- Email: [email protected]
Time interval from left ventricular stimulation to QRS onset is a novel predictor of nonresponse to cardiac resynchronization therapy
Accepted Manuscript Time interval from left ventricular stimulation to QRS onset is a novel predictor of nonresponse to cardiac resynchronization therapy Daigo Yagishita, MD., Morio Shoda, MD. PhD., Yoshimi Yagishita, MD. PhD., Koichiro Ejima, MD. PhD., Nobuhisa Hagiwara, MD.PhD PII:
S1547-5271(18)30906-8
DOI:
10.1016/j.hrthm.2018.08.035
Reference:
HRTHM 7730
To appear in:
Heart Rhythm
Received Date: 1 June 2018
Please cite this article as: Yagishita D, Shoda M, Yagishita Y, Ejima K, Hagiwara N, Time interval from left ventricular stimulation to QRS onset is a novel predictor of non-response to cardiac resynchronization therapy, Heart Rhythm (2018), doi: 10.1016/j.hrthm.2018.08.035. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Native QRS
A
100ms
Ⅰ
Left Ventricular ACCEPTED MANUSCRIPT
B
pacing
C
100ms
Ⅰ
Biventricular pacing Ⅰ
Ⅱ
Ⅱ
Ⅲ
Ⅲ
Ⅲ
aVR
aVR
aVL
aVL
aVF
aVF
V1
V1
SC M AN U
V2
Q
V3
V4
V4
V5
V5
V6
V6
RA
RA
EP
V3
TE D
V2
RI PT
Ⅱ
Q
LV2-3 LV3-4 LV Q-LV QRSd
aVL aVF V1 V2 V3 V4 V5 V6
RV
RV
LV1-2
aVR
RA
AC C
RV
100ms
LV1-2
LV1-2
LV2-3
LV2-3
LV3-4
LV3-4 STIM
S-QRS LV pacing QRSd
Actual LV pacing QRSd
STIM
BiV pacing QRSd
ACCEPTED MANUSCRIPT 1
Time interval from left ventricular stimulation to QRS onset is a novel
2
predictor of non-response to cardiac resynchronization therapy
RI PT
1
3
Daigo Yagishita MD.*, Morio Shoda MD. PhD.*†, Yoshimi Yagishita MD. PhD.*,
5
Koichiro Ejima MD. PhD.*†, Nobuhisa Hagiwara MD.PhD.*
M AN U
SC
4
6 7
Total word count: 4999
10
Brief title: Impact of left ventricular stimulation to QRS interval on CRT
EP
9
TE D
8
* Department of Cardiology, Tokyo Women’s Medical University
12
† Clinical Research Division for Heart Rhythm Management, Tokyo Women’s
13
Medical University
AC C
11
14 15
The authors do not have grants, conflicts, or relationships with industry.
ACCEPTED MANUSCRIPT
Correspondence:
2
Morio Shoda M.D., Ph.D.
3
8-1 Kawada-cho, Shinjuku-ku, Tokyo, Japan
4
Tel: +81.3.3353.8111
5
Fax: +81.3.3356.0441
6
Email: [email protected]
7
11 12 13 14 15
EP
10
AC C
9
TE D
8
M AN U
SC
1
RI PT
2
ACCEPTED MANUSCRIPT 3
Abstract
2
Background: A left ventricular (LV) lead placement at the late activation site (LAS)
3
has been proposed as an optimal LV pacing site (i.e. Q-LV interval). However, LAS
4
may be relevant to local electrical conduction, measured as an interval from LV
5
pacing stimulation to QRS onset (S-QRS interval).
6
Objectives: This study sought to evaluate the prognostic value of S-QRS for reverse
7
remodeling, and the impact of S-QRS on pacing QRS configuration in patients
8
undergoing cardiac resynchronization therapy (CRT).
9
Methods: Sixty consecutive heart failure patients with a wide QRS complex
TE D
M AN U
SC
RI PT
1
underwent CRT. A site with Q-LV≥95ms was targeted for LV lead placement. A
11
responder was defined as one with greater than 15% reduction in LV end systolic
12
volume at six months after CRT.
13
Results: An LV lead placement with Q-LV≥95ms was achieved in 52 of 60 patients
14
(86.7%). Thirty-two of 52 patients were responders (61.5%). S-QRS was
15
significantly shorter in responders than non-responders (P<0.01), whereas Q-LV
AC C
EP
10
ACCEPTED MANUSCRIPT 4
was not significantly different. A cutoff value of 37ms for S-QRS had a sensitivity
2
and specificity of 81% and 90%, respectively. Shorter S-QRS (<37ms) showed
3
significantly narrower LV pacing QRS width and biventricular pacing QRS width
4
compared to longer S-QRS. After multivariate analysis, PQ interval (OR=0.97,
5
P=0.01) and long S-QRS greater than 37ms (OR=0.014, P<0.01) were independent
6
predictors for response to CRT.
7
Conclusion: In addition to a sufficient Q-LV, S-QRS can be a useful indicator of
8
optimal LV lead position to achieve reverse remodeling. S-QRS contributes to the
9
pacing QRS configuration associated with CRT response.
SC
M AN U
TE D
EP
10
RI PT
1
Key words: Cardiac resynchronization therapy, Heart failure, Electrocardiograms,
12
Left ventricular lead, Prognosis, Scar
13 14 15
AC C
11
ACCEPTED MANUSCRIPT 5
Introduction
2
Cardiac resynchronization therapy (CRT) is an established treatment for heart
3
failure patients due to left ventricular (LV) dysfunction and QRS prolongation
4
However, some candidates fail to achieve clinical or echocardiographic benefit from
5
CRT, which has been attributed in part to a suboptimal LV lead position 5.
6
LV lead placement at the latest electrical activation site has been proposed as an
7
important factor for a superior CRT response compared with conventional
8
anatomical approach. The time interval from QRS onset to the local LV activation at
9
the LV lead (Q-LV) is a common indicator determining optimal LV pacing site, and is
10
associated with acute hemodynamic improvement, LV reverse remodeling and
11
clinical outcome 6-9.
12
LV late activation site (LAS) may be due to scarring or fibrosis within the viable
13
myocardium where an anatomical or functional conduction block can occur.
14
Therefore, LV pacing within those areas may be less effective for CRT 10-13.
15
Local conduction disturbance during LV pacing (i.e. stimulus to QRS interval,
1-4
.
AC C
EP
TE D
M AN U
SC
RI PT
1
ACCEPTED MANUSCRIPT 6
S-QRS) may indicate that the LV lead has been placed on a scar or fibrotic region.
2
Furthermore, S-QRS may influence the QRS configuration during LV pacing or
3
biventricular pacing, which are predictors for CRT response 14-16.
4
We therefore hypothesized that long S-QRS would also predict CRT non-response.
5
The present study aimed 1) to evaluate the prognostic value of S-QRS on LV
6
reverse remodeling; and 2) to elucidate the relevance of S-QRS on the QRS
7
configurations during LV and biventricular pacing in CRT patients with optimal LV
8
lead position at LAS.
TE D
M AN U
SC
RI PT
1
9 Methods
11
Ethics Statement
12
This study was a retrospective study approved by the Ethics Committee of the
13
Tokyo Women’s Medical University and conducted in accordance with the
14
Declaration of Helsinki. Written informed consent was given by all patients or their
15
guardians.
AC C
EP
10
ACCEPTED MANUSCRIPT 7
1 Study Population
3
All heart failure patients who underwent CRT implantation between January 2012
4
and February 2016 at our hospital were reviewed for study eligibility. Inclusion
5
criteria were based on our country’s guideline, specifically for sinus rhythm, New
6
York Heart Association functional class II to IV heart failure on optimal
7
pharmacological therapy, QRS duration (QRSd)≥120ms, and left ventricular
8
ejection fraction (LVEF)≤35%. Exclusion criteria were age≤18 years, chronic atrial
9
tachycardia/fibrillation,
TE D
M AN U
SC
RI PT
2
and
complete
atrioventricular
block.
Patients
with
Q-LV<95ms at the LV pacing site were excluded because recent reports indicate
11
this was an inadequate LV pacing site 6, 7.
AC C
12
EP
10
13
CRT Implantation Procedure
14
CRT devices were implanted in the standard fashion, and LV leads were
15
transvenously implanted in a coronary vein branch. During the procedure,
ACCEPTED MANUSCRIPT 8
simultaneous recording of 12 lead ECG and intracardiac electrograms were
2
continuously obtained using a digital recording system (Pruka CardioLab System,
3
GE Healthcare, Waukesha, WI). The recorded intracardiac electrograms were
4
bandpass filtered between 0.5Hz and 500Hz. Q-LV mapping was performed with a
5
decapolar catheter or LV lead electrodes, and the latest LV activation site was
6
targeted as the optimal pacing site in all patients.
M AN U
SC
RI PT
1
7
Electrocardiographic measurements
9
The electrocardiographic measurements were retrospectively performed by two
10
independent reviewers blinded to the outcome. Q-LV was defined as the time
11
interval from the earliest QRS onset to the first major peak of the local bipolar
12
electrogram. The %Q-LV was calculated as the percentage of Q-LV in native QRSd.
13
S-QRS was defined as the time from the LV pacing to the earliest onset of QRS. LV
14
pacing QRSd (LVp-QRSd) was measured from the pacing to the offset of QRS,
15
while the subtraction of S-QRS from LVp-QRSd was termed as the actual LV pacing
AC C
EP
TE D
8
ACCEPTED MANUSCRIPT 9
QRSd (actual LVp-QRSd). Similarly, simultaneous biventricular pacing QRSd
2
(BiV-QRSd) was measured from the pacing to the offset of QRS (Figure1). Finally,
3
the difference between native QRSd and BiV-QRSd (Delta-QRSd) was calculated.
4
These parameters were measured with a caliper in the recording system at a sweep
5
speed of 200mm/s. LV pacing was performed with output at pacing threshold (1.1±
6
0.6V at 0.4ms), and pacing rate was programmed at 10 beats per minute faster than
7
the sinus rhythm.
M AN U
SC
RI PT
1
9
TE D
8 Echocardiographic assessment All
patients
had
a
comprehensive
echocardiographic
assessment
using
11
commercially available echocardiographic systems and data at baseline and six
12
months after CRT implantation were stored digitally. Standard LV volumetric
13
measurements were performed using the recommended Simpson’s biplane method.
14
Intraventricular dyssynchrony was assessed by radial speckle tracking analysis of
15
the anteroseptal and posterolateral mid LV segments
AC C
EP
10
17
. A CRT responder was
ACCEPTED MANUSCRIPT 10
defined as one who had a greater than 15% reduction in LV end systolic volume
2
(LVESV) at six months compared with baseline.
3
RI PT
1
Optimization of CRT device programing
5
VVD optimization was performed to minimize BiV-QRSd during procedure,
6
thereafter, echocardiography was performed to maximize pulsed wave LV outflow
7
tract velocity time integral. AVD optimization was performed by Doppler
8
echocardiography to provide the maximum left ventricular filling time without
9
interference of atrial wave and diastolic mitral regurgitation. Following to AVD
M AN U
TE D
EP
11
optimization, BiV-QRSd and QRS morphology were confirmed again.
