Importance of Adjunctive Heart Failure Optimization Immediately
After Implantation to Improve Long-Term Outcomes With Cardiac
Resynchronization Therapy
Wilfried Mullens, MD, PhD
a,b,
*, Jacek Kepa, MS
b
, Philippe De Vusser, MD
a
, Jan Vercammen, RN
a
,
Maximo Rivero-Ayerza, MD, PhD
a
, Patrick Wagner, PhD
b
, Joseph Dens, MD, PhD
a
,
Mathias Vrolix, MD
a
, Pieter Vandervoort, MD
a
, and W.H. Wilson Tang, MD
c
Despite improvement in morbidity and mortality with cardiac resynchronization therapy
(CRT), disease progression continues to affect a subset of patients and there is limited effort
to identify contributing factors. Our objective was to investigate if a protocol-driven
approach incorporated in a management strategy of heart failure immediately after im-
plantation would provide incremental benefits beyond usual care after implantation. We
reviewed 114 consecutive patients with CRT implanted from 2005 through 2009 who
received usual care after implantation or underwent protocol-driven CRT care after
implantation. Preimplantation characteristics in patients receiving usual versus protocol-
driven care were similar in left ventricular (LV) dimension (LV internal diastolic diameter
6.2 ⴞ 0.8 vs 6.4 ⴞ 1.0 cm), LV ejection fraction (26 ⴞ 8% vs 25% ⴞ 8%), QRS width, and
medication usage. Major adjustments during the protocol-driven approach were uptitration
of neurohormonal blockers (64%), echocardiographically guided atrioventricular optimi-
zation (50%), heart failure education (42%), arrhythmia management (19%), and LV lead
repositioning (7%). Although positive LV remodeling was noted in the 2 groups at 6
months, extent was significantly greater in the protocol-driven approach compared to usual
care (change in LV internal diastolic diameter 0.7 ⴞ 0.6 cm vs 0.2 ⴞ 1.2 cm, p ⴝ 0.01;
change in LV ejection fraction 11 ⴞ 7% vs 7 ⴞ 9%, p ⴝ 0.01), which was associated with
fewer major adverse events (14% vs 53%, p <0.001). In conclusion, a protocol-driven
approach for patients with CRT started immediately after implantation is associated with
incremental favorable effects on reverse remodeling and fewer adverse events compared to
usual care after implantation. These effects appeared to be driven not only by changes in
device settings and arrhythmia management but also by concomitant medication optimi-
zation and heart failure education. © 2011 Elsevier Inc. All rights reserved. (Am J
Cardiol 2011;108:409 – 415)
Most therapeutic effects of cardiac resynchronization
therapy (CRT) have been attributed to device-induced de-
crease in dyssynchrony. For those who do not demonstrate
noticeable improvements immediately after implantation,
further optimization of specific device programming, ar-
rhythmia management, and considerations for lead reposi-
tioning, if suboptimal, can be beneficial.
1
However, it is
often assumed that only those who did not respond over
time would benefit from such optimization strategies be-
cause the maximal therapeutic effect of CRT should be
derived if implantation and device programming are opti-
mal. Factors favorably or adversely affecting CRT efficacy,
particularly those in the immediate postimplantation setting,
have not been extensively studied. Therefore, the objective
of this study was to investigate if a protocol-driven ap-
proach incorporated in a management strategy of heart fail-
ure immediately after implantation would provide incre-
mental benefits beyond usual care after implantation
including changes in exercise and echocardiographic pa-
rameters and long-term adverse cardiac events.
Methods
This study included consecutive patients with CRT using
an implanted pacemaker or defibrillator under standard clin-
ical indications in a single tertiary cardiac care institution
from November 2005 through February 2010. All patients
demonstrated stable but advanced heart failure symptoms
(New York Heart Association functional class III or IV
symptoms) despite receiving optimal medical therapy as
tolerated by the patient, decreased left ventricular (LV)
ejection fraction (ⱕ35%), and prolonged QRS duration
(ⱖ120 ms) at time of implantation. Because of stringent
reimbursement criteria in Belgium, CRT using implanted
defibrillators was performed only cases of previous episodes
of sustained ventricular tachycardia or inducible ventricular
a
Department of Cardiovascular Medicine, Ziekenhuis Oost Limburg,
Genk, Belgium;
b
School of Life Sciences, Transnational University Lim-
burg, Belgium;
c
Department of Cardiovascular Medicine, Heart and Vas-
cular Institute, Cleveland Clinic, Cleveland, Ohio. Manuscript received
January 12, 2011; revised manuscript received and accepted March 22,
2011.
