Download Survival After Rate-Responsive Programming in Patients With

Original Article
Survival After Rate-Responsive Programming in Patients
With Cardiac Resynchronization Therapy-Defibrillator
Implants Is Associated With a Novel Parameter
The Heart Rate Score
Brian Olshansky, MD; Mark Richards, PhD, MD; Arjun Sharma, MD; Nicholas Wold, MS;
Paul Jones, MS; David Perschbacher, BS, MD; Bruce L. Wilkoff, MD
Downloaded from http://circep.ahajournals.org/ by guest on May 12, 2017
Background—Rate-responsive pacing (DDDR) versus nonrate-responsive pacing (DDD) has shown no survival benefit for
patients undergoing cardiac resynchronization therapy defibrillator (CRT-D) implants. The heart rate score (HRSc), an
indicator of heart rate variation, may predict survival. We hypothesized that high-risk HRSc CRT-D patients will have
improved survival with DDDR versus DDD alone.
Methods and Results—All CRT-D patients in LATITUDE remote monitoring (2006–2011), programmed DDD, had HRSc
calculated at first data upload after implant (median 1.4 months). Patients subsequently reprogrammed to DDDR 7.6
median months later were compared with a propensity-matched DDD group and followed for 21.4 median months by
remote monitoring. Data were adjusted for age, sex, lower rate limit, percent atrial pacing, percent biventricular pacing,
and implant year. The social security death index was used to identify deaths. Remote monitoring provided programming
and histogram data. DDDR programming in CRT-D patients was associated with improved survival (adjusted hazard ratio
=0.77; P<0.001). However, only those with baseline HRSc ≥70% (2308/6164) had improved HRSc with DDDR (from
88±9% to 78±15%; P<0.001) and improved survival (hazard ratio =0.74; P<0.001). Patients with a high baseline HRSc
and significant improvement over time were more likely to survive (hazard ratio =0.63; P=0.006). For patients with HRSc
<70%, DDDR reprogramming increased the HRSc from 46±11% to 50±15% (P<0.001); survival did not change. The
HRSc did not change with DDD pacing over time.
Conclusions—In CRT-D patients with HRSc ≥70%, DDDR reprogramming improved the HRSc and was associated with
survival. Patients with lower HRSc had no change in survival with DDDR programming. (Circ Arrhythm Electrophysiol.
2016;9:e003806. DOI: 10.1161/CIRCEP.115.003806)
Key Words: cardiac resynchronization therapy ◼ heart rate score ◼ mortality ◼ pacemaker optimization
◼ pacing ◼ rate-responsive pacing
C
ardiac resynchronization therapy (CRT) has been shown
to improve quality-of-life, functional status, exercise
capacity, and survival in patients with left ventricular systolic
dysfunction, moderate to severe heart failure symptoms, and
a wide QRS.1–4 Controlled, randomized CRT trials have used
atrial sensing to provide atrial-synchronized biventricular pacing, but have not focused on atrial pacing or rate response.
Current CRT devices are capable of atrial pacing for atrial rate
support and rate-responsive pacing (DDDR), and these are 2
commonly programmed modes in CRT devices. However, no
randomized prospective clinical trial has shown survival benefit from these pacing modes.
Increasing heart rate through atrial support pacing in heart
failure patients can increase cardiac output. However, higher
resting and mean heart rates are associated with increased mortality.5,6 In a heart failure population, although lower resting
rates are associated with better outcomes, no one has specifically considered outcomes in those with lower resting heart
rates but better heart rate response with exercise. Similarly, in
an implantable cardioverter defibrillator (ICD) population, the
mean heart rate predicts outcomes of heart failure hospitalization and total mortality, but this generally represents resting
rate and does not provide information about heart rate changes
with exertion or in other circumstances. Chronotropic incompetence, either pathophysiologic or iatrogenic (eg, β-blocker
therapy), is common in heart failure patients. In these patients,
it is reasonable to suspect that the hemodynamic benefits of
biventricular pacing may be enhanced by increasing atrial
rates with increased physical demands.
Large, remote monitoring databases offer the opportunity to
assess associations between programming parameters and clinical outcomes. A novel parameter, the heart rate score (HRSc),
Received September 15, 2015; accepted June 30, 2016.
From the Mercy Hospital, Mason City, IA (B.O.); ProMedica Cardiology, Toledo, OH (M.R.); Boston Scientific, St Paul, MN (A.S., N.W., P.J., D.P.);
and Cleveland Clinic, Cleveland, OH (B.L.W.).
Correspondence to Brian Olshansky, MD, Professor Emeritus, University of Iowa Hospitals, 200 Hawkins Dr, 4426a JCP, Iowa City, IA 52242. E-mail
[email protected]
© 2016 American Heart Association, Inc.