AC C
10
SC
4
12
Statistical Analysis
13
All statistical analyses were performed using JMP version 13.0 (SAS Institute, Cary,
14
NC, USA). Continuous variables are expressed as the mean±SD, while categorical
15
variables are presented as frequency and percentage. A linear regression analysis
ACCEPTED MANUSCRIPT 11
was applied to study the correlation between percent reduction in LVESV and
2
S-QRS or Q-LV. For the univariate analysis, Fisher’s exact test and Wilcoxon exact
3
test was used. Receiver operating characteristic (ROC) curve analysis was
4
performed to determine the optimal cut-off value of the S-QRS for predicting CRT
5
non-responders. A multivariate logistic regression model was constructed to assess
6
the association of clinical variables to predict the response to CRT. P < 0.05 was
7
used to select variables from the univariate analysis. All statistical tests were two
8
tailed, and P value < 0.05 was considered as significant.
9
TE D
M AN U
SC
RI PT
1
Results
11
Baseline patient characteristics
12
Sixty consecutive CRT patients were enrolled in the present study. Of these, eight
13
patients were excluded due to suboptimal LV lead position without sufficient
14
electrical latency. The final study population included 52 patients (79% male) with a
15
mean age of 62.6±13.9 years (Table 1). Mean native QRSd was 159.5±34.0ms,
AC C
EP
10
ACCEPTED MANUSCRIPT 12
and mean LVEF was 26.0±6.2%. In the QRS morphological features, 36 patients
2
(69%) had LBBB, while 16 patients (31%) had non-LBBB. Non-ischemic etiology
3
and ischemic disease was present in 75% and 25% of patients, respectively. There
4
were 18 patients classified as NYHA
5
optimal medical therapy before CRT implantation. According to LV volumetric
6
evaluation results before and six months after CRT implantation, 32 patients
7
(61.5%) were responders. Patients were divided into a responder and a
8
non-responder group to compare their characteristics. Baseline characteristics were
9
not significantly different except defibrillator device indication.
TE D
M AN U
SC
, and all patients were under the
EP
10
and
RI PT
1
Echocardiographic parameters before and six months after CRT implantation
12
At baseline, patients had severely reduced LV systolic function with enlarged LV
13
with mean LVESV at 197.0±96.1ml. Severe radial dyssynchrony was present at
14
204.5±120.1ms, and mean GLS was reduced at 9.0±3.5%. There were no
15
significant differences in LV volumetric evaluation and radial dyssynchrony between
AC C
11
ACCEPTED MANUSCRIPT 13
responders and non-responders, however, responders showed significantly greater
2
GLS than non-responders (9.9±3.8% vs. 7.6±2.6%, P=0.02). LVESV was
3
significantly improved in responders (from 196.5±106.0ml to 134.8±67.4ml, P<0.01),
4
but not in non-responders (from 197.8±80.4ml to 200.7±80.8ml, P=0.74). Similarly,
5
LVEF was significantly improved in responders, but not in non-responders (from
6
26.6±6.7% to 34.6±8.9%, P<0.01, from 25.1±5.3% to 26.6±7.3%, P=0.15,
7
respectively, Supplementary table S1). There were no significant differences in
8
sensed/paced AVD and VVD between responders and non-responders
9
(Supplementary table S2). In addition, patient who needed an excessively short
10
AVD setting due to short intrinsic atrioventricular conduction was not observed.
SC
M AN U
TE D
EP AC C
11
RI PT
1
12
Association of electrocardiographic parameters with response to CRT
13
LV lead was placed at an LAS with mean Q-LV of 128.0±31.7ms, and mean %Q-LV
14
of 78.7±13.2%. There were no significant differences in Q-LV and %Q-LV between
15
responders and non-responders (Table2, Figure 2C). In addition, there was no
ACCEPTED MANUSCRIPT 14
significant correlation between Q-LV and percent reduction in LVESV (Figure 2D).
2
During LV pacing, a mean S-QRS of 39.0±17.5ms was observed. Non-responders
3
showed a significantly longer S-QRS when compared with responders
4
(53.1±17.4ms vs. 30.8±9.5ms, P<0.01, Figure 2A). There was a significant
5
correlation between S-QRS and percent reduction in LVESV (Figure 2B).
6
Non-responders had significantly longer LVp-QRSd than responders, however,
7
actual LVp-QRSd was similar between the two groups. Therefore, the difference in
8
LVp-QRSd arose from a difference in S-QRS. BiVp-QRSd was significantly shorter
9
and Delta-QRSd was significantly greater in responders than in non-responders
SC
M AN U
TE D
EP
11
(Table2).
AC C
10
RI PT
1
12
Long S-QRS interval is associated with non-response to CRT
13
After ROC curve analysis for optimal cutoff value of S-QRS for prediction of CRT
14
response, and 37ms of S-QRS had a sensitivity and specificity of 81% and 90%,
15
respectively, with the area under the curve of 0.91. Patients were then divided into a
ACCEPTED MANUSCRIPT 15
short S-QRS group(<37ms) and a long S-QRS group(≥37ms). The responder rate
2
was significantly higher in the short S-QRS group compared with the long S-QRS
3
group (96% vs. 32%, P<0.01). Both groups had a significant improvement in
4
LVESV and LVEF (Table3), however, percent reduction in LVESV and change in
5
LVEF were significantly greater in short S-QRS group (Figure3). LVp-QRSd was
6
significantly wider in the long S-QRS group, with no significant difference in actual
7
LVp-QRSd compared with short S-QRS group. BiVp-QRSd was significantly
8
narrower, and Delta-QRSd was significantly greater in short S-QRS group than in
9
long S-QRS group (Table3). After optimization of device programing, there were no
TE D
M AN U
SC
RI PT
1
significant differences in AVD and VVD between short S-QRS and long S-QRS
11
(Table3). The significant predictors in the univariable models for response in LVESV
12
were GLS, PQ interval, long S-QRS(≥37ms), LVp-QRSd, BiVp-QRSd, and
13
Delta-QRSd. After multivariate analysis, PQ interval and long S-QRS(≥37ms) were
14
independent predictors for LVESV response (Table4).
15
AC C
EP
10
ACCEPTED MANUSCRIPT 16
Discussion
2
To evaluate the additional impact of S-QRS on CRT response, this study focused
3
on CRT patients with an LV lead placed at LAS. Our major findings are: 1) mapping
4
of LAS enabled successful LV lead placement at a site with Q-LV≥95ms in 86.7% of
5
patients; 2) there was nearly 40% of non-responders in our study population; 3) the
6
optimal cutoff point of S-QRS for non-response prediction was 37ms, and the
7
patients with short S-QRS(<37ms) showed considerably higher response rate at
8
96%; 4) shorter LVp-QRSd was associated with CRT response and mainly
9
contributed by S-QRS; 5) significant shortening in BiVp-QRSd resulting in greater
10
Delta-QRSd was observed in patients with shorter S-QRS; 6) PQ interval and long
11
S-QRS(≥37ms) were independent predictors of CRT response.
SC
M AN U
TE D
EP
AC C
12
RI PT
1
13
Left ventricular pacing at the late activation site and cardiac
14
resynchronization therapy response
ACCEPTED MANUSCRIPT 17
CRT has been shown to promote reverse remodeling, improve clinical outcome by
2
restoring the electrical and mechanical dyssynchrony. LV pacing at the LAS is
3
essential to decrease the non-response rate. However, LV lead placement by
4
conventional anatomical approach is limited for optimal lead placement on the LAS,
5
and a suboptimal LV lead position may contribute to non-responses. Q-LV
6
or %Q-LV has been proposed as a simple intra-procedural evaluation of LAS. Singh
7
et al reported that CRT patients with %Q-LV≥50% showed significant improvement
8
in hemodynamic response and mortality 9. Similarly, Kandala et al reported
9
that %Q-LV≥50% was a predictor of better clinical outcome in CRT patients 8. In
TE D
M AN U
SC
RI PT
1
these studies, however, successful LV lead implantation on the LAS by
11
conventional anatomical approach was only 62% and 76%, respectively.
12
Furthermore, Q-LV≥95ms also showed a remarkable predictive value for improved
13
reverse remodeling and better clinical outcome 7, however, successful LV lead
14
placement on LAS was achieved in less than 50% of patients. In the present study,
15
Q-LV mapping during the procedure enabled a successful LV lead placement on
AC C
EP
10
ACCEPTED MANUSCRIPT 18
LAS in 87% of patients. Despite an optimal LV lead position, Gold et al reported that
2
the responder rate of CRT patients with Q-LV≥95ms was 62.7% 7, which was
3
similarly observed in our study population. Importantly, there were nearly 40%
4
non-responders even in our selected CRT patients, suggesting other reasons for
5
suboptimal LV lead position. Particularly, the LAS may represent a longer S-QRS
6
due to localized scar or fibrotic legion, which should be avoided in a suitable LV
7
pacing site.
M AN U
SC
RI PT
1
9
TE D
8
Long S-QRS interval is associated with non-response to cardiac resynchronization therapy
11
LV lead implantation away from the scar area using cardiac imaging modalities has
12
been shown to benefit CRT response and mortality 10-13. However, an
13
intra-procedural electrocardiographic predictor may be more ideal to locate the scar
14
area. Therefore, this study focused on S-QRS, which may indicate localized tissue
15
property. A longer S-QRS has been frequently observed during pace mapping for
AC C
EP
10
ACCEPTED MANUSCRIPT 19
ventricular tachycardia ablation. Bogun et al reported that pacing at the scar area
2
with delayed potential showed a longer S-QRS with good paced QRS configuration
3
18
4
fractionated potential showed a S-QRS longer than 40ms 19, 20. Therefore, the
5
optimal cutoff point of 37ms for CRT non-response is reasonable to distinguish the
6
scar or fibrotic region. In this study, LVGLS, which is a surrogate marker of scar
7
burden 21, 22, was significantly smaller in the long S-QRS group. Patients with lower
8
GLS may be more likely to have an LV lead on the scar area as a result of larger
9
scar burden. However, GLS, which reflects whole LV scar burden is fundamentally
RI PT
1
TE D
M AN U
SC
. In addition, Stevenson et al reported that pacing at the scar area with
different from S-QRS reflecting local conduction property. Furthermore, the final
11
AVD and VVD were not significantly different between short and long S-QRS group,
12
while the responder rate was significantly higher in short S-QRS group. Therefore,
13
S-QRS rather than Q-LV or GLS is more useful for the evaluation of local
14
conduction property. In addition, Q-LV which is insensitive to tissue impedance
AC C
EP
10
ACCEPTED MANUSCRIPT 20
does not reflect where the activation delay is occurring. However, S-QRS, a
2
measurement during pacing, is clearly affected by localized tissue property.
3
The scar distribution and scar properties may vary between individuals, therefore,
4
S-QRS mapping may help to avoid pacing at scar lesion. Responders showed a
5
shorter S-QRS and a significant correlation was observed between S-QRS and
6
reduction in LVESV, while similar Q-LV and %Q-LV were observed. These findings
7
might suggest that S-QRS mapping should be used instead of longer Q-LV, rather
8
than in addition to longer Q-LV, if a sufficient LAS (QLV≥95ms) was obtained.