*Corresponding author: Tel: 32-8932-7087; fax: 32-8932-7918.
E-mail address: wilfried.mullens@zol.be (W. Mullens).
0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. www.ajconline.org
doi:10.1016/j.amjcard.2011.03.060
arrhythmia. The study complied with the Declaration of
Helsinki, the locally appointed ethics committee approved
the research protocol, and informed consent was obtained
from the subjects.
A protocol-driven CRT optimization protocol was estab-
lished in August 2008; thus patients were stratified in 2
groups according to CRT implantation before August 2008
(“usual care”) versus after August 2008 (“CRT optimization
protocol”). Before August 2008 patients received usual care
after implantation performed generally by different hospital
staffs. In this model, an electrophysiology nurse performed
a standard device check. Afterward a patient was seen by a
treating cardiologist to assess the patient’s current health
status and symptoms, often also performing echocardiogra-
phy (Figure 1). Changes in device settings and heart failure
therapy were at the discretion of the treating cardiologist.
A postimplantation stepwise CRT optimization protocol
was established as part of a multidisciplinary approach
toward postimplantation CRT care incorporated in manage-
ment program of heart failure in August 2008. In the CRT
optimization protocol, number of scheduled clinic visits was
similar to that of the usual-care group (6 weeks, 3 and 6
months after implantation). Beyond the standard device
check, a more thorough CRT optimization clinic protocol
was conducted, which included a wider variety of measure-
ments and prespecified optimization guidelines performed
in a designated clinic staffed with physicians and nurses
with a broad interest in heart failure and cardiac devices as
previously described (Figure 2).
1
In summary, a heart fail-
ure nurse recorded an electrocardiogram to assess heart rate,
QRS width, and AV/PR intervals. This was performed 2
times with the implanted pacemaker turned on and off to
ensure adequate biventricular pacing. Next, an anterior–
posterior and lateral chest x-ray was carried out to deter-
mine optimal positioning of the right atrial, right ventricu-
lar, and LV leads (in basal or midlateral and posterior
position). In the meantime, routine laboratory tests were
done to detect occult hematologic and metabolic derange-
ments. After these measurements, the designated cardiolo-
gist recorded a detailed history on heart failure symptoms,
occurrence of arrhythmias, and potential device-related is-
sues, checked for compliance to medication usage and salt/
fluid restriction, and completed a full physical cardiovascu-
lar examination.
Afterward a comprehensive 2-dimensional echocardio-
graphic examination was performed (Philips Medical Sys-
tems, Andover, Massachusetts) with nominal settings of the
CRT device. All reported echocardiographic measurements
including LV size/function and mitral regurgitation were
averaged from ⱖ3 consecutive cycles as recommended by
the American Society of Echocardiography.
2
Then an effort
was made to optimize LV diastolic filling when it differed
from stage I by altering AV timing using conventional
Doppler echocardiography. Optimal AV interval was deter-
mined by sampling mitral inflow with pulse-wave Doppler
to correspond with the shortest AV interval that dissociated
the E and A waves but did not interrupt the end of the A
wave.
3–5
To evaluate a patient’s physical fitness and ensure biven-
tricular pacing was persistent even during exercise, a cyc-
loergometric bicycle test with maximum oxygen uptake
recording was performed. Based on the findings, a recom-
mendation was proposed to the patient to maximize the
potential of CRT. These recommendations were not mutu-
ally exclusive because actions could be categorized by re-
positioning of the LV lead to correct inappropriate lead
positioning, changes in device programming for suboptimal
device programming (mostly AV timing), or treatment of
arrhythmias medically or invasively.
Thorough efforts were taken toward optimization of
medical therapy, i.e., uptitration of neurohormonal blockers
to guideline-recommended doses, which were often not
tolerated before implantation. In addition, adequate heart
failure education was provided to familiarize a patient with
heart failure risk factors and modifications in lifestyle. Patients
were informed through dietary consults about salt-free diets (2
to 3 g/day) and fluid restriction (1 to 1.5 L/day), which often
coincided with a progressive decrease in loop diuretic doses.
Importantly, all these adjustments were implemented in close
collaboration with general practitioners who were informed
through telephone contact the day of the patient’s clinic visit
and provided with the findings and recommendations of the
clinic through an on-line letter sent immediately after the CRT
clinic visit to ensure optimization of medical therapy was
accomplished at home under close supervision.
We prespecified the primary end points for analysis as
time to first occurrence of any of the following: all-cause
mortality, cardiac transplantation, and/or first readmission
Figure 1. Flow chart presenting usual care after implantation.