Circ Arrhythm Electrophysiol is available at http://circep.ahajournals.org
1
DOI: 10.1161/CIRCEP.115.003806
2 Olshansky et al Heart Rate Score Is Associated With Survival
WHAT IS KNOWN
• Rate-responsive
pacing has been utilized in various clinical circumstances, even though randomized
clinical trials do not show benefit.
• Chronotropic incompetence can occur in patients
with cardiac resynchronization therapy defibrillators
(CRT-D), and its presence may identify a population
at greater risk of dying.
WHAT THE STUDY ADDS
• The
Downloaded from http://circep.ahajournals.org/ by guest on May 12, 2017
Heart Rate Score, a measure of chronotropic
incompetence, identified CRT-D patients who have
improved survival with rate-responsive (DDDR)
pacing.
• Rate-responsive pacing improves survival in patients
with chronotropic incompetence.
may help distinguish which patients are in need of DDDR pacing. A high HRSc (≥70%) has been associated with both chronotropic incompetence and increased mortality.7,8 We hypothesized
that CRT defibrillator (CRT-D) patients with a HRSc ≥70%
identified via remote monitoring would have improved survival
with DDDR versus nonrate-responsive (DDD) pacing.
Methods
HRSc was defined as the percent of all paced and sensed atrial events
in the single tallest 10 beats per minute rate histogram bin (Figure 1).
A HRSc of 100% requires that all paced and sensed beats occur in
the same histogram bin. Lower HRScs occur when atrial events are
distributed over a range of rates. Therefore, the HRSc is a geometric
measure of heart rate recorded by device histograms. It is related to
the long-term variation of heart rate (Figure 1). The first observations that led to the development of the concept of the HRSc originated from the LIFE trial (Limiting Chronotropic Incompetence for
Pacemaker Recipients), in which chronotropic incompetence was associated with a HRSc ≥70%.7 Following this, a prospective analysis
(unpublished) revealed that survival in patients with ICDs or CRT-Ds
was decreased in those with higher HRSc scores.8
Patients included in this study underwent implantation of dualchamber Boston Scientific Corporation CRT-Ds and were monitored
remotely via the LATITUDE remote monitoring system from 2006 to
2011. The data were part of the ALTITUDE study database, which is
a deidentified database whose use has been approved by the Boston
Scientific governance board. No individual patients were studied, and,
therefore, no institutional review board approval was necessary. Patients
programmed DDD as an initial setting had a HRSc calculated at the
time of the first LATITUDE upload. The baseline HRSc was determined
from the first data upload post device implant. HRScs were divided into
3 groups: <30%, 30% to 69%, and ≥70% for simplicity, but were also
considered as a continuous variable (unpublished) and in deciles.8 The
rationale for the 3 HRSc groups was based on the LIFE study analysis,
with only HRSc group ≥70% having a reduction in HRSc with DDDR
pacing.7 Deaths as reported through the manufacturer’s device tracking
system and through the Social Security Death Index.
Study Cohort
All patients programmed to the DDD mode at time of initial data upload were included (Figure 2). Devices implanted before US market
release of the LATITUDE system in 2006 were excluded. Patients
with a baseline HRSc, defined by a histogram rate bin above 170
beats per minute, were excluded. That is, patients were excluded if
the tallest histogram bin occurred at a rate above the nominal setting
for atrial tachycardia response mode switch, likely indicating persistent or chronic atrial fibrillation. Patients with programming changes
to the lower rate limit parameter during follow-up were excluded
(Figure 2) because this parameter has been shown to be independently associated with survival.
Patients initially programmed to DDD mode and reprogrammed
to DDDR mode during follow-up were propensity score–matched to
patients who remained in DDD mode throughout follow-up. For purposes of the match and subsequent analyses, a transition date was
determined for each patient. For patients reprogrammed to DDDR,
the date of transition was defined as the date of reprogramming to
DDDR. For patients remaining in DDD throughout follow-up, the
date of transition was defined as the average date of reprogramming
in the DDDR group: 15 months post implant. The HRSc at the date of
transition was defined as the baseline HRSc.
On average, baseline HRScs were calculated using >16 000 000
beats, with 96% of patients having their baseline HRScs calculated
with at least 1 000 000 beats. DDDR patients were matched 1:1 to
DDD patients by baseline HRSc, age, sex, implant year, biventricular
Figure 1. Example of a patient with heart rate (HR) score (HRSc) calculated during baseline nonrate-responsive pacing (DDD) mode and after
programming to rate-responsive pacing (DDDR) mode with reduction in HRSc because of more atrial pacing at rates above the lower rate limit.
3 Olshansky et al Heart Rate Score Is Associated With Survival
Figure 2. Study design. Note that patients
included in the analysis need to be on
remote follow-up, have a cardiac resynchronization therapy defibrillators (CRT-D)
implanted after 2006, have no atrial fibrillation, have a constant lower rate limit
programmed, and be propensity score
matched. DDD indicates nonrate-responsive
pacing; DDDR, rate-responsive pacing; HR,
heart rate; and LRL, lower rate limit.