9
Moreover, mapping of both shorter S-QRS and longer Q-LV should be feasible,
TE D
M AN U
SC
RI PT
1
however, it is still unknown whether patients exhibiting a long S-QRS have an
11
alternative LV pacing site with short S-QRS in addition to sufficient Q-LV.
12
Electrophysiological study might be useful for optimal patient selection before CRT
13
implantation. Furthermore, high output pacing or multiple sites/points pacing might
14
resolve the influence of long S-QRS as recently reported 23, 24. However, these
15
concepts require further validation.
AC C
EP
10
ACCEPTED MANUSCRIPT 21
1 Impact of S-QRS interval on QRS configurations during LV and biventricular
3
pacing
4
LVp-QRSd has been reported as an independent predictor of CRT response 14. The
5
wider LVp-QRSd is also an indicator of LV pacing proximity to the scar region with
6
conduction delay leading to insufficient resynchronization. In this study,
7
non-responders showed significantly wider LVp-QRSd, longer S-QRS, lower
8
LVGLS, and similar actual LVp-QRSd compared with responders. Thus, actual
9
LVp-QRSd may represent total ventricular conduction rather than local conduction.
TE D
M AN U
SC
RI PT
2
While BiV-QRSd and Delta-QRSd are still controversial predictors for CRT
11
response 15, 16, our study population showed significantly shorter BiV-QRSd and
12
greater Delta-QRSd in responders as well as in short S-QRS group. However, both
13
BiV-QRSd and Delta-QRSd were not significant predictors in multivariate model.
14
Lecoq et al has reported the prognostic value of BiV-QRSd and Delta-QRSd on
15
CRT response, but their definition of CRT responder was NYHA improvement and
AC C
EP
10
ACCEPTED MANUSCRIPT 22
peak VO2>10% increase, which is clinical responder 16. From the perspective of
2
mechanical response to CRT, longer S-QRS may reflect larger scar burden and
3
produce ineffective CRT pacing, therefore, S-QRS could be a stronger predictor
4
than BiV-QRSd or Delta-QRSd. Although BiV-QRSd represents electrical
5
resynchronization after CRT, longer BiV-QRSd may also imply insufficient electrical
6
resynchronization due to longer S-QRS.
M AN U
SC
RI PT
1
7 Limitation
9
This study has several limitations. First, this was a single center retrospective
TE D
8
cohort study and was not sufficiently sized to evaluate clinical outcomes. Second,
11
this study included more patients with non-ischemic etiology and non-LBBB QRS
12
configuration. In addition, patients with relatively narrow QRSd were indicated for
13
CRT according to our country's guideline. Third, LV scar assessment using cardiac
14
imaging modalities was not performed. Therefore, S-QRS does not prove
15
histopathological scar or fibrosis at the LV pacing site. Finally, in this study, most of
AC C
EP
10
ACCEPTED MANUSCRIPT 23
the right ventricular (RV) leads were placed on the RV mid septum, therefore, Q-LV
2
was preferable rather than RV-LV interval 25.
3
RI PT
1
Conclusion
5
S-QRS longer than 37ms at LV pacing site can be a novel independent predictor of
6
CRT non-response. S-QRS also contributes to QRS configuration during LV and
7
biventricular pacing and is associated with CRT response. Both Q-LV and S-QRS
8
guided LV lead placement may benefit CRT outcome. Further studies are warranted
9
to establish our concept.
TE D
M AN U
SC
4
EP
10 Acknowledgment
12
none
13 14 15
AC C
11
ACCEPTED MANUSCRIPT 24
1
References
2
1.
chronic heart failure. N Engl J Med 2002;346:1845-1853. 2.
Bristow MR, Saxon LA, Boehmer J, et al. Cardiac-resynchronization therapy
SC
4
RI PT
3
Abraham WT, Fisher WG, Smith AL, et al. Cardiac resynchronization in
with or without an implantable defibrillator in advanced chronic heart failure.
6
N Engl J Med 2004;350:2140-2150.
7
3.
M AN U
5
Cleland JG, Daubert JC, Erdmann E, Freemantle N, Gras D, Kappenberger L, Tavazzi L, Cardiac Resynchronization-Heart Failure Study I. The effect of
9
cardiac resynchronization on morbidity and mortality in heart failure. N Engl J
TE D
8
4.
12 13
EP
11
Med 2005;352:1539-1549.
Moss AJ, Hall WJ, Cannom DS, et al. Cardiac-resynchronization therapy for
AC C
10
the prevention of heart-failure events. N Engl J Med 2009;361:1329-1338.
5.
Mullens W, Grimm RA, Verga T, Dresing T, Starling RC, Wilkoff BL, Tang WH.
14
Insights from a cardiac resynchronization optimization clinic as part of a heart
15
failure disease management program. J Am Coll Cardiol 2009;53:765-773.
ACCEPTED MANUSCRIPT 25
1
6.
Chatterjee NA, Gold MR, Waggoner AD, Picard MH, Stein KM, Yu Y, Meyer TE, Wold N, Ellenbogen KA, Singh JP. Longer Left Ventricular Electric Delay
3
Reduces Mitral Regurgitation After Cardiac Resynchronization Therapy:
4
Mechanistic Insights From the SMART-AV Study (SmartDelay Determined
5
AV Optimization: A Comparison to Other AV Delay Methods Used in Cardiac
6
Resynchronization Therapy). Circ Arrhythm Electrophysiol
7
2016;9(11):e004346.
SC
M AN U
7.
Gold MR, Birgersdotter-Green U, Singh JP, Ellenbogen KA, Yu Y, Meyer TE,
TE D
8
RI PT
2
Seth M, Tchou PJ. The relationship between ventricular electrical delay and
10
left ventricular remodelling with cardiac resynchronization therapy. Eur Heart
11
J 2011;32:2516-2524.
13
8.
AC C
12
EP
9
Kandala J, Upadhyay GA, Altman RK, Parks KA, Orencole M, Mela T, Kevin
Heist E, Singh JP. QRS morphology, left ventricular lead location, and clinical
14
outcome in patients receiving cardiac resynchronization therapy. Eur Heart J
15
2013;34:2252-2262.
ACCEPTED MANUSCRIPT 26
1
9.
Singh JP, Fan D, Heist EK, Alabiad CR, Taub C, Reddy V, Mansour M, Picard MH, Ruskin JN, Mela T. Left ventricular lead electrical delay predicts
3
response to cardiac resynchronization therapy. Heart Rhythm
4
2006;3:1285-1292.
SC
10.
Adelstein EC, Saba S. Scar burden by myocardial perfusion imaging predicts
M AN U
5
RI PT
2
6
echocardiographic response to cardiac resynchronization therapy in
7
ischemic cardiomyopathy. Am Heart J 2007;153:105-112. 11.
Bakos Z, Markstad H, Ostenfeld E, Carlsson M, Roijer A, Borgquist R.
TE D
8 9
Combined preoperative information using a bullseye plot from speckle tracking echocardiography, cardiac CT scan, and MRI scan: targeted left
11
ventricular lead implantation in patients receiving cardiac resynchronization
12 13
AC C
EP
10
therapy. Eur Heart J Cardiovasc Imaging 2014;15:523-531.
12.
Leyva F, Foley PW, Chalil S, Ratib K, Smith RE, Prinzen F, Auricchio A.
14
Cardiac resynchronization therapy guided by late gadolinium-enhancement
15
cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2011;13:29.
ACCEPTED MANUSCRIPT 27
1
13.
Wong JA, Yee R, Stirrat J, et al. Influence of pacing site characteristics on response to cardiac resynchronization therapy. Circ Cardiovasc Imaging
3
2013;6:542-550. 14.
Hsing JM, Selzman KA, Leclercq C, Pires LA, McLaughlin MG, McRae SE,
SC
4
RI PT
2
Peterson BJ, Zimetbaum PJ. Paced left ventricular QRS width and ECG
6
parameters predict outcomes after cardiac resynchronization therapy:
7
PROSPECT-ECG substudy. Circ Arrhythm Electrophysiol 2011;4:851-857. 15.
Iler MA, Hu T, Ayyagari S, Callahan TDt, Civello KC, Thal SG, Wilkoff BL,
TE D
8
M AN U
5
9
Chung MK. Prognostic value of electrocardiographic measurements before and after cardiac resynchronization device implantation in patients with heart
11
failure due to ischemic or nonischemic cardiomyopathy. Am J Cardiol
12 13
AC C
EP
10
2008;101:359-363.
16.
Lecoq G, Leclercq C, Leray E, Crocq C, Alonso C, de Place C, Mabo P,
14
Daubert C. Clinical and electrocardiographic predictors of a positive
15
response to cardiac resynchronization therapy in advanced heart failure. Eur
ACCEPTED MANUSCRIPT 28
2
Heart J 2005;26:1094-1100. 17.
Ypenburg C, van Bommel RJ, Delgado V, Mollema SA, Bleeker GB,
RI PT
1
Boersma E, Schalij MJ, Bax JJ. Optimal left ventricular lead position predicts
4
reverse remodeling and survival after cardiac resynchronization therapy. J
5
Am Coll Cardiol 2008;52:1402-1409. 18.
M AN U
6
SC
3
Bogun F, Good E, Reich S, Elmouchi D, Igic P, Lemola K, Tschopp D, Jongnarangsin K, Oral H, Chugh A, Pelosi F, Morady F. Isolated potentials
8
during sinus rhythm and pace-mapping within scars as guides for ablation of
9
post-infarction ventricular tachycardia. J Am Coll Cardiol
TE D
7
19.
12
I. Relation of pace mapping QRS configuration and conduction delay to
13
ventricular tachycardia reentry circuits in human infarct scars. Journal of the
14 15
Stevenson WG, Sager PT, Natterson PD, Saxon LA, Middlekauff HR, Wiener
AC C
11
2006;47:2013-2019.
EP
10
American College of Cardiology 1995;26:481-488. 20.
Stevenson WG, Weiss JN, Wiener I, Rivitz SM, Nademanee K, Klitzner T,
ACCEPTED MANUSCRIPT 29
Yeatman L, Josephson M, Wohlgelernter D. Fractionated endocardial
2
electrograms are associated with slow conduction in humans: evidence from
3
pace-mapping. J Am Coll Cardiol 1989;13:369-376. 21.
Delgado-Montero A, Tayal B, Goda A, Ryo K, Marek JJ, Sugahara M, Qi Z,
SC
4
RI PT
1
Althouse AD, Saba S, Schwartzman D, Gorcsan J, 3rd. Additive Prognostic
6
Value of Echocardiographic Global Longitudinal and Global Circumferential
7
Strain to Electrocardiographic Criteria in Patients With Heart Failure
8
Undergoing Cardiac Resynchronization Therapy. Circ Cardiovasc Imaging
9
2016;9(6):e004241.
TH. Echocardiographic predictors of reverse remodeling after cardiac
12
resynchronization therapy and subsequent events. Circ Cardiovasc Imaging
13
15
Park JH, Negishi K, Grimm RA, Popovic Z, Stanton T, Wilkoff BL, Marwick
AC C
11
14
TE D
22.
