Shortening of atrioventricular delay at increased atrial paced heart rates improves diastolic filling and functional class in patients with biventricular pacing.

Abstract

Use of rate adaptive atrioventricular (AV) delay remains controversial in patients with biventricular (Biv) pacing. We hypothesized that a shortened AV delay would provide optimal diastolic filling by allowing separation of early and late diastolic filling at increased heart rate (HR) in these patients.
34 patients (75 ± 11 yrs, 24 M, LVEF 34 ± 12%) with Biv and atrial pacing had optimal AV delay determined at baseline HR by Doppler echocardiography. Atrial pacing rate was then increased in 10 bpm increments to a maximum of 90 bpm. At each atrial pacing HR, optimal AV delay was determined by changing AV delay until best E and A wave separation was seen on mitral inflow pulsed wave (PW) Doppler (defined as increased atrial duration from baseline or prior pacemaker setting with minimal atrial truncation). Left ventricular (LV) systolic ejection time and velocity time integral (VTI) at fixed and optimal AV delay was also tested in 13 patients. Rate adaptive AV delay was then programmed according to the optimal AV delay at the highest HR tested and patients were followed for 1 month to assess change in NYHA class and Quality of Life Score as assessed by Minnesota Living with Heart Failure Questionnaire.
81 AV delays were evaluated at different atrial pacing rates. Optimal AV delay decreased as atrial paced HR increased (201 ms at 60 bpm, 187 ms at 70 bpm, 146 ms at 80 bpm and 123 ms at 90 bpm (ANOVA F-statistic = 15, p = 0.0010). Diastolic filling time (P < 0.001 vs. fixed AV delay), mitral inflow VTI (p < 0.05 vs fixed AV delay) and systolic ejection time (p < 0.02 vs. fixed AV delay) improved by 14%, 5% and 4% respectively at optimal versus fixed AV delay at the same HR. NYHA improved from 2.6 ± 0.7 at baseline to 1.7 ± 0.8 (p < 0.01) 1 month post optimization. Physical component of Quality of Life Score improved from 32 ± 17 at baseline to 25 ± 12 (p < 0.05) at follow up.
Increased heart rate by atrial pacing in patients with Biv pacing causes compromise in diastolic filling time which can be improved by AV delay shortening. Aggressive AV delay shortening was required at heart rates in physiologic range to achieve optimal diastolic filling and was associated with an increase in LV ejection time during optimization. Functional class improved at 1 month post optimization using aggressive AV delay shortening algorithm derived from echo-guidance at the time of Biv pacemaker optimization.

Full text

Page 1

RESEARCHOpen Access
Shortening of atrioventricular delay at increased
atrial paced heart rates improves diastolic filling
and functional class in patients with biventricular
pacing
Reza Rafie1, Salima Qamruddin1, Ali Ozhand2, Nima Taha1and Tasneem Z Naqvi1*
Abstract
Background: Use of rate adaptive atrioventricular (AV) delay remains controversial in patients with biventricular
(Biv) pacing. We hypothesized that a shortened AV delay would provide optimal diastolic filling by allowing
separation of early and late diastolic filling at increased heart rate (HR) in these patients.
Methods: 34 patients (75 ± 11 yrs, 24 M, LVEF 34 ± 12%) with Biv and atrial pacing had optimal AV delay
determined at baseline HR by Doppler echocardiography. Atrial pacing rate was then increased in 10 bpm
increments to a maximum of 90 bpm. At each atrial pacing HR, optimal AV delay was determined by changing AV
delay until best E and A wave separation was seen on mitral inflow pulsed wave (PW) Doppler (defined as
increased atrial duration from baseline or prior pacemaker setting with minimal atrial truncation). Left ventricular
(LV) systolic ejection time and velocity time integral (VTI) at fixed and optimal AV delay was also tested in 13
patients. Rate adaptive AV delay was then programmed according to the optimal AV delay at the highest HR
tested and patients were followed for 1 month to assess change in NYHA class and Quality of Life Score as
assessed by Minnesota Living with Heart Failure Questionnaire.
Results: 81 AV delays were evaluated at different atrial pacing rates. Optimal AV delay decreased as atrial paced HR
increased (201 ms at 60 bpm, 187 ms at 70 bpm, 146 ms at 80 bpm and 123 ms at 90 bpm (ANOVA F-statistic =
15, p = 0.0010). Diastolic filling time (P < 0.001 vs. fixed AV delay), mitral inflow VTI (p < 0.05 vs fixed AV delay) and
systolic ejection time (p < 0.02 vs. fixed AV delay) improved by 14%, 5% and 4% respectively at optimal versus
fixed AV delay at the same HR. NYHA improved from 2.6 ± 0.7 at baseline to 1.7 ± 0.8 (p < 0.01) 1 month post
optimization. Physical component of Quality of Life Score improved from 32 ± 17 at baseline to 25 ± 12 (p < 0.05)
at follow up.
Conclusions: Increased heart rate by atrial pacing in patients with Biv pacing causes compromise in diastolic filling
time which can be improved by AV delay shortening. Aggressive AV delay shortening was required at heart rates
in physiologic range to achieve optimal diastolic filling and was associated with an increase in LV ejection time
during optimization. Functional class improved at 1 month post optimization using aggressive AV delay shortening
algorithm derived from echo-guidance at the time of Biv pacemaker optimization.
Keywords: Biventricular pacemaker, Rate Adaptive AV Delay, Diastolic Function, Echocardiography, Doppler
* Correspondence: tnaqvi@usc.edu
1Echocardiographic Laboratories, Department of Medicine, Keck School of
Medicine, University of Southern California, Los Angeles, California, 90033
USA
Full list of author information is available at the end of the article
Rafie et al. Cardiovascular Ultrasound 2012, 10:2
http://www.cardiovascularultrasound.com/content/10/1/2
CARDIOVASCULAR
ULTRASOUND
© 2012 Rafie et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.