Downloaded from http://circep.ahajournals.org/ by guest on May 12, 2017
pacing percentage, presence of device-determined paroxysmal
atrial fibrillation, lower rate-limit programming, and the time to
LATITUDE set-up. A DDD match candidate was required to be active and alive at the time of DDDR reprogramming.
Statistical Analysis
Demographics for the DDDR and DDD groups were compared using standardized difference scores.9,10 Kaplan–Meier survival curves
were constructed for baseline HRSc group (<30%, 30% to 69%,
and ≥70%). Stratified log-rank tests were performed to compare patients initially programmed to DDD mode to those reprogrammed
to DDDR mode during follow-up. Stratification by propensity score
quintile was used because it cannot be assumed that the DDDR
and DDD groups are independent.11 Time to event was analyzed
from date of transition (ie, DDDR programming date for DDDR
group and at 15-month postimplant date for each DDD match).
Multivariate Cox proportional hazards models were performed adjusted for age, sex, lower rate limit, percent atrial pacing, percent
biventricular pacing, and implant year. To account for the matched
nature of the sample, each Cox model was stratified by propensity score quintile and used a robust sandwich variance estimator.12
Change in HRSc was assessed for each baseline HRSc group using a paired t test, with each pair consisting of a DDDR patient
and a DDD patient. The paired t test was used because it cannot
be assumed that the DDDR and DDD groups are independent.13
Programming of the lower rate limit was constant during the baseline period. Follow-up was considered through December 31, 2012,
to allow for a lag in reporting of deaths.
Results
The entire ALTITUDE CRT-D population included 42 894
patients. After excluding patients who had devices implanted
before 2006 (N=4748), those who had a predominant
HRSc>170 beats per minute likely because of chronic atrial
fibrillation (N=1678), and those with inconsistent lower rate
limit programming between implant and the transition date
(N=1838), the propensity-matched study cohort included 6164
patients. There were 3082 patients consistently programmed
to DDD throughout follow-up and 3082 reprogrammed from
DDD to DDDR during follow-up (Figure 2). The median time
from implant to first LATITUDE upload was 1.4 (interquartile
range 0.7–3.2) months, and the median time from implant to
DDDR reprogramming was 7.6 (interquartile range 3.4–17.2)
months. The median follow-up time after the transition date
was 21.4 (interquartile range 9.6–36.8) months, and the total
follow-up time included 12 119 patient-years.
Baseline demographics are shown in Table. Age was
associated with the HRSc, and the majority of patients in
the matched cohort had a baseline HRSc of 30% to 69%.
However, 37% had a HRSc ≥70%.
In the propensity-matched DDD group, follow-up times were
comparable to the those of the DDDR group (DDDR, 2.0 years;
DDD, 1.9 years). Over time, the mean HRSc for all patients in
each group was nearly constant (DDDR, 1.0% decrease, DDD,
0.5% increase). However, when analyzed by 10% increments in
HRSc, there were clear differences in the influence of DDDR
programming on HRSc (Figure 3). In patients with baseline
HRSc ≥70%, DDDR pacing decreased HRSc significantly
(Figure 4). The most substantial decrease occurred in the 91%
to 100% group (>13% decrease in HRSc Figure 3), but it was
present throughout the HRSc ≥70% group (a 10% decrease in
HRSc for the ≥70% group). For the other 2 HRSc groups, 30%
to 69% and <30% DDDR programming actually led to modest
increases in the HRSc group (Figure 4; P<0.001 for each group).
Time-to-event analyses, comparing propensity-matched
groups based on the baseline HRSc, showed that only patients
with a baseline HRSc ≥70% had survival benefit with DDDR
programming (Figures 5A and 6; hazard ratio =0.74; 95%
confidence interval 0.63–0.87; P<0.001). The HRSc 30% to
69% group (Figures 5B and 6) also showed minor nonsignificant survival effect with DDDR pacing (hazard ratio =0.90,
95% confidence interval 0.78, 1.02; P=0.106). The lowest
HRSc group (<30%) showed no survival benefit with DDDR
programming (P=0.403; Figures 5C and 6).