EP
10
M AN U
5
2013;6:864-872.
23.
Forleo GB, Santini L, Giammaria M, et al. Multipoint pacing via a quadripolar left-ventricular lead: preliminary results from the Italian registry on multipoint
ACCEPTED MANUSCRIPT 30
left-ventricular pacing in cardiac resynchronization therapy (IRON-MPP).
2
Europace 2017;19:1170-1177.
3
24.
RI PT
1
Ishibashi K, Kubo T, Kitabata H, et al. Improvement of cardiac function by increasing stimulus strength during left ventricular pacing in cardiac
5
resynchronization therapy. Int Heart J 2015;56:62-66. 25.
M AN U
6
SC
4
Michael R. Gold JPS, Kenneth A. Ellenbogen, Yinghong Yu, Nicholas Wold, Timothy E. Meyer, Ulrinka Birgersdoter-Green. Interventricular electrical
8
delay is predictive of response to cardiac resynchronization therapy. J Am
9
Coll Cardiol EP 2016;2(4):438-447.
TE D
7
12 13
AC C
11
EP
10
14 15
Figure Legend
ACCEPTED MANUSCRIPT 31
Figure 1. Examples of 12 lead ECG and intracardiac electrograms during sinus
2
rhythm with native QRS (A), during left ventricular (LV) pacing (B), and during
3
biventricular pacing (C). The calipers are aligned with the onset of QRS, off-set of
4
QRS, the peak of the latest LV electrogram, and the pacing stimulation signal. BiV =
5
biventricular; Q = onset of QRS; Q-LV = interval from QRS onset to latest LV
6
electrogram; QRSd = QRS duration; S-QRS = interval from stimulus to QRS onset;
7
STIM = stimulation signal.
TE D
8
M AN U
SC
RI PT
1
Figure 2. Distribution and comparison of S-QRS interval (A) and Q-LV interval (C)
10
in non-responders and responders, and the correlation between percent reduction
11
in LVESV and S-QRS interval (B) and Q-LV interval (D). LVESV indicates left
12
ventricular end systolic volume.
AC C
13
EP
9
14
Figure 3. Comparison of the percent reduction in LVESV (A) and the change in
15
LVEF (B) between the short S-QRS group and the long S-QRS group. LVESV
ACCEPTED MANUSCRIPT 32
indicates left ventricular end systolic volume; LVEF, left ventricular ejection
2
fraction.
RI PT
1
3
SC
4
M AN U
5 6 7
TE D
8 9
EP
10
12 13 14
AC C
11
Figure 1
ACCEPTED MANUSCRIPT
TE D
M AN U
SC
RI PT
33
1 2
EP
3
5 6
AC C
4
7 8
Figure 2.
ACCEPTED MANUSCRIPT 34
B
120
P < 0.01
120
60
40
0
Responder
D
250
250
100
Responder
TE D AC C
4
7 8
1
Figure 3.
0
20
40
60
y = 0.29x + 123.7 R2 = 0.04 P =0.16
200
150
100
50 -40
0
EP
3
6
Q-LV interval (ms)
150
Non-Responder
2
5
M AN U
Q-LV interval (ms)
200
-20
% Reduction in LVESV (%)
P = 0.45
1
40
0 -40
1
Non-Responder
50
60
20
20
C
80
RI PT
80
0
y = -0.42x + 46.9 R2 = 0.23 P < 0.01
100
S-QRS interval (ms)
S-QRS interval (ms)
100
SC
A
-20
0
20
40
% Reduction in LVESV (%)
60
ACCEPTED MANUSCRIPT
3 4 5 6 7
EP
2
AC C
1
TE D
M AN U
SC
RI PT
35
ACCEPTED MANUSCRIPT 36
Table 1. Baseline clinical characteristics in total population and those with
2
and without response to cardiac resynchronization therapy Total (N=52)
Responder
62.6 ± 13.9
Male gender (%)
41 (79)
Ischemic (%) Non-Ischemic (%)
62.6 ± 13.7
0.78
24 (75)
17 (85)
0.5
13 (25)
9 (28)
4 (20)
0.74
39 (75)
23 (72)
16 (80) 0.59
0 (0)
0 (0)
0 (0)
34 (65)
21 (66)
13 (65)
(%)
10 (19)
5 (16)
5 (25)
(%)
8 (15)
6 (19)
2 (10)
159.5±34.0
161.4±28.4
157.0±32.6
EP
(%)
62.6 ± 14.7
TE D
NYHA
AC C
(%)
QRS duration (ms)
p value
nder (N=20)
M AN U
Age (years)
Non-Respo
SC
(N=32)
RI PT
1
0.44
ACCEPTED MANUSCRIPT 37
QRS block type
0.67 4 (13)
LBBB (%)
36 (69)
21 (66)
IVCD (%)
9 (17)
7 (22)
3 (15)
RI PT
7 (13)
15 (75) 2 (10)
SC
RBBB (%)
M AN U
Device type 8 (15)
6 (19)
2 (10)
CRT-D (%)
44 (85)
26 (81)
18 (90)
Primary prevention (%)
35 (80)
24 (92)
11 (61)
2 (8)
7 (39)
TE D
CRT-P (%)
Secondary prevention (%)
9 (20)
0.46
0.04
Values are expressed as mean ± SD or number (%). CRT indicates cardiac
2
resynchronization therapy; IVCD, intraventricular conduction delay; LBBB, left
3
bundle branch block; NYHA, New York heart association; RBBB, right bundle
4
branch block.
AC C
EP
1
ACCEPTED MANUSCRIPT 38
Table 2. Comparison of electrocardiographic assessment during the
2
procedure in responders vs. non-responders
RI PT
1
Responder (N=32)
Non-Responder (N=20)
p value
Native QRSd (ms)
159.5 ± 34.0
161.4 ± 28.4
157.0 ± 32.6
0.44
PQ interval (ms)
223.2 ± 57.1
209.2 ± 42.5
Q-LV interval (ms)
128.0 ± 31.7
%Q-LV (%)
78.7 ± 13.2
LV pacing QRSd (ms)
245.5 ± 46.9
S-QRS interval (ms)
39.0 ± 17.5
Actual LV pacing QRSd (ms)
206.6± 39.4
3 4
< 0.01
130.8 ± 28.0
126.1 ± 29.6
0.46
81.1 ± 9.9
80.5 ± 9.2
0.76
238.5 ± 29.1
264.0 ± 40.5
0.016
30.8 ± 9.5
53.1 ± 17.4
< 0.01
207.7 ± 28.6
210.9 ± 33.6
0.41
146.3 ± 18.5
165.6 ± 21.8
< 0.01
−15.2 ± 26.0
+8.6 ± 30.0
< 0.01
M AN U
250.4± 55.4
TE D
EP
Delta-QRS (ms)
152.1 ± 28.6
AC C
Biventricular pacing QRSd (ms)
SC
Total (N=52)
−6.7 ± 29.9
Values are expressed as mean±SD. QRSd indicates QRS duration
ACCEPTED MANUSCRIPT 39
Table 3. Comparison of clinical characteristics between short S-QRS group
2
(<37ms) and long S-QRS group (≥37ms)
RI PT
1
Long S-QRS (≥37ms) (N=28)
p value
64.5 ± 13.3
60.9 ± 14.4
0.51
Age (years)
17 (71)
24 (86)
0.31
Ischemic etiology (%)
9 (37)
4 (14)
0.11
174.3 ± 80.0
216.5 ± 105.5
0.06
27.8 ± 6.3
24.5 ± 5.8
0.03
191.4 ± 79.9
< 0.01
35.4 ± 9.6
28.2 ± 7.4
< 0.01
% reduction in LVESV (%)
13.3 ± 2.7
-8.5 ± 18.3
< 0.01
Radial Dyssynchrony (ms)
215.2 ± 134.6
195.4 ± 107.9
0.67
10.5 ± 3.6
7.7 ± 3.0
< 0.01
163.1 ± 26.6
156.9 ± 32.6
0.3
M AN U
Male Gender (%)
SC
Short S-QRS (<37ms) (N=24)
LVESV at baseline (mL)
(mL)
AC C
LVEF after six months (%)
123.6 ± 61.3
EP
LVESV after six months
TE D
LVEF at baseline (%)
LV GLS (%) Native QRSd (ms)
ACCEPTED MANUSCRIPT 40
S-QRS interval (ms)
132.8 ± 28.7
125.6 ± 28.3
0.33
26.2 ± 5.5
50.6 ± 15.1
< 0.01
259.8 ± 35.4
< 0.01
209.1 ± 29.7
0.49
234.9 ± 31.9
Actual LV pacing QRSd
208.8 ± 31.8
SC
LV pacing QRSd (ms)
RI PT
Q-LV interval (ms)
M AN U
(ms) Biventricular pacing QRSd
145.0 ± 18.0
161.1 ± 22.3
< 0.01
-18.0 ± 26.7
+4.2 ± 28.6
< 0.01
114.6 ± 25.4
122.9 ± 31.3
0.5
153.3 ± 29.7
162.5 ± 35.0
0.55
-12.3 ± 17.3
-11.8 ± 18.1
0.83
23 (96)
9 (32)
< 0.01
(ms)
Sensed AVD (ms)
EP
Paced AVD (ms)
TE D
Delta-QRS (ms)
AC C
VVD (ms) Responder rate (%) 1
Values are mean±SD or number (%). AVD indicates atrioventricular delay; LVEF,
2
left ventricular ejection fraction; LVESV, left ventricular end systolic volume; LVGLS,
ACCEPTED MANUSCRIPT 41
left ventricular global longitudinal strain; QRSd, QRS duration, VVD; interventricular
2
delay. P<0.01 when compared with baseline.
AC C
EP
TE D
M AN U
SC
RI PT
1
ACCEPTED MANUSCRIPT 42
Table 4. Multivariate analysis for prediction of response to cardiac
2
resynchronization therapy Univariate analysis
Multivariate analysis
95% CI
p value
LVEDV (mL)
1
0.99 - 1.01
0.91
LVESV (mL)
1
0.99 - 1.01
0.96
LVEF (%)
1.04
0.95 - 1.14
0.41
Radial Dyssynchrony
1
1.00 - 1.01
0.24
Odds ratio
95% CI
p value
1.14
0.85 - 1.51
0.39
LV GLS (%)
1.24
Native QRSd (ms)
EP
TE D
M AN U
SC
Odds ratio
(ms)
RI PT
1
AC C
1.01
1.04 - 1.52
0.023
0.99 - 1.03
0.6
PQ interval (ms)
0.98
0.96 - 0.99
0.012
0.97
0.94 - 1.0
0.01
Long S-QRS interval
0.02
0.002 - 0.18
< 0.01
0.014
0.001 - 0.26
< 0.01
0.98
0.96 - 1.0
0.02
(≥37ms) LV pacing QRSd (ms)
ACCEPTED MANUSCRIPT
Actual LV pacing QRSd
1
0.98 - 1.02
0.71
0.95
0.92 - 0.98
< 0.01
0.97
0.94 - 0.99
< 0.01
(ms) Biventricular pacing
M AN U
Delta-QRS (ms)
0.97
SC
QRSd (ms)
RI PT
43
0.99
0.92 - 1.02
0.17
0.95 - 1.02
0.44
1
CI indicates confidence interval. Other abbreviations are same as those in Tables 2
2
and 3.