Page 2

Cardiac resynchronization therapy (CRT) in patients
with advanced heart failure leads to an improvement in
cardiac function during rest, exercise [1] and pharmaco-
logical stress [2]. Previous studies have shown that in
patients with CRT, optimizing atrioventricular (AV)
delay further improves stroke volume [3-6] and that
optimal AV delay changes over time requiring periodic
optimization [7]. There is controversy on optimal AV
delay at increase in heart rate with exercise. Studies
have shown conflicting data suggesting AV delay should
be prolonged [8] or shortened [9,10] as heart rate
increases. In patients with heart failure, an increase in
heart rate can compromise diastolic filling to preserve
systolic ejection [11]. This diastolic filling time can be
manipulated in heart failure patients with CRT by
adjusting AV delay at increased heart rates by using rate
adaptive algorithms of CRT devices. In this study in
patients with CRT, we evaluated a) the effect of
increased atrial pacing rate on diastolic filling at a fixed
AV delay and b) to examine the effect of manipulating
AV delay on diastolic filling. Our secondary goal was to
examine the effect of programming AV delay that pro-
vided most optimal diastolic filling at increased heart
rate on cardiac ejection acutely and on functional class
at 1 month follow up.
Methods
Study Subjects
34 heart failure patients with Biv-CRT who were atrially
pacing at resting heart rates were included in the study.
Exclusion criteria
Patients with atrial fibrillation, atrial pacing of less than
50% on pacemaker interrogation, recent hospital admis-
sion within 2 months for congestive heart failure and
acute coronary event requiring revascularization within
3 months were excluded from the study. All patients
signed an informed consent approved by the Institu-
tional Review Board. 14 patients from this study were
part of the 2 studies reported earlier [5,6].
Pacing Protocol
AV delay and interventricular (VV) delays were opti-
mized at rest using Doppler echocardiography by con-
ventional methods. AV delay that provided that most
optimal mitral inflow filling without atrial truncation
and with best E and A separation was chosen at optimal
AV delay. Left ventricular outflow tract (LVOT) pulsed
wave (PW) Doppler was measured at rest and at each
AV delay along with pulmonary vein PW Doppler. After
optimal AV delay was determined at rest, atrial pacing
rate was increased at 10 bpm increments from baseline
heart rate of 60 or 70 bpm to a maximum of 90 bpm.
At each atrial pacing rate, mitral inflow PW Doppler
was first evaluated at the AV delay found optimal at the
previous paced atrial rate. Then the AV delay was
decreased in decrements of 20 ms until best E and A
wave separation was seen on mitral inflow without atrial
truncation. If E and A separation did not occur despite
decreasing AV delay, AV delay was increased in incre-
ments of 20 ms until E and A separation was noted.
Atrial pacing rate ranged from 60-90 beats per minute.
In 13 patients LV ejection time was also evaluated at
each paced atrial rate before and after optimal AV delay
change. PW Doppler assessment was performed 5 beats
of after change in pacemaker setting.
Echocardiography
We used transthoracic echocardiography using GE Vivid
7 or Vivid 9 ultrasound systems (Horten, Norway)
equipped with a variable frequency 2.5-5 MHz transducer
with harmonics. Echocardiogram was performed in the
left lateral decubitus position. AV delay was first opti-
mized at resting heart rate partially adopting the recent
guidelines from American Society of Echocardiography
[12]. PW Doppler sample volume placed at the tip of the
mitral leaflets for LV diastolic filling parameters. PW
Doppler sample volume was placed 0.5-1 cm below the
aortic valve at the LV outflow tract (LVOT) to obtain the
LV ejection duration and aortic velocity time integral
(VTI). Frame rate was kept above 100 fps by using single
focus, narrow imaging sector, appropriate depth and
frame rate. Parallel Doppler beam alignment to myocar-
dial segments and color Doppler was used for all Doppler
data acquisition. ECG was displayed on the ultrasound
system and 3 cardiac cycles were used for each data
acquisition. Raw data was stored digitally as DICOM cine
loops and transferred for offline analysis to a customized
dedicated workstation equipped with custom built soft-
ware (Echo PAC PC Dimension version 6.0.1, GE
Vingmed Ultrasound, Horten, Norway) via internet.
Offline Measurements
All routine chamber dimensions and LV ejection frac-
tion at baseline were measured off line. Mitral inflow
peak E and A velocities, E wave deceleration time and
mitral inflow VTI were measured at baseline and at
each paced heart rate. We also measured LVOT PW
Doppler ejection time and VTI in 13 patients at fixed
AV delay and optimal AV delay at increased atrial
pacing rates. Two blinded independent investigators
performed offline analysis of all echo data. Investigators
performing offline analysis were blinded to online
assessment of optimal AV delay settings.
Functional Class Assessment
NYHA class was evaluated at baseline and Quality of
Life was assessed by Minnesota Living with Heart
Rafie et al. Cardiovascular Ultrasound 2012, 10:2
http://www.cardiovascularultrasound.com/content/10/1/2
Page 2 of 11