We considered the relationship between HRSc improvement and survival by patient rather than population. There
was a tight relationship between HRSc improvement and
outcomes in patients reprogrammed to DDDR with baseline
HRSc ≥70% (hazard ratio =0.63; P=0.006). Furthermore, in
the group with HRSc ≥70%, despite reduction in their HRSc,
4 Olshansky et al Heart Rate Score Is Associated With Survival
Table. Baseline Patient Demographics for HRSc Group and Programming Mode
Total
HR Score <30%
Statistic
DDD→DDDR
(N=3082)
DDD Only
(N=3082)
Standardized
Difference*
DDD→DDDR
(N=117)
DDD Only
(N=117)
Female
N (%)
933 (30%)
929 (30%)
0.00
40 (34%)
40 (34%)
Age, y
Mean±SD
70±10
70±10
−0.02
63±12
63±12
HR score at transition
Median (IQR)
56%
(42%, 85%)
55%
(42%, 82%)
0.04
27%
(24%, 28%)
27%
(25%, 29%)
Paroxysmal AF before
transition
N (%)
205 (7%)
232 (8%)
−0.03
11 (9%)
11 (9%)
Lower rate limit, bpm
Median (IQR)
60 (60, 60)
60 (60, 60)
−0.01
60 (50, 60)
60 (50, 60)
Implant year
Median (IQR)
2009
(2007, 2010)
2009
(2007, 2010)
−0.01
2008
(2007, 2009)
2008
(2007, 2009)
LV pacing, %
Median (IQR)
98.9%
(95.9%, 99.8%)
99.0%
(96.9%, 99.8%)
−0.07
96.8%
(91.9%, 99.4%)
98.2%
(94.7%, 99.5%)
Time to LATITUDE
set-up, months
Median (IQR)
1.3 (0.7, 3.1)
1.4 (0.7, 3.4)
−0.07
1.3 (0.8, 2.7)
1.3 (0.9, 2.4)
Characteristic
Downloaded from http://circep.ahajournals.org/ by guest on May 12, 2017
(Continued )
the mean HRSc after DDDR programming was still high at
78%, reflecting the possibility for even more improvement
with optimization of rate response. The changes in HRSc
in individual patients varied from a decrease of 69% to an
increase of 25%. In 28% of patients, the HRSc transitioned to
a lower risk zone of <70% after DDDR programming.
The HRSc increased for those programed DDDR with
a baseline HRSc <70% but decreased for those with a HRSc
≥70%. Although statistical regression to the mean is a remote
possibility, this is unlikely in such a large population followed
long term for the following reasons: the differences were highly
statistically significant, and, yet, the baseline measurements were
stable and were made over a long period of time with many beats
recorded. It would, therefore, be highly unusual, but not totally
inconceivable, that further HRSc measurements would show a
fluctuation, and regression to the mean only with rate response
turned on and, for the group with higher HRSc, that, with rate
response turned on, HRS would drop. Additionally, there were
small changes in a matched group continuing as DDD programming (Figure 3). Therefore, majority of the change in HRSc
does not seem consistent with regression to the mean.
We evaluated percent biventricular pacing in both groups
and found that programming rate response on did not affect
the percent RV/LV pacing in the DDDR versus DDD group
(Table). Thus, it was unlikely that changes in biventricular pacing affected outcomes. We considered the lower rate limit in
both groups and found that there were no significant differences in lower rate limit values in the DDD and DDDR groups.
Discussion
We show, for the first time, that rate-responsive dual-chamber
(DDDR) programming in CRT-D patients is associated with
significant improvement in survival but only for a select highrisk population identified by a novel parameter, the HRSc.
Figure 3. Change in heart rate score (HRSc)
by pacing mode. The absolute change in
HRSc (%) is plotted against the baseline
HRSc for patients programmed DDD only
(DDD) and those initially programmed
DDD and then reprogrammed to DDDR
(DDD→DDDR). The greatest reduction in
HRSc with programming DDDR (−13.2%),
occurs in patients with the highest baseline
DDD HRSc (91%–100%). Note the increase
in HRSc with DDDR programming (9.7%)
seen in patients with baseline HRSc in DDD
<30%. DDD indicates nonrate-responsive
pacing; and DDDR, rate-responsive pacing.
5 Olshansky et al Heart Rate Score Is Associated With Survival
Table. Continued
HR Score ≥70%
HR Score 30% to 69%
Downloaded from http://circep.ahajournals.org/ by guest on May 12, 2017
Standardized
Difference*
DDD→DDDR
(N=1811)
DDD Only
(N=1811)
Standardized
Difference*
DDD→DDDR
(N=1154)
DDD Only
(N=1154)
Standardized
Difference*
0.00
616 (34%)
616 (34%)
0.00
277 (24%)
273 (24%)
0.01
−0.00
69±11
69±10
−0.01
72±9
73±9
−0.04
−0.17
46% (39%, 54%)
45% (39%, 54%)
0.03
90% (81%, 97%)
87% (80%, 95%)
0.19
0.00
115 (6%)
114 (6%)
0.00
79 (7%)
107 (9%)
−0.09
0.01
60 (50, 60)
60 (50, 60)
0.01
60 (60, 70)
60 (60, 70)
−0.04
0.00
2008 (2007, 2010)
2008 (2007, 2010)
−0.01
2009 (2008, 2010)
2009 (2008, 2010)
−0.02
−0.27
98.8%
(95.5%, 99.7%)
98.9%
(96.9%, 99.8%)
−0.10
99.1%
(96.8%, 99.8%)
99.1%
(97.0%, 99.8%)
−0.00
0.00
1.4 (0.8, 3.4)
1.4 (0.8, 3.4)
−0.02
1.1 (0.6, 2.6)
1.4 (0.7, 3.6)
−0.14
AF indicates atrial fibrillation; DDD, nonrate-responsive pacing; DDDR, rate-responsive pacing; HR, heart rate; HRSc, heart rate score; IQR, interquartile range;
and LV, left ventricle.