6 7 8 9 10
EP
5
AC C
4
TE D
3
S1547-5271(18)30906-8
DOI:
10.1016/j.hrthm.2018.08.035
Reference:
HRTHM 7730
To appear in:
Heart Rhythm
Received Date: 1 June 2018
Please cite this article as: Yagishita D, Shoda M, Yagishita Y, Ejima K, Hagiwara N, Time interval from left ventricular stimulation to QRS onset is a novel predictor of non-response to cardiac resynchronization therapy, Heart Rhythm (2018), doi: 10.1016/j.hrthm.2018.08.035. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Native QRS
A
100ms
Ⅰ
Left Ventricular ACCEPTED MANUSCRIPT
B
pacing
C
100ms
Ⅰ
Biventricular pacing Ⅰ
Ⅱ
Ⅱ
Ⅲ
Ⅲ
Ⅲ
aVR
aVR
aVL
aVL
aVF
aVF
V1
V1
SC M AN U
V2
Q
V3
V4
V4
V5
V5
V6
V6
RA
RA
EP
V3
TE D
V2
RI PT
Ⅱ
Q
LV2-3 LV3-4 LV Q-LV QRSd
aVL aVF V1 V2 V3 V4 V5 V6
RV
RV
LV1-2
aVR
RA
AC C
RV
100ms
LV1-2
LV1-2
LV2-3
LV2-3
LV3-4
LV3-4 STIM
S-QRS LV pacing QRSd
Actual LV pacing QRSd
STIM
BiV pacing QRSd
ACCEPTED MANUSCRIPT 1
Time interval from left ventricular stimulation to QRS onset is a novel
2
predictor of non-response to cardiac resynchronization therapy
RI PT
1
3
Daigo Yagishita MD.*, Morio Shoda MD. PhD.*†, Yoshimi Yagishita MD. PhD.*,
5
Koichiro Ejima MD. PhD.*†, Nobuhisa Hagiwara MD.PhD.*
M AN U
SC
4
6 7
Total word count: 4999
10
Brief title: Impact of left ventricular stimulation to QRS interval on CRT
EP
9
TE D
8
* Department of Cardiology, Tokyo Women’s Medical University
12
† Clinical Research Division for Heart Rhythm Management, Tokyo Women’s
13
Medical University
AC C
11
14 15
The authors do not have grants, conflicts, or relationships with industry.
ACCEPTED MANUSCRIPT
Correspondence:
2
Morio Shoda M.D., Ph.D.
3
8-1 Kawada-cho, Shinjuku-ku, Tokyo, Japan
4
Tel: +81.3.3353.8111
5
Fax: +81.3.3356.0441
6
Email: [email protected]
7
11 12 13 14 15
EP
10
AC C
9
TE D
8
M AN U
SC
1
RI PT
2
ACCEPTED MANUSCRIPT 3
Abstract
2
Background: A left ventricular (LV) lead placement at the late activation site (LAS)
3
has been proposed as an optimal LV pacing site (i.e. Q-LV interval). However, LAS
4
may be relevant to local electrical conduction, measured as an interval from LV
5
pacing stimulation to QRS onset (S-QRS interval).
6
Objectives: This study sought to evaluate the prognostic value of S-QRS for reverse
7
remodeling, and the impact of S-QRS on pacing QRS configuration in patients
8
undergoing cardiac resynchronization therapy (CRT).
9
Methods: Sixty consecutive heart failure patients with a wide QRS complex
TE D
M AN U
SC
RI PT
1
underwent CRT. A site with Q-LV≥95ms was targeted for LV lead placement. A
11
responder was defined as one with greater than 15% reduction in LV end systolic
12
volume at six months after CRT.
13
Results: An LV lead placement with Q-LV≥95ms was achieved in 52 of 60 patients
14
(86.7%). Thirty-two of 52 patients were responders (61.5%). S-QRS was
15
significantly shorter in responders than non-responders (P<0.01), whereas Q-LV
AC C
EP
10
ACCEPTED MANUSCRIPT 4
was not significantly different. A cutoff value of 37ms for S-QRS had a sensitivity
2
and specificity of 81% and 90%, respectively. Shorter S-QRS (<37ms) showed
3
significantly narrower LV pacing QRS width and biventricular pacing QRS width
4
compared to longer S-QRS. After multivariate analysis, PQ interval (OR=0.97,
5
P=0.01) and long S-QRS greater than 37ms (OR=0.014, P<0.01) were independent
6
predictors for response to CRT.
7
Conclusion: In addition to a sufficient Q-LV, S-QRS can be a useful indicator of
8
optimal LV lead position to achieve reverse remodeling. S-QRS contributes to the
9
pacing QRS configuration associated with CRT response.
SC
M AN U
TE D
EP
10
RI PT
1
Key words: Cardiac resynchronization therapy, Heart failure, Electrocardiograms,
12
Left ventricular lead, Prognosis, Scar
13 14 15
AC C
11
ACCEPTED MANUSCRIPT 5
Introduction
2
Cardiac resynchronization therapy (CRT) is an established treatment for heart
3
failure patients due to left ventricular (LV) dysfunction and QRS prolongation
4
However, some candidates fail to achieve clinical or echocardiographic benefit from
5
CRT, which has been attributed in part to a suboptimal LV lead position 5.
6
LV lead placement at the latest electrical activation site has been proposed as an
7
important factor for a superior CRT response compared with conventional
8
anatomical approach. The time interval from QRS onset to the local LV activation at
9
the LV lead (Q-LV) is a common indicator determining optimal LV pacing site, and is
10
associated with acute hemodynamic improvement, LV reverse remodeling and
11
clinical outcome 6-9.
12
LV late activation site (LAS) may be due to scarring or fibrosis within the viable
13
myocardium where an anatomical or functional conduction block can occur.
14
Therefore, LV pacing within those areas may be less effective for CRT 10-13.
15
Local conduction disturbance during LV pacing (i.e. stimulus to QRS interval,
1-4
.
AC C
EP
TE D
M AN U
SC
RI PT
1
ACCEPTED MANUSCRIPT 6
S-QRS) may indicate that the LV lead has been placed on a scar or fibrotic region.
2
Furthermore, S-QRS may influence the QRS configuration during LV pacing or
3
biventricular pacing, which are predictors for CRT response 14-16.
4
We therefore hypothesized that long S-QRS would also predict CRT non-response.
5
The present study aimed 1) to evaluate the prognostic value of S-QRS on LV
6
reverse remodeling; and 2) to elucidate the relevance of S-QRS on the QRS
7
configurations during LV and biventricular pacing in CRT patients with optimal LV
8
lead position at LAS.
TE D
M AN U
SC
RI PT
1
9 Methods
11
Ethics Statement
12
This study was a retrospective study approved by the Ethics Committee of the
13
Tokyo Women’s Medical University and conducted in accordance with the
14
Declaration of Helsinki. Written informed consent was given by all patients or their
15
guardians.
AC C
EP
10
ACCEPTED MANUSCRIPT 7
1 Study Population
3
All heart failure patients who underwent CRT implantation between January 2012
4
and February 2016 at our hospital were reviewed for study eligibility. Inclusion
5
criteria were based on our country’s guideline, specifically for sinus rhythm, New
6
York Heart Association functional class II to IV heart failure on optimal
7
pharmacological therapy, QRS duration (QRSd)≥120ms, and left ventricular
8
ejection fraction (LVEF)≤35%. Exclusion criteria were age≤18 years, chronic atrial
9
tachycardia/fibrillation,
TE D
M AN U
SC
RI PT
2
and
complete
atrioventricular
block.
Patients
with
Q-LV<95ms at the LV pacing site were excluded because recent reports indicate
11
this was an inadequate LV pacing site 6, 7.
AC C
12
EP
10
13
CRT Implantation Procedure
14
CRT devices were implanted in the standard fashion, and LV leads were
15
transvenously implanted in a coronary vein branch. During the procedure,
ACCEPTED MANUSCRIPT 8
simultaneous recording of 12 lead ECG and intracardiac electrograms were
2
continuously obtained using a digital recording system (Pruka CardioLab System,
3
GE Healthcare, Waukesha, WI). The recorded intracardiac electrograms were
4
bandpass filtered between 0.5Hz and 500Hz. Q-LV mapping was performed with a
5
decapolar catheter or LV lead electrodes, and the latest LV activation site was
6
targeted as the optimal pacing site in all patients.
M AN U
SC
RI PT
1
7
Electrocardiographic measurements
9
The electrocardiographic measurements were retrospectively performed by two
10
independent reviewers blinded to the outcome. Q-LV was defined as the time
11
interval from the earliest QRS onset to the first major peak of the local bipolar
12
electrogram. The %Q-LV was calculated as the percentage of Q-LV in native QRSd.
13
S-QRS was defined as the time from the LV pacing to the earliest onset of QRS. LV
14
pacing QRSd (LVp-QRSd) was measured from the pacing to the offset of QRS,
15
while the subtraction of S-QRS from LVp-QRSd was termed as the actual LV pacing
AC C
EP
TE D
8
ACCEPTED MANUSCRIPT 9
QRSd (actual LVp-QRSd). Similarly, simultaneous biventricular pacing QRSd
2
(BiV-QRSd) was measured from the pacing to the offset of QRS (Figure1). Finally,
3
the difference between native QRSd and BiV-QRSd (Delta-QRSd) was calculated.
4
These parameters were measured with a caliper in the recording system at a sweep
5
speed of 200mm/s. LV pacing was performed with output at pacing threshold (1.1±
6
0.6V at 0.4ms), and pacing rate was programmed at 10 beats per minute faster than
7
the sinus rhythm.
M AN U
SC
RI PT
1
9
TE D
8 Echocardiographic assessment All
patients
had
a
comprehensive
echocardiographic
assessment
using
11
commercially available echocardiographic systems and data at baseline and six
12
months after CRT implantation were stored digitally. Standard LV volumetric
13
measurements were performed using the recommended Simpson’s biplane method.
14
Intraventricular dyssynchrony was assessed by radial speckle tracking analysis of
15
the anteroseptal and posterolateral mid LV segments
AC C
EP
10
17
. A CRT responder was
ACCEPTED MANUSCRIPT 10
defined as one who had a greater than 15% reduction in LV end systolic volume
2
(LVESV) at six months compared with baseline.
3
RI PT
1
Optimization of CRT device programing
5
VVD optimization was performed to minimize BiV-QRSd during procedure,
6
thereafter, echocardiography was performed to maximize pulsed wave LV outflow
7
tract velocity time integral. AVD optimization was performed by Doppler
8
echocardiography to provide the maximum left ventricular filling time without
9
interference of atrial wave and diastolic mitral regurgitation. Following to AVD
M AN U
TE D
EP
11
optimization, BiV-QRSd and QRS morphology were confirmed again.
