the basis of anteroposterior and right anterior oblique views, and as anterior, lateral, posterior, or inferior on the basis of the left anterior oblique view. The agreement between optimal pacing site, defined on echocardiography, and final LV lead position was described as concordant if positioned at the optimal site, adjacent if placed within 1 segment of the optimal site, or remote if ≥2 segments from the optimal site. In cases in which the final LV lead position was apical, this was also described as either adjacent or remote with respect to the optimal site defined on echocardiography.
Data from patients with a concordant (n = 139) or adjacent (n = 63) (i.e., optimal) LV lead position were combined for further analysis and compared to those from the group with a remote (suboptimal) lead position. This was undertaken to reflect clinical practice, in which achieving a concordant lead position may be limited by venous anatomy, lead stability, or phrenic nerve stimulation and an adjacent lead position may be the only feasible option. Furthermore, optimization in patients with an adjacent LV lead position can enhance response and negate potential deleterious effects, in contrast to optimization in patients with a remote LV lead position (17).
Continuous data are presented as median (interquartile range) or mean ± SD, and categorical data, as n (%). Continuous variables were compared using the Mann-Whitney U and Student t tests as appropriate. Categorical variables were compared by the use of the chi-square test. Kaplan-Meier curves were plotted to describe all-cause mortality according to LV lead position (either concordant/adjacent or remote from the optimal pacing site), and data from these 2 groups were compared using the log-rank test. Cox univariate proportional hazards analysis was performed to identify the baseline characteristics that predicted all-cause mortality. Important clinical and echocardiographic characteristics (p < 0.1) were then included in a forward stepwise multivariate model. Analysis was performed using SPSS statistical software version 19.0 (IBM SPSS Statistics, IBM Corporation, Armonk, New York) and GraphPad Prism version 6 (GraphPad Software, Inc., San Diego, California). A p value of <0.05 was considered significant.
Baseline patient characteristics according to final LV lead position are presented in Table 1. Patients with a remote LV lead were older than those with a concordant/adjacent LV lead position. Clinical and echocardiographic characteristics were otherwise similar between the 2 groups, and there was no difference in the proportions of patients receiving a CRT defibrillator.
The distribution of final LV lead anatomic location is shown in Table 2, stratified according to concordant/adjacent or remote position. Although a greater proportion of LV leads in the remote group were of anterior location (21% vs. 4%; p < 0.001) the proportion of remote LV leads (77%) located either laterally or posteriorly was similar to that in the concordant/adjacent group (89%; p = 0.06). The majority of patients had a nonapical LV lead position (88% vs. 11%). The proportions of patients in whom the LV lead was placed at the apex did not differ between the 2 groups (11% in the adjacent group vs. 10% in the remote group; p = 0.8).
Clinical and LV remodeling response to CRT
At 6-month follow-up, there were overall improvements in LVEDV (193 ± 65 ml vs. 162 ± 59 ml; p < 0.001), LVESV (149 ± 56 ml vs. 116 ± 49 ml; p < 0.001), and EF (23 ± 7% vs. 30 ± 10%; p < 0.001) compared with baseline, and 63% (n = 157) of patients were considered CRT responders (≥15% reduction in LVESV). Patients with a concordant/adjacent LV lead position demonstrated a superior response to CRT (72% vs. 25% response rate in the remote group; p < 0.001) (Fig. 2), with a greater relative reduction in LVESV (–24 ± 15% vs. –12 ± 17%; p < 0.001) and greater absolute increase in EF (+7 ± 8% vs. +4 ± 7%; p = 0.02). Patients with a concordant/adjacent LV lead also had a greater improvement in NYHA functional class, heart failure questionnaire score, and 6-min walk test score (Table 3).