Page 3

Failure Questionnaire (MHFQ) at baseline. NYHA class
and MHFQ were re-evaluated by a questionnaire given
to the patient in a self addressed envelope to be mailed
in at 1 month post optimization. A hall walk as toler-
ated by the patient was performed on all patients after
optimization to ensure no deterioration of exercise capa-
city from baseline.
Statistical analysis
All measurements were displayed as mean ± SD. Mea-
surements before and after optimization for AV delay
for each heart rate as well as NHYA and functional
class scores were analyzed by Paired student t- test. For
comparison of different heart rates, we used analysis of
variance (ANOVA). We used repeated measures
ANOVA for optimal measured AV delay at more than
two different heart rates. P < 0.05 was considered statis-
tically significant.
Results
34 patients were included in our study. Baseline charac-
teristics of the study population are shown in Table 1.
The mean duration of CRT implant prior to optimiza-
tion was 23 ± 27 months.
Effect of Optimal AV delay Programming at Resting
Paced Atrial Rate
Change in diastolic filling and systolic ejection PW Dop-
pler parameters and on peak pulmonary artery pressure
from baseline AV delay to optimal AV delay at resting
HR is shown in Table 2. Figure 1 is a study example of
a patient showing mitral inflow, pulmonary vein inflow
and LVOT PW Doppler measurement at baseline AV
delay and at optimal AV delay at resting HR of 60 bpm.
Increased Atrial Pacing
Patients tolerated increasing atrial pacing rate without
developing arrhythmia or hypotension. Mean duration
of atrial pacing during which Doppler echo data for dia-
stolic filling time at fixed and variable AV delays was
collected was 2 minutes. In 13 patients in whom echo
Doppler systolic ejection time before and after AV delay
optimization was also obtained, mean duration of atrial
pacing was 3 minutes.
Changes in Optimal AV Delay With Increasing Heart Rate
81 AV delays were evaluated at different atrial pacing
rates. Data on optimal AV delay at each paced heart
rate in the overall study population is shown in Table 3.
In 32 patients, optimal AV delay decreased as atrial
paced rate increased. Optimal AV delay was 201 ms at
60 bpm, 187 ms at 70 bpm, 146 ms at 80 bpm and 123
ms at 90 bpm (p = 0.001 by ANOVA for all) and is
shown in Figure 2. Repeated measures ANOVA showed
a statistically significant shortening of AV delay at three
sets of heart rates (60, 70 and 80 bpm, P < 0.001). Opti-
mal AV delay at baseline and increased paced atrial rate
in each patient is shown in Figure 3.
Effect of Optimal AV Delay on Diastolic Parameters
Data on diastolic filling time, mitral inflow VTI, E wave
deceleration time, E and A wave velocities, E/A ratio,
LVOT ejection time and LVOT VTI at various heart
rates during fixed and optimal AV delays is shown in
Table 4. Figures 4 and 5 show mitral inflow PW Dop-
pler tracings with increased atrial pacing rate at fixed
and optimal AV delay at a heart rate of 70 bpm (Figure
4) and then 80 bpm (Figure 5) in a study patient. Abso-
lute and percent changes in PW Doppler mitral inflow
and of LVOT at fixed and optimal AV delay at 81 AV
delays tested are shown in Table 5. Diastolic filling time
was significantly longer at optimal AV delay with a
mean difference at fixed vs. optimal AV delay of 40.3 ±
25 ms, p < 0.001. Decreasing optimal AV delay with