AC C
10
SC
4
12
Statistical Analysis
13
All statistical analyses were performed using JMP version 13.0 (SAS Institute, Cary,
14
NC, USA). Continuous variables are expressed as the mean±SD, while categorical
15
variables are presented as frequency and percentage. A linear regression analysis
ACCEPTED MANUSCRIPT 11
was applied to study the correlation between percent reduction in LVESV and
2
S-QRS or Q-LV. For the univariate analysis, Fisher’s exact test and Wilcoxon exact
3
test was used. Receiver operating characteristic (ROC) curve analysis was
4
performed to determine the optimal cut-off value of the S-QRS for predicting CRT
5
non-responders. A multivariate logistic regression model was constructed to assess
6
the association of clinical variables to predict the response to CRT. P < 0.05 was
7
used to select variables from the univariate analysis. All statistical tests were two
8
tailed, and P value < 0.05 was considered as significant.
9
TE D
M AN U
SC
RI PT
1
Results
11
Baseline patient characteristics
12
Sixty consecutive CRT patients were enrolled in the present study. Of these, eight
13
patients were excluded due to suboptimal LV lead position without sufficient
14
electrical latency. The final study population included 52 patients (79% male) with a
15
mean age of 62.6±13.9 years (Table 1). Mean native QRSd was 159.5±34.0ms,
AC C
EP
10
ACCEPTED MANUSCRIPT 12
and mean LVEF was 26.0±6.2%. In the QRS morphological features, 36 patients
2
(69%) had LBBB, while 16 patients (31%) had non-LBBB. Non-ischemic etiology
3
and ischemic disease was present in 75% and 25% of patients, respectively. There
4
were 18 patients classified as NYHA
5
optimal medical therapy before CRT implantation. According to LV volumetric
6
evaluation results before and six months after CRT implantation, 32 patients
7
(61.5%) were responders. Patients were divided into a responder and a
8
non-responder group to compare their characteristics. Baseline characteristics were
9
not significantly different except defibrillator device indication.
TE D
M AN U
SC
, and all patients were under the
EP
10
and
RI PT
1
Echocardiographic parameters before and six months after CRT implantation
12
At baseline, patients had severely reduced LV systolic function with enlarged LV
13
with mean LVESV at 197.0±96.1ml. Severe radial dyssynchrony was present at
14
204.5±120.1ms, and mean GLS was reduced at 9.0±3.5%. There were no
15
significant differences in LV volumetric evaluation and radial dyssynchrony between
AC C
11
ACCEPTED MANUSCRIPT 13
responders and non-responders, however, responders showed significantly greater
2
GLS than non-responders (9.9±3.8% vs. 7.6±2.6%, P=0.02). LVESV was
3
significantly improved in responders (from 196.5±106.0ml to 134.8±67.4ml, P<0.01),
4
but not in non-responders (from 197.8±80.4ml to 200.7±80.8ml, P=0.74). Similarly,
5
LVEF was significantly improved in responders, but not in non-responders (from
6
26.6±6.7% to 34.6±8.9%, P<0.01, from 25.1±5.3% to 26.6±7.3%, P=0.15,
7
respectively, Supplementary table S1). There were no significant differences in
8
sensed/paced AVD and VVD between responders and non-responders
9
(Supplementary table S2). In addition, patient who needed an excessively short
10
AVD setting due to short intrinsic atrioventricular conduction was not observed.
SC
M AN U
TE D
EP AC C
11
RI PT
1
12
Association of electrocardiographic parameters with response to CRT
13
LV lead was placed at an LAS with mean Q-LV of 128.0±31.7ms, and mean %Q-LV
14
of 78.7±13.2%. There were no significant differences in Q-LV and %Q-LV between
15
responders and non-responders (Table2, Figure 2C). In addition, there was no
ACCEPTED MANUSCRIPT 14
significant correlation between Q-LV and percent reduction in LVESV (Figure 2D).
2
During LV pacing, a mean S-QRS of 39.0±17.5ms was observed. Non-responders
3
showed a significantly longer S-QRS when compared with responders
4
(53.1±17.4ms vs. 30.8±9.5ms, P<0.01, Figure 2A). There was a significant
5
correlation between S-QRS and percent reduction in LVESV (Figure 2B).
6
Non-responders had significantly longer LVp-QRSd than responders, however,
7
actual LVp-QRSd was similar between the two groups. Therefore, the difference in
8
LVp-QRSd arose from a difference in S-QRS. BiVp-QRSd was significantly shorter
9
and Delta-QRSd was significantly greater in responders than in non-responders
SC
M AN U
TE D
EP
11
(Table2).
AC C
10
RI PT
1
12
Long S-QRS interval is associated with non-response to CRT
13
After ROC curve analysis for optimal cutoff value of S-QRS for prediction of CRT
14
response, and 37ms of S-QRS had a sensitivity and specificity of 81% and 90%,
15
respectively, with the area under the curve of 0.91. Patients were then divided into a
ACCEPTED MANUSCRIPT 15
short S-QRS group(<37ms) and a long S-QRS group(≥37ms). The responder rate
2
was significantly higher in the short S-QRS group compared with the long S-QRS
3
group (96% vs. 32%, P<0.01). Both groups had a significant improvement in
4
LVESV and LVEF (Table3), however, percent reduction in LVESV and change in
5
LVEF were significantly greater in short S-QRS group (Figure3). LVp-QRSd was
6
significantly wider in the long S-QRS group, with no significant difference in actual
7
LVp-QRSd compared with short S-QRS group. BiVp-QRSd was significantly
8
narrower, and Delta-QRSd was significantly greater in short S-QRS group than in
9
long S-QRS group (Table3). After optimization of device programing, there were no
TE D
M AN U
SC
RI PT
1
significant differences in AVD and VVD between short S-QRS and long S-QRS
11
(Table3). The significant predictors in the univariable models for response in LVESV
12
were GLS, PQ interval, long S-QRS(≥37ms), LVp-QRSd, BiVp-QRSd, and
13
Delta-QRSd. After multivariate analysis, PQ interval and long S-QRS(≥37ms) were
14
independent predictors for LVESV response (Table4).
15
AC C
EP
10
ACCEPTED MANUSCRIPT 16
Discussion
2
To evaluate the additional impact of S-QRS on CRT response, this study focused
3
on CRT patients with an LV lead placed at LAS. Our major findings are: 1) mapping
4
of LAS enabled successful LV lead placement at a site with Q-LV≥95ms in 86.7% of
5
patients; 2) there was nearly 40% of non-responders in our study population; 3) the
6
optimal cutoff point of S-QRS for non-response prediction was 37ms, and the
7
patients with short S-QRS(<37ms) showed considerably higher response rate at
8
96%; 4) shorter LVp-QRSd was associated with CRT response and mainly
9
contributed by S-QRS; 5) significant shortening in BiVp-QRSd resulting in greater
10
Delta-QRSd was observed in patients with shorter S-QRS; 6) PQ interval and long
11
S-QRS(≥37ms) were independent predictors of CRT response.
SC
M AN U
TE D
EP
AC C
12
RI PT
1
13
Left ventricular pacing at the late activation site and cardiac
14
resynchronization therapy response
ACCEPTED MANUSCRIPT 17
CRT has been shown to promote reverse remodeling, improve clinical outcome by
2
restoring the electrical and mechanical dyssynchrony. LV pacing at the LAS is
3
essential to decrease the non-response rate. However, LV lead placement by
4
conventional anatomical approach is limited for optimal lead placement on the LAS,
5
and a suboptimal LV lead position may contribute to non-responses. Q-LV
6
or %Q-LV has been proposed as a simple intra-procedural evaluation of LAS. Singh
7
et al reported that CRT patients with %Q-LV≥50% showed significant improvement
8
in hemodynamic response and mortality 9. Similarly, Kandala et al reported
9
that %Q-LV≥50% was a predictor of better clinical outcome in CRT patients 8. In
TE D
M AN U
SC
RI PT
1
these studies, however, successful LV lead implantation on the LAS by
11
conventional anatomical approach was only 62% and 76%, respectively.
12
Furthermore, Q-LV≥95ms also showed a remarkable predictive value for improved
13
reverse remodeling and better clinical outcome 7, however, successful LV lead
14
placement on LAS was achieved in less than 50% of patients. In the present study,
15
Q-LV mapping during the procedure enabled a successful LV lead placement on
AC C
EP
10
ACCEPTED MANUSCRIPT 18
LAS in 87% of patients. Despite an optimal LV lead position, Gold et al reported that
2
the responder rate of CRT patients with Q-LV≥95ms was 62.7% 7, which was
3
similarly observed in our study population. Importantly, there were nearly 40%
4
non-responders even in our selected CRT patients, suggesting other reasons for
5
suboptimal LV lead position. Particularly, the LAS may represent a longer S-QRS
6
due to localized scar or fibrotic legion, which should be avoided in a suitable LV
7
pacing site.
M AN U
SC
RI PT
1
9
TE D
8
Long S-QRS interval is associated with non-response to cardiac resynchronization therapy
11
LV lead implantation away from the scar area using cardiac imaging modalities has
12
been shown to benefit CRT response and mortality 10-13. However, an
13
intra-procedural electrocardiographic predictor may be more ideal to locate the scar
14
area. Therefore, this study focused on S-QRS, which may indicate localized tissue
15
property. A longer S-QRS has been frequently observed during pace mapping for
AC C
EP
10
ACCEPTED MANUSCRIPT 19
ventricular tachycardia ablation. Bogun et al reported that pacing at the scar area
2
with delayed potential showed a longer S-QRS with good paced QRS configuration
3
18
4
fractionated potential showed a S-QRS longer than 40ms 19, 20. Therefore, the
5
optimal cutoff point of 37ms for CRT non-response is reasonable to distinguish the
6
scar or fibrotic region. In this study, LVGLS, which is a surrogate marker of scar
7
burden 21, 22, was significantly smaller in the long S-QRS group. Patients with lower
8
GLS may be more likely to have an LV lead on the scar area as a result of larger
9
scar burden. However, GLS, which reflects whole LV scar burden is fundamentally
RI PT
1
TE D
M AN U
SC
. In addition, Stevenson et al reported that pacing at the scar area with
different from S-QRS reflecting local conduction property. Furthermore, the final
11
AVD and VVD were not significantly different between short and long S-QRS group,
12
while the responder rate was significantly higher in short S-QRS group. Therefore,
13
S-QRS rather than Q-LV or GLS is more useful for the evaluation of local
14
conduction property. In addition, Q-LV which is insensitive to tissue impedance
AC C
EP
10
ACCEPTED MANUSCRIPT 20
does not reflect where the activation delay is occurring. However, S-QRS, a
2
measurement during pacing, is clearly affected by localized tissue property.
3
The scar distribution and scar properties may vary between individuals, therefore,
4
S-QRS mapping may help to avoid pacing at scar lesion. Responders showed a
5
shorter S-QRS and a significant correlation was observed between S-QRS and
6
reduction in LVESV, while similar Q-LV and %Q-LV were observed. These findings
7
might suggest that S-QRS mapping should be used instead of longer Q-LV, rather
8
than in addition to longer Q-LV, if a sufficient LAS (QLV≥95ms) was obtained.
9
Moreover, mapping of both shorter S-QRS and longer Q-LV should be feasible,
TE D
M AN U
SC
RI PT
1
however, it is still unknown whether patients exhibiting a long S-QRS have an
11
alternative LV pacing site with short S-QRS in addition to sufficient Q-LV.