Long-term survival according to LV lead location and scar
During a median follow-up duration of 39 months (range: <1 to 61 months) there were 70 deaths (49 in the concordant/adjacent group, 21 in the remote group). The survival rate in the group with concordant/adjacent LV lead position was 70% compared with 38% in the group with a remote LV lead position (log-rank p = 0.003) (Fig. 3). A total of 7 patients (15%) in the remote LV lead group had died before 6-month follow-up compared with 4 patients (2%) in the concordant/adjacent group (p < 0.001). Analysis according to the presence of low strain amplitude (scar) at the LV lead position demonstrated an impact on longer-term survival. At 5-year follow-up, patients without low strain amplitude at the LV lead site had a higher survival rate (log-rank p < 0.01) (Fig. 4). An apical (n = 28) versus nonapical (n = 222) LV lead position did not affect survival (log-rank p = 0.48). On Cox proportional hazards analysis, remote LV lead position was an independent predictor of impaired survival (Table 4).
The beneficial effect of CRT on all-cause mortality in patients with a concordant or adjacent LV lead position persisted during long-term follow-up. Our definition of optimum LV lead position combines the assessment of both dyssynchrony (the latest segment of mechanical activation) and absence of scar (preserved strain amplitude). This is the longest reported prospective follow-up using this definition of optimum pacing site and extends previous findings from our group and those of others (5,6). In addition, a greater extent of LV remodeling, resulting in a greater proportion of CRT volume responders, and greater symptomatic improvement were observed in the concordant/adjacent group after 6 months. A remote LV lead position was an independent predictor of all-cause mortality.
The early divergence of the Kaplan-Meier curves, stratified according to LV lead position, suggests that the disadvantageous effect of a remotely placed LV lead occurs soon after implantation. After a follow-up duration of 46 months, survival in the remote group was 54.3%, qualitatively this is comparable to that in the medical-therapy arm in the CARE-HF (Cardiac Resynchronization–Heart Failure) trial (1). Data to support the concept that correction of mechanical dyssynchrony is the major mechanism of benefit from CRT are accumulating (14,18,19); however, the potential for CRT to worsen mechanical dyssynchrony remains largely unaddressed. It is conceivable that a suboptimal (remote) LV lead position may increase electromechanical activation time, leading to either no change or even deterioration in dyssynchrony and LV performance.
This potential for a deleterious effect of CRT has been alluded to in retrospective observational data from Auger et al. (20) and Sanderson and Yu (21) showing that mechanical dyssynchrony (defined by tissue Doppler imaging) not present at baseline can actually be induced in some patients after CRT and results in less favorable outcomes. Further mechanistic insight can be derived from the dyssynchrony data presented in the STARTER trial (12), in which greater mechanical resynchronization was observed in a group of patients receiving an echo-guided LV lead, comprising a smaller proportion of remote leads than that in the control group. In addition, a reduction in mechanical dyssynchrony was associated with reduced mortality and heart failure–related hospitalizations (12).
Our current findings add further support for an individualized lead-targeting strategy to optimize CRT device implantation, matching the LV lead position to the underlying ventricular mechanical dysfunction. The importance of this concept is emphasized by the high proportion of patients categorized in the remote group with either a lateral or posterior LV lead, which is in contrast with early data regarding anatomic LV lead placement that would actually favor these sites. The corollary is that an anterior LV lead position may in fact be optimal in some cases, although in the majority of patients it should be avoided. A pacing strategy on the basis of pre-procedure STE is one approach that, with supporting randomized data, can be used to achieve a prospectively targeted LV lead to the latest activated myocardial segment while avoiding scar. Our group is experienced in this analysis, which is performed on routinely acquired 2D grayscale short-axis echocardiographic views. The available published reports suggest that achieving satisfactory image quality to perform this type of analysis in this patient group is feasible in the vast majority (∼85% to 90%) of cases, in keeping with our clinical experience (12,15). The performance of this approach in a nonspecialist center and in a more general setting is unknown. An alternative strategy using cardiac magnetic resonance imaging to avoid pacing segments of scar has also been shown to improve clinical outcomes, including all-cause mortality, cardiovascular disease–related death, and heart failure–related hospitalization (22). A multimodality imaging approach incorporating cardiac magnetic resonance scar and dyssynchrony assessment integrated with fluoroscopy has also been utilized to facilitate lead targeting (23). Ultimately, the optimum imaging modality may depend on local availability and expertise. The fundamental concept underpinning this approach is that CRT implantation is enhanced when tailored to treat patient-specific LV mechanical dysfunction.