*Standardized Difference=difference in means or proportions divided by standard error. Imbalance defined as absolute value >0.20 (small imbalance, 0.2–0.49;
medium imbalance, 0.5–0.79; large imbalance, ≥0.8).
Prior Clinical Trials Examining Outcomes in
Relationship to Pacing Modes
An assessment of potential benefit of DDDR logically is
confined to those pacemaker patients who have chronotropic
incompetence (ie, sick sinus syndrome). Unfortunately, no
gold standard determines exactly what chronotropic incompetence is, and no tool has been developed yet to determine
who will improve with rate-response programming. When
considering outcomes of pacing in sick sinus syndrome, 10
retrospective studies showed a 50% reduction in mortality
with dual-chamber versus single-chamber pacing.14
The Danish trial of Anderson et al stands out as the only
randomized controlled clinical trial showing a significant
reduction in mortality, albeit at a study duration longer than
the predetermined period.15 This study is remarkable because
all 225 patients had sinus node dysfunction and were randomized to atrial or ventricular pacing. Unnecessary right ventricular pacing was essentially eliminated in the atrial pacing group,
whereas there was significant ventricular pacing in the VVIR
(single-chamber, ventricular-paced, rate-responsive) group. The
difference in unnecessary right ventricular pacing likely affected
mortality differences. Furthermore, because all patients received
rate response, this study cannot assess the value of rate response.
The INTRINSIC RV trial (Inhibition of Unnecessary RV
Pacing With AV Search Hysteresis in ICDs) compared DDDR-60–
130 to VVI-40 (essentially intrinsic rate), with little unnecessary
Figure 4. Change in heart rate score (HRSc)
by baseline HRSc <30%, 30% to 69%, and
≥70%. A significant decrease in HRSc with
rate-responsive pacing (DDDR) programming was only seen in the patients with
baseline nonrate-responsive pacing (DDD)
HRSc ≥70%. These differences in response
to DDDR programming formed the basis for
the grouping.
6 Olshansky et al Heart Rate Score Is Associated With Survival
Downloaded from http://circep.ahajournals.org/ by guest on May 12, 2017
Figure 5. Survival from reprogramming by baseline
heart rate score (HRSc) group. A, Baseline HRSc
≥70%. Survival(%) is plotted over 4 years followup with statistically significant greater survival in
the DDDR group. B, Baseline HRSc 30% to 69%.
No survival benefit of DDDR programming. C,
Baseline HRSc <30%. No significant differences in
survival. AF indicates atrial fibrillation; DDD, nonrate-responsive pacing; DDDR, rate-responsive
pacing; and HR, heart rate.
right ventricular pacing in an ICD population not targeted to
have chronotropic incompetence.16 It showed a trend toward survival benefit with DDDR pacing. The PEGASUS trial (Pacing
Evaluation—Atrial Support Study in Cardiac Resynchronization
Therapy) found no benefit of atrial support pacing or rate-responsive pacing versus DDD-40 pacing in patients with CRT-D
implants.17 In fact, no randomized prospective trial concluded
that DDDR reduces mortality. However, appropriate patient
selection may determine who benefits and who does not.
Potential Reasons Why Prior Trials Did Not Show
Benefit With DDDR
No prior randomized, controlled, clinical trial has focused specifically on DDDR in the CRT-D population.18–20 Potential problems
7 Olshansky et al Heart Rate Score Is Associated With Survival
Figure 6. Multivariate survival model Cox proportional hazard model adjusted for age, sex, programmed lower rate limit, percent atrial
pacing, percent biventricular pacing, and implant year. Heart rate score (HRSc) ≥70% is associated with greater survival. CI indicates
confidence interval; DDD, nonrate-responsive pacing; DDDR, rate-responsive pacing; and HR, heart rate.
Downloaded from http://circep.ahajournals.org/ by guest on May 12, 2017
with the prior large, randomized, controlled, clinical trials that
have attempted to evaluate outcomes based on programming and
device characteristics regarding atrial-based and DDDR include
that they have not been able to factor out the impact of right ventricular pacing (DAVID I [The Dual Chamber and VVI Implantable Defibrillation]18) or have only considered nonrate-response
atrial pacing (DAVID II21). Unopposed right ventricular pacing
that had potentially confounded the analysis of some prior trials
(DAVID I) was not an issue in our study because all patients had a
high degree of biventricular pacing (Table). Further, no study has
yet considered targeting a high-risk heart failure group with chronotropic incompetence, in whom there is little heart rate variation
and for whom DDDR may help.
Too fast a heart rate could have an adverse effect.5 Based
on results from the INTRINSIC-RV Trial, higher mean heart
rates were associated with adverse outcomes in patients undergoing ICD implant. However, transient increases in heart rate
from a lower baseline resting rate, as needed, were not evaluated and, based on our results shown here, may be beneficial.