12
Electrophysiological study might be useful for optimal patient selection before CRT
13
implantation. Furthermore, high output pacing or multiple sites/points pacing might
14
resolve the influence of long S-QRS as recently reported 23, 24. However, these
15
concepts require further validation.
AC C
EP
10
ACCEPTED MANUSCRIPT 21
1 Impact of S-QRS interval on QRS configurations during LV and biventricular
3
pacing
4
LVp-QRSd has been reported as an independent predictor of CRT response 14. The
5
wider LVp-QRSd is also an indicator of LV pacing proximity to the scar region with
6
conduction delay leading to insufficient resynchronization. In this study,
7
non-responders showed significantly wider LVp-QRSd, longer S-QRS, lower
8
LVGLS, and similar actual LVp-QRSd compared with responders. Thus, actual
9
LVp-QRSd may represent total ventricular conduction rather than local conduction.
TE D
M AN U
SC
RI PT
2
While BiV-QRSd and Delta-QRSd are still controversial predictors for CRT
11
response 15, 16, our study population showed significantly shorter BiV-QRSd and
12
greater Delta-QRSd in responders as well as in short S-QRS group. However, both
13
BiV-QRSd and Delta-QRSd were not significant predictors in multivariate model.
14
Lecoq et al has reported the prognostic value of BiV-QRSd and Delta-QRSd on
15
CRT response, but their definition of CRT responder was NYHA improvement and
AC C
EP
10
ACCEPTED MANUSCRIPT 22
peak VO2>10% increase, which is clinical responder 16. From the perspective of
2
mechanical response to CRT, longer S-QRS may reflect larger scar burden and
3
produce ineffective CRT pacing, therefore, S-QRS could be a stronger predictor
4
than BiV-QRSd or Delta-QRSd. Although BiV-QRSd represents electrical
5
resynchronization after CRT, longer BiV-QRSd may also imply insufficient electrical
6
resynchronization due to longer S-QRS.
M AN U
SC
RI PT
1
7 Limitation
9
This study has several limitations. First, this was a single center retrospective
TE D
8
cohort study and was not sufficiently sized to evaluate clinical outcomes. Second,
11
this study included more patients with non-ischemic etiology and non-LBBB QRS
12
configuration. In addition, patients with relatively narrow QRSd were indicated for
13
CRT according to our country's guideline. Third, LV scar assessment using cardiac
14
imaging modalities was not performed. Therefore, S-QRS does not prove
15
histopathological scar or fibrosis at the LV pacing site. Finally, in this study, most of
AC C
EP
10
ACCEPTED MANUSCRIPT 23
the right ventricular (RV) leads were placed on the RV mid septum, therefore, Q-LV
2
was preferable rather than RV-LV interval 25.
3
RI PT
1
Conclusion
5
S-QRS longer than 37ms at LV pacing site can be a novel independent predictor of
6
CRT non-response. S-QRS also contributes to QRS configuration during LV and
7
biventricular pacing and is associated with CRT response. Both Q-LV and S-QRS
8
guided LV lead placement may benefit CRT outcome. Further studies are warranted
9
to establish our concept.
TE D
M AN U
SC
4
EP
10 Acknowledgment
12
none
13 14 15
AC C
11
ACCEPTED MANUSCRIPT 24
1
References
2
1.
chronic heart failure. N Engl J Med 2002;346:1845-1853. 2.
Bristow MR, Saxon LA, Boehmer J, et al. Cardiac-resynchronization therapy
SC
4
RI PT
3
Abraham WT, Fisher WG, Smith AL, et al. Cardiac resynchronization in
with or without an implantable defibrillator in advanced chronic heart failure.
6
N Engl J Med 2004;350:2140-2150.
7
3.
M AN U
5
Cleland JG, Daubert JC, Erdmann E, Freemantle N, Gras D, Kappenberger L, Tavazzi L, Cardiac Resynchronization-Heart Failure Study I. The effect of
9
cardiac resynchronization on morbidity and mortality in heart failure. N Engl J
TE D
8
4.
12 13
EP
11
Med 2005;352:1539-1549.
Moss AJ, Hall WJ, Cannom DS, et al. Cardiac-resynchronization therapy for
AC C
10
the prevention of heart-failure events. N Engl J Med 2009;361:1329-1338.
5.
Mullens W, Grimm RA, Verga T, Dresing T, Starling RC, Wilkoff BL, Tang WH.
14
Insights from a cardiac resynchronization optimization clinic as part of a heart
15
failure disease management program. J Am Coll Cardiol 2009;53:765-773.
ACCEPTED MANUSCRIPT 25
1
6.
Chatterjee NA, Gold MR, Waggoner AD, Picard MH, Stein KM, Yu Y, Meyer TE, Wold N, Ellenbogen KA, Singh JP. Longer Left Ventricular Electric Delay
3
Reduces Mitral Regurgitation After Cardiac Resynchronization Therapy:
4
Mechanistic Insights From the SMART-AV Study (SmartDelay Determined
5
AV Optimization: A Comparison to Other AV Delay Methods Used in Cardiac
6
Resynchronization Therapy). Circ Arrhythm Electrophysiol
7
2016;9(11):e004346.
SC
M AN U
7.
Gold MR, Birgersdotter-Green U, Singh JP, Ellenbogen KA, Yu Y, Meyer TE,
TE D
8
RI PT
2
Seth M, Tchou PJ. The relationship between ventricular electrical delay and
10
left ventricular remodelling with cardiac resynchronization therapy. Eur Heart
11
J 2011;32:2516-2524.
13
8.
AC C
12
EP
9
Kandala J, Upadhyay GA, Altman RK, Parks KA, Orencole M, Mela T, Kevin
Heist E, Singh JP. QRS morphology, left ventricular lead location, and clinical
14
outcome in patients receiving cardiac resynchronization therapy. Eur Heart J
15
2013;34:2252-2262.
ACCEPTED MANUSCRIPT 26
1
9.
Singh JP, Fan D, Heist EK, Alabiad CR, Taub C, Reddy V, Mansour M, Picard MH, Ruskin JN, Mela T. Left ventricular lead electrical delay predicts
3
response to cardiac resynchronization therapy. Heart Rhythm
4
2006;3:1285-1292.
SC
10.
Adelstein EC, Saba S. Scar burden by myocardial perfusion imaging predicts
M AN U
5
RI PT
2
6
echocardiographic response to cardiac resynchronization therapy in
7
ischemic cardiomyopathy. Am Heart J 2007;153:105-112. 11.
Bakos Z, Markstad H, Ostenfeld E, Carlsson M, Roijer A, Borgquist R.
TE D
8 9
Combined preoperative information using a bullseye plot from speckle tracking echocardiography, cardiac CT scan, and MRI scan: targeted left
11
ventricular lead implantation in patients receiving cardiac resynchronization
12 13
AC C
EP
10
therapy. Eur Heart J Cardiovasc Imaging 2014;15:523-531.
12.
Leyva F, Foley PW, Chalil S, Ratib K, Smith RE, Prinzen F, Auricchio A.
14
Cardiac resynchronization therapy guided by late gadolinium-enhancement
15
cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2011;13:29.
ACCEPTED MANUSCRIPT 27
1
13.
Wong JA, Yee R, Stirrat J, et al. Influence of pacing site characteristics on response to cardiac resynchronization therapy. Circ Cardiovasc Imaging
3
2013;6:542-550. 14.
Hsing JM, Selzman KA, Leclercq C, Pires LA, McLaughlin MG, McRae SE,
SC
4
RI PT
2
Peterson BJ, Zimetbaum PJ. Paced left ventricular QRS width and ECG
6
parameters predict outcomes after cardiac resynchronization therapy:
7
PROSPECT-ECG substudy. Circ Arrhythm Electrophysiol 2011;4:851-857. 15.
Iler MA, Hu T, Ayyagari S, Callahan TDt, Civello KC, Thal SG, Wilkoff BL,
TE D
8
M AN U
5
9
Chung MK. Prognostic value of electrocardiographic measurements before and after cardiac resynchronization device implantation in patients with heart
11
failure due to ischemic or nonischemic cardiomyopathy. Am J Cardiol
12 13
AC C
EP
10
2008;101:359-363.
16.
Lecoq G, Leclercq C, Leray E, Crocq C, Alonso C, de Place C, Mabo P,
14
Daubert C. Clinical and electrocardiographic predictors of a positive
15
response to cardiac resynchronization therapy in advanced heart failure. Eur
ACCEPTED MANUSCRIPT 28
2
Heart J 2005;26:1094-1100. 17.
Ypenburg C, van Bommel RJ, Delgado V, Mollema SA, Bleeker GB,
RI PT
1
Boersma E, Schalij MJ, Bax JJ. Optimal left ventricular lead position predicts
4
reverse remodeling and survival after cardiac resynchronization therapy. J
5
Am Coll Cardiol 2008;52:1402-1409. 18.
M AN U
6
SC
3
Bogun F, Good E, Reich S, Elmouchi D, Igic P, Lemola K, Tschopp D, Jongnarangsin K, Oral H, Chugh A, Pelosi F, Morady F. Isolated potentials
8
during sinus rhythm and pace-mapping within scars as guides for ablation of
9
post-infarction ventricular tachycardia. J Am Coll Cardiol
TE D
7
19.
12
I. Relation of pace mapping QRS configuration and conduction delay to
13
ventricular tachycardia reentry circuits in human infarct scars. Journal of the
14 15
Stevenson WG, Sager PT, Natterson PD, Saxon LA, Middlekauff HR, Wiener
AC C
11
2006;47:2013-2019.
EP
10
American College of Cardiology 1995;26:481-488. 20.
Stevenson WG, Weiss JN, Wiener I, Rivitz SM, Nademanee K, Klitzner T,
ACCEPTED MANUSCRIPT 29
Yeatman L, Josephson M, Wohlgelernter D. Fractionated endocardial
2
electrograms are associated with slow conduction in humans: evidence from
3
pace-mapping. J Am Coll Cardiol 1989;13:369-376. 21.
Delgado-Montero A, Tayal B, Goda A, Ryo K, Marek JJ, Sugahara M, Qi Z,
SC
4
RI PT
1
Althouse AD, Saba S, Schwartzman D, Gorcsan J, 3rd. Additive Prognostic
6
Value of Echocardiographic Global Longitudinal and Global Circumferential
7
Strain to Electrocardiographic Criteria in Patients With Heart Failure
8
Undergoing Cardiac Resynchronization Therapy. Circ Cardiovasc Imaging
9
2016;9(6):e004241.
TH. Echocardiographic predictors of reverse remodeling after cardiac
12
resynchronization therapy and subsequent events. Circ Cardiovasc Imaging
13
15
Park JH, Negishi K, Grimm RA, Popovic Z, Stanton T, Wilkoff BL, Marwick
AC C
11
14
TE D
22.
EP
10
M AN U
5
2013;6:864-872.
23.
Forleo GB, Santini L, Giammaria M, et al. Multipoint pacing via a quadripolar left-ventricular lead: preliminary results from the Italian registry on multipoint
ACCEPTED MANUSCRIPT 30
left-ventricular pacing in cardiac resynchronization therapy (IRON-MPP).