Using the current standard transvenous implantation technique, final LV lead position is inherently dependent on coronary venous anatomy, accessibility, and lead stability. In the echo-guided LV lead groups in both the TARGET and STARTER studies, 10% and 15% of patients, respectively, had a remotely positioned LV lead. For this relatively small proportion of patients, an alternative implantation strategy may be advantageous. For example, a surgically placed epicardial LV lead, or the emerging approach of an endocardial LV lead, may overcome constraints relating to the coronary sinus and its tributaries. Given that a remotely placed LV lead may be no better than medical therapy alone when considering long-term survival, the option of avoiding implantation of a remote LV lead, or using one of these less conventional approaches, warrants consideration. This may be particularly important as CRT guidelines have been extended to include less symptomatic patients in NYHA functional class II (24,25).
The limitations of the TARGET study were described in the publication, and we have subsequently reported the utility of 2D radial strain imaging to identify the optimal LV pacing site to enhance response to CRT (11). Recent advances in imaging technology were not used as the initial characterization of subjects was based on echocardiographic analysis at the time of recruitment into the The TARGET study in 2008 to 2010. The data at 5 years was complete, aided by the fact that the subjects were attending 2 teaching hospitals for device follow-up and accurate recording of death through the primary care system in the United Kingdom.
We have shown that the initial benefits derived from CRT according to final LV lead position are sustained over longer-term follow-up. Remote placement of the LV lead was associated with early reduction in survival after implantation and was an independent predictor of impaired prognosis. Optimal device implantation requires a strategy incorporating the assessment of myocardial substrate and underlying myocardial dysfunction to guide LV lead position, and this can be readily achieved using pre-procedure 2D STE. In patients in whom anatomic or lead-stability constraints prevent an optimal location being achieved, an alternative strategy is perhaps warranted.
This work was supported by the Cambridge Biomedical Research Centre, funded from the U.K. National Institute for Health Research. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiac resynchronization therapy
- ejection fraction
- left ventricle
- left ventricular end-diastolic volume
- left ventricular end-systolic volume
- New York Heart Association
- speckle-tracking echocardiography
- Received November 1, 2013.
- Accepted November 7, 2013.
- American College of Cardiology Foundation
- Bristow M.R.,
- Saxon L.A.,
- Boehmer J.,
- et al.,
- Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) Investigators
- Daubert J.C.,
- Saxon L.,
- Adamson P.B.,
- et al.,
- European Heart Rhythm Association (EHRA); European Society of Cardiology (ESC); Heart Rhythm Society; Heart Failure Society of America (HFSA); American Society of Echocardiography (ASE); American Heart Association (AHA); European Association of Echocardiography (EAE) of ESC; Heart Failure Association of ESC (HFA)
- Ypenburg C.,
- van Bommel R.J.,
- Delgado V.,
- et al.
- Delgado V.,
- van Bommel R.J.,
- Bertini M.,
- et al.
- Singh J.P.,
- Klein H.U.,
- Huang D.T.,
- et al.
- Becker M.,
- Hoffmann R.,
- Kuhl H.P.,
- et al.
- Khan F.Z.,
- Virdee M.S.,
- Palmer C.R.,
- et al.
- Saba S.,
- Marek J.,
- Schwartzman D.,
- et al.
- Gorcsan J. 3rd.,
- Oyenuga O.,
- Habib P.J.,
- et al.
- Suffoletto M.S.,
- Dohi K.,
- Cannesson M.,
- Saba S.,
- Gorcsan J. 3rd.
- Khan F.Z.,
- Virdee M.S.,
- Read P.A.,
- et al.
- Tanaka H.,
- Nesser H.J.,
- Buck T.,
- et al.
- van Bommel R.J.,
- Gorcsan J. 3rd.,
- Chung E.S.,
- et al.
- Auger D.,
- Bleeker G.B.,
- Bertini M.,
- et al.
- Sanderson J.E.,
- Yu C.M.
- Shetty A.K.,
- Duckett S.G.,
- Ginks M.R.,
- et al.
- Tracy C.M.,
- Epstein A.E.,
- Darbar D.,
- et al.
- McMurray J.J.,
- Adamopoulos S.,
- Anker S.D.,
- et al.