410 The American Journal of Cardiology (www.ajconline.org)
for heart failure after implantation. Patients in the usual-care
group were followed from implantation to the date of the
first visit to the protocol-driven care clinic, whereas patients
in the protocol-driven group were followed until April 30,
2010.
Collected data are expressed as mean ⫾ SD for contin-
uous data and as ratio for categorical data. Paired-sample t
tests were performed for variables between related patient
data groups and independent-sample t tests were performed
for variables between unrelated patient data groups. Statis-
tical significance was set at a 2-tailed probability level with
alpha equal to 0.05. Kaplan–Meier survival curves were
calculated with combined end points for all patients strati-
fied in 2 groups. Cox proportional hazards regression model
was used to determine which variables were related signif-
icantly to the different end point during the follow-up pe-
riod. The authors had full access to the data and take
responsibility for the integrity of the data. All authors have
read and agreed to the report as written. All statistical
analyses were performed using SPSS 17.0 for Windows
(SPSS, Inc., Chicago, Illinois).
Results
Preimplantation patient characteristics are presented in
Table 1 and were similar in the 2 groups (n ⫽ 53 in
usual-care group and 61 in protocol-driven group) including
degree of LV remodeling and maximum exercise capacity.
There was a large and similar proportion of use of neuro-
hormonal blockers. For the CRT optimization group, mean
clinic visit duration was 40 minutes with involvement of a
designated nurse (⫾20 minutes) and a cardiologist (⫾20
minutes).
All patients had a lead implanted in the right atrium, right
ventricle, and left ventricle through the coronary sinus
(72%) or epicardially (28%). X-rays demonstrated no lead
dislodgement but indicated a suboptimal positioning of the
LV lead for 8 patients. One patient was scheduled for LV
lead repositioning.
Device interrogation was successful in all patients,
which was paced mostly in an atrial sensing–ventricu-
lar pacing mode. No battery depletion or lead integrity
was noted.
Figure 2. Flow chart presenting protocol-driven postimplantation management in cardiac resynchronization therapy clinic. AV opt. ⫽ atrioventricular
optimization; biv. ⫽ biventricular; ECG ⫽ electrocardiography; ECHO ⫽ echocardiography; EF ⫽ ejection fraction; LVIDd ⫽ left ventricular internal
diastolic diameter.
411Heart Failure/Protocol-Driven Cardiac Resynchronization Clinic
Arrhythmias were present in 19% of patients, mostly
atrial fibrillation, but in 7% frequent ventricular ectopy was
present, leading to ⬍100% biventricular pacing in 19% and
⬍90% in 17% of patients. All arrhythmias were treated
accordingly at least to ensure ⬎90% biventricular pacing in
95% instead of 83% of patients.
An additional 50% of patients were found to be pro-
grammed with suboptimal AV timing settings. AV timings
were always optimized in these patients after an improve-
ment in LV filling. These improvements were confirmed
during the next clinic visit, with only 2 patients needing an
additional change in their AV timings.
Uptitration in neurohormonal blockers was possible for
64% of patients, although ⬎90% were already receiving
angiotensin-converting enzyme inhibitor and

-blocker
medication at time of implantation. Interestingly, all these
patients were already taking a similar dosage of neurohor-
monal blockers for ⬎3 months before device implantation
and only noted an improved tolerance toward uptitration of
these drugs after implantation (Figure 3).
Failure in patient compliance with regard to salt/water
restriction and stringent intake of medication was high be-
cause 42% patients testified to having poorly followed their
daily medication intake and dietary advice. After implemen-
tation of heart failure education and dietary consult, in
collaboration with general practitioners 22% of patients had
a decrease in dosage of loop diuretic (Figure 4).
In addition, 63% patients were obese, 13% of patients
had anemia, but only 1 patient had hemoglobin ⬍10 g/dl,
which was treated with transfusion or erythropoietin agents.
Starting from August 2008 all patients receiving usual
care were also referred to the protocol-driven CRT clinic.