2
Europace 2017;19:1170-1177.
3
24.
RI PT
1
Ishibashi K, Kubo T, Kitabata H, et al. Improvement of cardiac function by increasing stimulus strength during left ventricular pacing in cardiac
5
resynchronization therapy. Int Heart J 2015;56:62-66. 25.
M AN U
6
SC
4
Michael R. Gold JPS, Kenneth A. Ellenbogen, Yinghong Yu, Nicholas Wold, Timothy E. Meyer, Ulrinka Birgersdoter-Green. Interventricular electrical
8
delay is predictive of response to cardiac resynchronization therapy. J Am
9
Coll Cardiol EP 2016;2(4):438-447.
TE D
7
12 13
AC C
11
EP
10
14 15
Figure Legend
ACCEPTED MANUSCRIPT 31
Figure 1. Examples of 12 lead ECG and intracardiac electrograms during sinus
2
rhythm with native QRS (A), during left ventricular (LV) pacing (B), and during
3
biventricular pacing (C). The calipers are aligned with the onset of QRS, off-set of
4
QRS, the peak of the latest LV electrogram, and the pacing stimulation signal. BiV =
5
biventricular; Q = onset of QRS; Q-LV = interval from QRS onset to latest LV
6
electrogram; QRSd = QRS duration; S-QRS = interval from stimulus to QRS onset;
7
STIM = stimulation signal.
TE D
8
M AN U
SC
RI PT
1
Figure 2. Distribution and comparison of S-QRS interval (A) and Q-LV interval (C)
10
in non-responders and responders, and the correlation between percent reduction
11
in LVESV and S-QRS interval (B) and Q-LV interval (D). LVESV indicates left
12
ventricular end systolic volume.
AC C
13
EP
9
14
Figure 3. Comparison of the percent reduction in LVESV (A) and the change in
15
LVEF (B) between the short S-QRS group and the long S-QRS group. LVESV
ACCEPTED MANUSCRIPT 32
indicates left ventricular end systolic volume; LVEF, left ventricular ejection
2
fraction.
RI PT
1
3
SC
4
M AN U
5 6 7
TE D
8 9
EP
10
12 13 14
AC C
11
Figure 1
ACCEPTED MANUSCRIPT
TE D
M AN U
SC
RI PT
33
1 2
EP
3
5 6
AC C
4
7 8
Figure 2.
ACCEPTED MANUSCRIPT 34
B
120
P < 0.01
120
60
40
0
Responder
D
250
250
100
Responder
TE D AC C
4
7 8
1
Figure 3.
0
20
40
60
y = 0.29x + 123.7 R2 = 0.04 P =0.16
200
150
100
50 -40
0
EP
3
6
Q-LV interval (ms)
150
Non-Responder
2
5
M AN U
Q-LV interval (ms)
200
-20
% Reduction in LVESV (%)
P = 0.45
1
40
0 -40
1
Non-Responder
50
60
20
20
C
80
RI PT
80
0
y = -0.42x + 46.9 R2 = 0.23 P < 0.01
100
S-QRS interval (ms)
S-QRS interval (ms)
100
SC
A
-20
0
20
40
% Reduction in LVESV (%)
60
ACCEPTED MANUSCRIPT
3 4 5 6 7
EP
2
AC C
1
TE D
M AN U
SC
RI PT
35
ACCEPTED MANUSCRIPT 36
Table 1. Baseline clinical characteristics in total population and those with
2
and without response to cardiac resynchronization therapy Total (N=52)
Responder
62.6 ± 13.9
Male gender (%)
41 (79)
Ischemic (%) Non-Ischemic (%)
62.6 ± 13.7
0.78
24 (75)
17 (85)
0.5
13 (25)
9 (28)
4 (20)
0.74
39 (75)
23 (72)
16 (80) 0.59
0 (0)
0 (0)
0 (0)
34 (65)
21 (66)
13 (65)
(%)
10 (19)
5 (16)
5 (25)
(%)
8 (15)
6 (19)
2 (10)
159.5±34.0
161.4±28.4
157.0±32.6
EP
(%)
62.6 ± 14.7
TE D
NYHA
AC C
(%)
QRS duration (ms)
p value
nder (N=20)
M AN U
Age (years)
Non-Respo
SC
(N=32)
RI PT
1
0.44
ACCEPTED MANUSCRIPT 37
QRS block type
0.67 4 (13)
LBBB (%)
36 (69)
21 (66)
IVCD (%)
9 (17)
7 (22)
3 (15)
RI PT
7 (13)
15 (75) 2 (10)
SC
RBBB (%)
M AN U
Device type 8 (15)
6 (19)
2 (10)
CRT-D (%)
44 (85)
26 (81)
18 (90)
Primary prevention (%)
35 (80)
24 (92)
11 (61)
2 (8)
7 (39)
TE D
CRT-P (%)
Secondary prevention (%)
9 (20)
0.46
0.04
Values are expressed as mean ± SD or number (%). CRT indicates cardiac
2
resynchronization therapy; IVCD, intraventricular conduction delay; LBBB, left
3
bundle branch block; NYHA, New York heart association; RBBB, right bundle
4
branch block.
AC C
EP
1
ACCEPTED MANUSCRIPT 38
Table 2. Comparison of electrocardiographic assessment during the
2
procedure in responders vs. non-responders
RI PT
1
Responder (N=32)
Non-Responder (N=20)
p value
Native QRSd (ms)
159.5 ± 34.0
161.4 ± 28.4
157.0 ± 32.6
0.44
PQ interval (ms)
223.2 ± 57.1
209.2 ± 42.5
Q-LV interval (ms)
128.0 ± 31.7
%Q-LV (%)
78.7 ± 13.2
LV pacing QRSd (ms)
245.5 ± 46.9
S-QRS interval (ms)
39.0 ± 17.5
Actual LV pacing QRSd (ms)
206.6± 39.4
3 4
< 0.01
130.8 ± 28.0
126.1 ± 29.6
0.46
81.1 ± 9.9
80.5 ± 9.2
0.76
238.5 ± 29.1
264.0 ± 40.5
0.016
30.8 ± 9.5
53.1 ± 17.4
< 0.01
207.7 ± 28.6
210.9 ± 33.6
0.41
146.3 ± 18.5
165.6 ± 21.8
< 0.01
−15.2 ± 26.0
+8.6 ± 30.0
< 0.01
M AN U
250.4± 55.4
TE D
EP
Delta-QRS (ms)
152.1 ± 28.6
AC C
Biventricular pacing QRSd (ms)
SC
Total (N=52)
−6.7 ± 29.9
Values are expressed as mean±SD. QRSd indicates QRS duration
ACCEPTED MANUSCRIPT 39
Table 3. Comparison of clinical characteristics between short S-QRS group
2
(<37ms) and long S-QRS group (≥37ms)
RI PT
1
Long S-QRS (≥37ms) (N=28)
p value
64.5 ± 13.3
60.9 ± 14.4
0.51
Age (years)
17 (71)
24 (86)
0.31
Ischemic etiology (%)
9 (37)
4 (14)
0.11
174.3 ± 80.0
216.5 ± 105.5
0.06
27.8 ± 6.3
24.5 ± 5.8
0.03
191.4 ± 79.9
< 0.01
35.4 ± 9.6
28.2 ± 7.4
< 0.01
% reduction in LVESV (%)
13.3 ± 2.7
-8.5 ± 18.3
< 0.01
Radial Dyssynchrony (ms)
215.2 ± 134.6
195.4 ± 107.9
0.67
10.5 ± 3.6
7.7 ± 3.0
< 0.01
163.1 ± 26.6
156.9 ± 32.6
0.3
M AN U
Male Gender (%)
SC
Short S-QRS (<37ms) (N=24)
LVESV at baseline (mL)
(mL)
AC C
LVEF after six months (%)
123.6 ± 61.3
EP
LVESV after six months
TE D
LVEF at baseline (%)
LV GLS (%) Native QRSd (ms)
ACCEPTED MANUSCRIPT 40
S-QRS interval (ms)
132.8 ± 28.7
125.6 ± 28.3
0.33
26.2 ± 5.5
50.6 ± 15.1
< 0.01
259.8 ± 35.4
< 0.01
209.1 ± 29.7
0.49
234.9 ± 31.9
Actual LV pacing QRSd
208.8 ± 31.8
SC
LV pacing QRSd (ms)
RI PT
Q-LV interval (ms)
M AN U
(ms) Biventricular pacing QRSd
145.0 ± 18.0
161.1 ± 22.3
< 0.01
-18.0 ± 26.7
+4.2 ± 28.6
< 0.01
114.6 ± 25.4
122.9 ± 31.3
0.5
153.3 ± 29.7
162.5 ± 35.0
0.55
-12.3 ± 17.3
-11.8 ± 18.1
0.83
23 (96)
9 (32)
< 0.01
(ms)
Sensed AVD (ms)
EP
Paced AVD (ms)
TE D
Delta-QRS (ms)
AC C
VVD (ms) Responder rate (%) 1
Values are mean±SD or number (%). AVD indicates atrioventricular delay; LVEF,
2
left ventricular ejection fraction; LVESV, left ventricular end systolic volume; LVGLS,
ACCEPTED MANUSCRIPT 41
left ventricular global longitudinal strain; QRSd, QRS duration, VVD; interventricular
2
delay. P<0.01 when compared with baseline.
AC C
EP
TE D
M AN U
SC
RI PT
1
ACCEPTED MANUSCRIPT 42
Table 4. Multivariate analysis for prediction of response to cardiac
2
resynchronization therapy Univariate analysis
Multivariate analysis
95% CI
p value
LVEDV (mL)
1
0.99 - 1.01
0.91
LVESV (mL)
1
0.99 - 1.01
0.96
LVEF (%)
1.04
0.95 - 1.14
0.41
Radial Dyssynchrony
1
1.00 - 1.01
0.24
Odds ratio
95% CI
p value
1.14
0.85 - 1.51
0.39
LV GLS (%)
1.24
Native QRSd (ms)
EP
TE D
M AN U
SC
Odds ratio
(ms)
RI PT
1
AC C
1.01
1.04 - 1.52
0.023
0.99 - 1.03
0.6
PQ interval (ms)
0.98
0.96 - 0.99
0.012
0.97
0.94 - 1.0
0.01
Long S-QRS interval
0.02
0.002 - 0.18
< 0.01
0.014
0.001 - 0.26
< 0.01
0.98
0.96 - 1.0
0.02
(≥37ms) LV pacing QRSd (ms)
ACCEPTED MANUSCRIPT
Actual LV pacing QRSd
1
0.98 - 1.02
0.71
0.95
0.92 - 0.98
< 0.01
0.97
0.94 - 0.99
< 0.01
(ms) Biventricular pacing
M AN U
Delta-QRS (ms)
0.97
SC
QRSd (ms)
RI PT
43
0.99
0.92 - 1.02
0.17
0.95 - 1.02
0.44
1
CI indicates confidence interval. Other abbreviations are same as those in Tables 2
2
and 3.
6 7 8 9 10
EP
5
AC C
4
TE D
3