For others, such as patients with robust spontaneous heart rate
variation and low HRScs, additional, and unnecessary, DDDR
may have adverse effects.
Possible Mechanisms Explaining Reduced Mortality
With DDDR Pacing and Decreased HRSc
Chronotropic incompetence has been shown to be highly
prevalent in the heart failure population with reduced ejection
fraction, with numbers ranging from 25% to 70%, using conventional definitions of reductions in rate response reserve.22
However, its contribution to cardiovascular morbidity and
mortality remain underappreciated, likely because of limitations in identification of chronotropic incompetence and of the
fundamental responsible mechanisms.
Recent elegant work by Benes et al has dissected out chronotropic incompetence into a resting heart rate and heart rate
reserve components in a small cohort of CRT-D patients. Fully
two thirds of their patients displayed chronotropic incompetence.23 A combined outcome of death, transplantation, or
urgent assist device was associated with less heart rate reserve,
a measure of chronotropic incompetence.23 Resting heart rate
and heart rate reserve were not correlated with each other
and were associated with distinct biomarker profiles. Both
higher resting heart rate and lower heart rate reserve identified
patients more likely to die or require transplantation.
Along these lines, the HRSc might identify patients with low
heart rate reserve who might benefit from DDDR. This is particularly attractive because both the method and the intervention to
improve mortality are achieved easily. This merits further study
in a prospective fashion. Likewise, a more physiological sensor, such as minute ventilation, may produce the greatest clinical
benefit in high HRSc patient and needs to be evaluated.
At best, however, our hypothesis can only be tested in
a trial of high-risk heart failure patients with chronotropic
incompetence in which patients are randomized to DDD or
DDDR pacing modes with follow-up data collected prospectively in a blinded fashion. From these present data, HRSc
may predict benefit from DDDR pacing, realizing that benefit
has many meanings, including survival, exercise tolerance,
and improved symptoms not measured here.
Limitations
These data were from a post hoc (retrospective) analysis of
data that were not collected in a manner that allows testing of
the central hypothesis that patients with chronotropic incompetence have improved survival, which is causally related to
rate-adaptive pacing. The results are associations observed
in large populations. The database does not contain patient
symptoms or quality-of-life, and, thus, they are not assessed.
Because the data were propensity-matched based on select
criteria, perhaps not all the necessary criteria were selected.
However, one would suspect that patients programmed DDDR
may be a high-risk population because of specific symptoms,
leading to the need for rate-response programming, and yet,
outcomes improved at least in the highest risk group. On the
contrary, this group programmed DDDR may be a more motivated group who were expecting a greater response from their
implanted device, and, therefore, they strove to derive all the
benefits they could, and, perhaps, they were a healthier lot.
Other trials using propensity-matched analyses have the same
conundrum. We did not expect to address all issues regarding why, or if, programming based on HRSc is the cause for
improved outcomes; this is just a first step.
We do not know the shortest length of time necessary to
get the most robust and reproducible HRSc. The relationships
between rate-response programming characteristics and outcomes in individual patients may be more challenging to obtain
than in a prospective trial. The mechanism of death is uncertain
in those who have high HRScs. The best method to program,
8 Olshansky et al Heart Rate Score Is Associated With Survival
and how to program, rate response (minute ventilation or accelerometer) is unknown. Medication details remain uncertain.
Several questions remain unanswered: Were doctors who
reprogrammed devices more adept at pacer management?
Were patients remaining with DDD programming felt to be less
likely to benefit from rate-adaptive pacing? Did patients more
demanding of improved effort tolerance, thereby prompting
their physicians to activate the DDDR mode? Could reprogramming to the DDDR mode simply have selected a group fated to
have intrinsically better survival? Although these questions are
germane, they are presently unanswerable. It is possible that
rate response was programmed on to a group fated to do better
anyway, but it is not clear a priori how one could determine that.
Future controlled studies may be able to answer this.
Conclusions
Downloaded from http://circep.ahajournals.org/ by guest on May 12, 2017
The HRSc predicts mortality in a large cohort of CRT-D
patients. Reducing the HRSc, through adoption of DDDR, is
associated with markedly decreased mortality in patients with
high HRSc and presumed chronotropic incompetence. These
are the first data that indicate a survival benefit with rateresponse programming when properly targeting a high-risk
CRT-D population. Because this report is observational, a prospective, randomized, controlled clinical trial will be needed
to confirm these initial observations.
Disclosures
Dr Olshansky has received honoraria and, thus, has the following to disclose: Biotronik, Boehringer-Ingleheim, Amarin, On-X,
and Daiichi Sankyo (all <$10 000) and Lundbeck (>$10 000). Dr
Richards has received honoraria and, thus, has the following to
disclose: (all<$10 000) Medtronic, Boston Scientific, Biotronik,
Janssen, Boehringer-Ingleheim, Bristol Myers Squibb, and Pfizer.