increased heart rate was associated with an increase in
Table 1 Baseline Characteristics of the Study Population
Age
75 ± 11 years
Gender
Male/Female: 24(72%)/10(28%)
NYHA
2.6 ± 0.7 - I/II/III/IV: 3%/34%/47%/
16%
NYHA
2.6 ± 0.7 - I/II/III/IV: 3%/34%/47%/
16%
Cardiomyopathy
Ischemic/Non- Ischemic: 69%/
31%
QRS Duration
175.3 ± 38.6 ms
Left Ventricular Diastolic
Diameter
5.9 ± 0.93 cm
Left Ventricular Systolic Diameter
4.9 ± 1.1 cm
Left Ventricular Ejection Fraction
34.1 ± 12.4%
Left Atrial Diameter (Superior-
Inferior)
6 ± 1 cm
Left Atrial Diameter (Antero-
Posterior)
4.7 ± 0.8 cm
AV Delay
154.1 ± 40.1 (baseline)
Table 2 Effect of Biv Pacemaker Optimization on Echo
Doppler Variables at Baseline Heart Rate
Echo Doppler VariablesBaselineOptmialP Value
LV Diastolic Filling Time (ms)
396 ± 75420 ± 630.016
LVOT VTI (cm)
17.5 ± 4.8 19.4 ± 5.2 < 0.001
LV ejection time (ms)
287 ± 40298 ± 36 < 0.001
Pulmonary artery pressure (mm Hg) 52 ± 744 ± 80.001
Values are Mean ± SD LV: left ventricle.; LVOT: left ventricular outflow tract;
VTI: velocity time integral
Rafie et al. Cardiovascular Ultrasound 2012, 10:2
http://www.cardiovascularultrasound.com/content/10/1/2
Page 3 of 11

Page 4

mitral inflow PW Doppler VTI. No significant change
was seen in mitral inflow deceleration time at fixed vs.
optimal AV delay. Mitral inflow A wave and E/A ratio
showed statistically significant changes after optimiza-
tion of AV delay at different heart rates (Table 4).
LVOT systolic ejection time increased at optimal vs.
fixed AV delay at increased heart rates, with a mean
change of 11 ± 15 ms (p < 0.02). A trend for an
increased in LVOT VTI was noted before and after opti-
mization (Tables 4 and 5).
Final Pacemaker Settings
Rate adaptive AV delay feature was turned ON in all
patients and AV delay shortening was programmed
according to AV delays found optimal at increased
paced heart rate in each patient. In patients with St Jude
device, AV delay settings were changed to the “most
aggressive” AV delay shortening.
25 of the patients reported an improvement in their
ability to walk immediately after optimization compared
to before optimization and 5 patients were unable to tell
Figure 1 Effect of AV delay optimization on diastolic filling. Mitral inflow (A and B), pulmonary vein inflow (C and D), and LV outflow PW
Doppler (E and F) at baseline (Apace 60 AV delay 150, LV 16 ms): A, C, and E) and optimal (A-paced 60 beats/min, AV delay 190 ms, VV0 ms: B,
D, and F) pacemaker settings. An increase in A wave duration occurred (B), and pulmonary vein pattern changed so that systolic fraction
increased, S wave deceleration time increased (D) and increased pulmonary vein atrial reversal (C) was abolished indicating reduction in left atrial
pressure. An improvement in percentage LV ejection time also occurred (F).
Table 3 Optimal AV Delay at Different Paced Atrial Rates in the Study Patients
Heart rate
(bpm)
N Optimal AV delay(ms)*95% Confidence Interval Minimum
AV delay(ms)
Maximum
AV delay(ms)
609 201 ± 48 169.64-232.36190 230
70 36187 ± 44172.63-201.37 100280
8025 146 ± 42129.54-162.46 80200
90 21123 ± 49 102.04-143.9650 250
Values are Mean ± SD
*Optimal AV delay at different heart rates (P value = 0.001 ANOVA)
Rafie et al. Cardiovascular Ultrasound 2012, 10:2
http://www.cardiovascularultrasound.com/content/10/1/2
Page 4 of 11