Interestingly, clinical, electrophysiologic, and device-re-
lated interventions were similar overall compared to pa-
tients followed in the protocol-driven clinic immediately
after implantation. Indeed, uptitration of neurohormonal
Table 1
Baseline demographics
Total Usual Care Protocol-Driven p Value
(n ⫽ 114) (n ⫽ 51) (n ⫽ 63)
Demographics
Age (years) 71 ⫾ 10 72 ⫾ 10 71 ⫾ 11 NS
Men 64% 55% 73% 0.04
Cardiac resynchronization therapy with defibrillator 51% 52% 48% NS
Weight (kg) 77 ⫾ 17 76 ⫾ 16 79 ⫾ 17 NS
Body mass index (kg/m
2
) 28 ⫾ 527⫾ 528⫾ 5NS
⬎25 73% 70% 76% NS
⬎30 33% 30% 36% NS
Hypertension (⬎140/90 mm Hg) 42% 40% 43% NS
Hyperlipidemia (low-density lipoprotein ⬎110 mg/dl) 33% 28% 37% NS
Quit smoking 20% 21% 18% NS
Active smoking 10% 12% 9% NS
Diabetes mellitus 30% 30% 30% NS
Atrial fibrillation 42% 38% 44% NS
Medications
Aspirin 76% 76% 75% NS
Angiotensin-converting enzyme inhibitors and angiotensin receptor blockers 88% 84% 90% NS

Blockers 89% 82% 94% NS
Spironolactone 58% 39% 73% 0.001
Loop diuretic 75% 84% 67% 0.02
Statin 59% 55% 62% NS
Hydralazine 9% 2% 14% 0.02
Isosorbide dinitrate 6% 2% 10% NS
Digoxin 24% 24% 24% NS
Electrocardiographic data
Heart rate (beat/min) 72 ⫾ 21 75 ⫾ 19 70 ⫾ 22 NS
PR width (ms) 192 ⫾ 47 192 ⫾ 48 191 ⫾ 44 NS
QRS width (ms) 159 ⫾ 31 155 ⫾ 31 160 ⫾ 31 NS
Cycloergometric data
Maximum exercise capacity (watts) 89 ⫾ 32 84 ⫾ 22 90 ⫾ 34 NS
Systolic/diastolic blood pressure at rest (mm Hg) 126/74 115/71 127/74 NS
Systolic/diastolic blood pressure during exercise (mm Hg) 151/74 145/67 153/75 NS
Maximum heart rate (beats/min) 112 ⫾ 23 113 ⫾ 18 112 ⫾ 25 NS
Maximum volume (ml/kg/min) 13.9 ⫾ 4.6 12.4 ⫾ 3.2 13.9 ⫾ 4.7 NS
Echocardiographic data
Left ventricular internal diastolic diameter (cm) 6.3 ⫾ 1 6.2 ⫾ 0.8 6.4 ⫾ 1NS
Left ventricular ejection fraction (%) 25 ⫾ 826⫾ 825⫾ 8NS
Mitral valve regurgitation (grade ⬎II) 10% 11% 10% NS
Tricuspid valve regurgitation (grade ⬎II) 3% 2% 3% NS
412 The American Journal of Cardiology (www.ajconline.org)
drugs was possible for 1/2 the patient population, decrease
in dosage of loop diuretics for 20% of patients, and 43% of
patients confirmed to having poorly maintained their med-
ication and dietary guidelines. In addition, 51% of patients
were shown to be paced with suboptimal AV timings and
20% of patients presented with arrhythmias.
Positive remodeling was noted in the 2 groups with regard
to LV dimension and LV function. However, extent of positive
LV remodeling and improvement in LV ejection fraction were
significantly greater in the group receiving a protocol-driven
approach from the start (Figure 5). Moreover, as presented in
Table 2, improvement in maximum exercise capacity as mea-
sured through maximum oxygen consumption was more pro-
nounced for the protocol-driven group. Interestingly, these
positive effects did not seem to relate to a greater decrease in
dyssynchrony because the decrease in QRS and PR times did
not differ between groups.
At the end of the follow-up period (mean follow-up dura-
tion 19 ⫾ 11 months for the 2 groups), 36% of patients had
died, undergone cardiac transplantation, and/or were hospital-
ized for decompensated heart failure. Although overall mor-
tality/cardiac transplantation was similar for the 2 groups (3 vs
4 events, p ⫽ 1), patients receiving protocol-driven care had
Figure 3. Interventions performed during visit to protocol-driven optimization cardiac resynchronization therapy clinic. RV ⫽ right ventricular; Subopt ⫽
suboptimal. Other abbreviations as in Figure 2.
Figure 4. Bar graph presenting percentage of patients whose medication
was optimized in the protocol-drive optimization cardiac resynchronization
therapy clinic. Ace ⫽ angiotensin-converting enzyme.
Figure 5. Bar graph for left ventricular internal diameter in diastole (cen-
timeters) and left ventricular ejection fraction (percentage) at time of
implantation versus follow-up for the 2 study groups. Abbreviations as in
Figure 2.
413Heart Failure/Protocol-Driven Cardiac Resynchronization Clinic