Dr Sharma has received salary and stock from Boston Scientific
(>$10 000). N. Wold has received salary and stock from Boston
Scientific (>$10 000). P. Jones has received salary and stock from
Boston Scientific (>$10 000). Dr Perschbacher has received salary
and stock from Boston Scientific (>$10 000). Dr Wilkoff has received
honoraria and, thus, has the following to disclose: St Jude Medical,
Boston Scientific, and Spectranetics (all <$10 000).
References
1. Bristow MR, Saxon LA, Boehmer J, Krueger S, Kass DA, De Marco
T, Carson P, DiCarlo L, DeMets D, White BG, DeVries DW, Feldman
AM; Comparison of Medical Therapy, Pacing, and Defibrillation in Heart
Failure (COMPANION) Investigators. Cardiac-resynchronization therapy
with or without an implantable defibrillator in advanced chronic heart failure. N Engl J Med. 2004;350:2140–2150. doi: 10.1056/NEJMoa032423.
2. Cazeau S, Leclercq C, Lavergne T, Walker S, Varma C, Linde C, Garrigue
S, Kappenberger L, Haywood GA, Santini M, Bailleul C, Daubert JC;
Multisite Stimulation in Cardiomyopathies (MUSTIC) Study Investigators.
Effects of multisite biventricular pacing in patients with heart failure and
intraventricular conduction delay. N Engl J Med. 2001;344:873–880. doi:
10.1056/NEJM200103223441202.
3. Cleland JG, Daubert JC, Erdmann E, Freemantle N, Gras D, Kappenberger
L, Tavazzi L; Cardiac Resynchronization-Heart Failure (CARE-HF)
Study Investigators. The effect of cardiac resynchronization on morbidity
and mortality in heart failure. N Engl J Med. 2005;352:1539–1549. doi:
10.1056/NEJMoa050496.
4. Abraham WT, Fisher WG, Smith AL, Delurgio DB, Leon AR, Loh E,
Kocovic DZ, Packer M, Clavell AL, Hayes DL, Ellestad M, Trupp RJ,
Underwood J, Pickering F, Truex C, McAtee P, Messenger J; MIRACLE
Study Group. Multicenter InSync Randomized Clinical Evaluation. Cardiac
resynchronization in chronic heart failure. N Engl J Med. 2002;346:1845–
1853. doi: 10.1056/NEJMoa013168.
5. Ahmadi-Kashani M, Kessler DJ, Day J, Bunch TJ, Stolen KQ, Brown S, Sbaity
S, Olshansky B; INTRINSIC RV Study Investigators. Heart rate predicts outcomes in an implantable cardioverter-defibrillator population. Circulation.
2009;120:2040–2045. doi: 10.1161/CIRCULATIONAHA.108.847608.
6. Böhm M, Swedberg K, Komajda M, Borer JS, Ford I, Dubost-Brama A,
Lerebours G, Tavazzi L; SHIFT Investigators. Heart rate as a risk factor in
chronic heart failure (SHIFT): the association between heart rate and outcomes in a randomised placebo-controlled trial. Lancet. 2010;376:886–
894. doi: 10.1016/S0140-6736(10)61259-7.
7. Richards M, Olshansky B, Sharma A, Wold N, Jones P, Perschbacher D,
Wilkoff B. Heart rate score, an easy and intuitive measure of chronotropic
incompetence in device patients. Heart Rhythm. 2014;11:S69 AB29-05.
8. Wilkoff B, Richards M, Sharma A, Wold N, Jones P, Perschbacher D,
Olshansky B. A novel simple predictor of mortality risk in ICD and CRT-D
patients: the heart rate score. Heart Rhythm. 2014;11:S467 PO05–141.
9. Cohen J. The statistical power of abnormal-social psychological research:
a review. J Abnorm Soc Psychol. 1962;65:145–153.
10. Schacht A, Bogaerts K, Bluhmki E, Lesaffre E. A new nonparametric approach for baseline covariate adjustment for two-group comparative studies.
Biometrics. 2008;64:1110–1116. doi: 10.1111/j.1541-0420.2008.00994.x.
11. Klein JP, Moeschberger ML. Survival Analysis: Techniques for Censored
and Truncated Data. New York, NY: Springer-Verlag; 1997.
12. Austin PC. The use of propensity score methods with survival or time-toevent outcomes: reporting measures of effect similar to those used in randomized experiments. Stat Med. 2014;33:1242–1258. doi: 10.1002/sim.5984.
13. Austin PC. An introduction to propensity score methods for reducing the
effects of confounding in observational studies. Multivariate Behav Res.
2011;46:399–424. doi: 10.1080/00273171.2011.568786.
14. Lauer MS, Francis GS, Okin PM, Pashkow FJ, Snader CE, Marwick TH.
Impaired chronotropic response to exercise stress testing as a predictor of
mortality. JAMA. 1999;281:524–529.