Page 5

the difference. No patient reported deterioration in exer-
cise tolerance. NYHA improved from 2.5 ± 0.8 at base-
line to 1.7 ± 0.8 (p < 0.01) at one month post
optimization. MHFQA Failure Questionnaire was
returned at the end of one month by 20 subjects.
MHFQA improved from 49 ± 30 to 39 ± 30 (p = 0.2).
When emotional and physical components of assess-
ment were separated, a significant improvement was
noted in the physical component of quality of life from
32 ± 17 to 25 ± 12, p < 0.05.
NYHA Class by Device Manufacturer
There were 14 Medtronic devices, 7 Boston Scientific
devices and 10 St. Jude devices. Mean NYHA improve-
ment for St Jude was 0.10 compared to 0.87 for all
other brands (p = 0.068). No differences were found for
other brands.
Discussion
The salient findings of our study are that increased atrial
pacing rate causes a marked shortening of diastolic fill-
ing time at a fixed AV delay and a shortening of AV
delay at increased heart rate improves diastolic filling.
This shortening of AV delay is associated with an
improvement in LV ejection time and indirectly LV
stroke volume. Optimization of AV delay at rest and
programming of aggressive rate adaptive delay was asso-
ciated with an improvement in NYHA functional class
at 1 month follow up. The shortening of AV delay at
increased heart rates allows adequate diastolic filling
time by causing separation of early and late diastolic fill-
ing and in turn leads to an improvement in cardiac ejec-
tion as measured by LV systolic ejection time.
Aggressive shortening of AV delay at increased heart
rates is further supported by the improvement in func-
tional class that occurred in our patients at one month
follow up.
Modern pacemaker devices allow rate-response as well
as rate-adaptive AV delay shortening during exercise.
Current pacemakers do not allow AV delay lengthening
at increased heart rates. In our study the amount of AV
delay shortening needed to optimize diastolic filling was
more aggressive than current default automated pace-
maker algorithms provide. This could be related to cur-
rent rate-adaptive AV delay algorithms based on data
derived from patients with conduction abnormalities
who had otherwise preserved right ventricular and LV
systolic function whereas patients with CRT have
marked cardiac dysfunction.
We did not have a control group in whom rate adap-
tive AV delay shortening was not performed to differ-
entiate effects of optimization at resting heart rates vs.
optimization at resting heart rates and use of rate adap-
tive AV delay, however improvement in NYHA class
and quality of life in our patients suggests that program-
ming of aggressive rate adaptive delay shortening is not
deleterious, as some studies suggest, and is likely to be
beneficial. We selected patients who were atrial pacing
in addition to Biv pacing at baseline since a constant
heart rate provided by atrial pacing allowed patients to
act as their own control in whom fixed versus shortened
AV delay effects in a given cardiac cycle of constant
length could be tested.
In humans with normal cardiac function, there is
decreasing AV delay with increasing heart rate in a pre-
dictable manner (4 ms/10 bpm). Patients with dilated
Figure 2 Effect of increasing paced atrial rate on optimal AV delay in the study population.
Rafie et al. Cardiovascular Ultrasound 2012, 10:2
http://www.cardiovascularultrasound.com/content/10/1/2
Page 5 of 11