15.Andersen HR, Thuesen L, Bagger JP, Vesterlund T, Thomsen PE.
Prospective randomised trial of atrial versus ventricular pacing in sicksinus syndrome. Lancet. 1994;344:1523–1528.
16. Olshansky B, Day JD, Moore S, Gering L, Rosenbaum M, McGuire M,
Brown S, Lerew DR. Is dual-chamber programming inferior to singlechamber programming in an implantable cardioverter-defibrillator?
Results of the INTRINSIC RV (Inhibition of Unnecessary RV Pacing
With AVSH in ICDs) study. Circulation. 2007;115:9–16. doi: 10.1161/
CIRCULATIONAHA.106.629428.
17. Martin DO, Day JD, Lai PY, Murphy AL, Nayak HM, Villareal RP, Weiner
S, Kraus SM, Stolen KQ, Gold MR. Atrial support pacing in heart failure: results from the multicenter PEGASUS CRT trial. J Cardiovasc Electrophysiol.
2012;23:1317–1325. doi: 10.1111/j.1540-8167.2012.02402.x.
18. Wilkoff BL, Cook JR, Epstein AE, Greene HL, Hallstrom AP, Hsia H,
Kutalek SP, Sharma A; Dual Chamber and VVI Implantable Defibrillator
Trial Investigators. Dual-chamber pacing or ventricular backup pacing
in patients with an implantable defibrillator: the Dual Chamber and VVI
Implantable Defibrillator (DAVID) Trial. JAMA. 2002;288:3115–3123.
19. Link MS, Estes NA III, Griffin JJ, Wang PJ, Maloney JD, Kirchhoffer JB,
Mitchell GF, Orav J, Goldman L, Lamas GA. Complications of dual chamber
pacemaker implantation in the elderly. Pacemaker Selection in the Elderly
(PASE) Investigators. J Interv Card Electrophysiol. 1998;2:175–179.
20. Healey JS, Toff WD, Lamas GA, Andersen HR, Thorpe KE, Ellenbogen
KA, Lee KL, Skene AM, Schron EB, Skehan JD, Goldman L, Roberts RS,
Camm AJ, Yusuf S, Connolly SJ. Cardiovascular outcomes with atrialbased pacing compared with ventricular pacing: meta-analysis of randomized trials, using individual patient data. Circulation. 2006;114:11–17.
doi: 10.1161/CIRCULATIONAHA.105.610303.
21. Wilkoff BL, Kudenchuk PJ, Buxton AE, Sharma A, Cook JR, Bhandari
AK, Biehl M, Tomassoni G, Leonen A, Klevan LR, Hallstrom AP;
DAVID II Investigators. The DAVID (Dual Chamber and VVI Implantable
Defibrillator) II trial. J Am Coll Cardiol. 2009;53:872–880. doi: 10.1016/j.
jacc.2008.10.057.
22. Dobre D, Zannad F, Keteyian SJ, Stevens SR, Rossignol P, Kitzman DW,
Landzberg J, Howlett J, Kraus WE, Ellis SJ. Association between resting
heart rate, chronotropic index, and long-term outcomes in patients with
heart failure receiving β-blocker therapy: data from the HF-ACTION trial.
Eur Heart J. 2013;34:2271–2280. doi: 10.1093/eurheartj/ehs433.
23. Benes J, Kotrc M, Borlaug BA, Lefflerova K, Jarolim P, Bendlova B, Jabor
A, Kautzner J, Melenovsky V. Resting heart rate and heart rate reserve in
advanced heart failure have distinct pathophysiologic correlates and prognostic impact: a prospective pilot study. JACC Heart Fail. 2013;1:259–
266. doi: 10.1016/j.jchf.2013.03.008.
Survival After Rate-Responsive Programming in Patients With Cardiac
Resynchronization Therapy-Defibrillator Implants Is Associated With a Novel Parameter:
The Heart Rate Score
Brian Olshansky, Mark Richards, Arjun Sharma, Nicholas Wold, Paul Jones, David
Perschbacher and Bruce L. Wilkoff
Downloaded from http://circep.ahajournals.org/ by guest on May 12, 2017
Circ Arrhythm Electrophysiol. 2016;9:
doi: 10.1161/CIRCEP.115.003806
Circulation: Arrhythmia and Electrophysiology is published by the American Heart Association, 7272 Greenville
Avenue, Dallas, TX 75231
Copyright © 2016 American Heart Association, Inc. All rights reserved.
Print ISSN: 1941-3149. Online ISSN: 1941-3084
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circep.ahajournals.org/content/9/8/e003806
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation: Arrhythmia and Electrophysiology can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for which
permission is being requested is located, click Request Permissions in the middle column of the Web page
under Services. Further information about this process is available in the Permissions and Rights Question and
Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation: Arrhythmia and Electrophysiology is online at:
http://circep.ahajournals.org//subscriptions/