1. No category

Clinical profile and prognostic value of low systolic blood
pressure in patients hospitalized for heart failure with
reduced ejection fraction: Insights from the Efficacy of
Vasopressin Antagonism in Heart Failure: Outcome
Study with Tolvaptan (EVEREST) trial
Andrew P. Ambrosy, MD, a,i Muthiah Vaduganathan, MD, MPH, b,i Robert J. Mentz, MD, c,i Stephen J. Greene, MD, d,i
Haris Subačius, MA, e,i Marvin A. Konstam, MD, f,i Aldo P. Maggioni, MD, g,i Karl Swedberg, MD, h,i and
Mihai Gheorghiade, MD d,i Stanford, CA; Boston, MA; Durham, NC; Chicago, IL; Florence, Italy; and Göteborg, Sweden
rC
DR
Background Systolic blood pressure (SBP) is related to the pathophysiologic development and progression of heart
failure (HF) and is inversely associated with adverse outcomes during hospitalization for HF (HHF). The prognostic value of SBP
after initiating inhospital therapy and the mode of death and etiology of cardiovascular readmissions based on SBP have not
been well characterized in HHF.
Methods
or
ia
za
da
po
A post hoc analysis was performed of the placebo group (n = 2061) of the EVEREST trial, which enrolled patients
within 48 hours of admission for worsening HF with an ejection fraction (EF) ≤40% and an SBP ≥90 mm Hg, for a median followup of 9.9 months. Systolic blood pressure was measured at baseline, daily during hospitalization, and at discharge/day 7.
Patients were divided into the following quartiles by SBP at baseline: ≤105, 106 to 119, 120 to 130, and ≥131 mm Hg.
Outcomes were all-cause mortality (ACM) and the composite of cardiovascular mortality or HHF (CVM + HHF). The associations
between baseline, discharge, and inhospital change in SBP and ACM and CVM + HHF were assessed using multivariable Cox
proportional hazards regression models adjusted for known covariates.
Co
pi
aa
ut
Results Median (25th, 75th) SBP at baseline was 120 (105, 130) mm Hg and ranged from 82 to 202 mm Hg. Patients
with a lower SBP were younger and more likely to be male; had a higher prevalence of prior revascularization and ventricular
arrhythmias; had a lower EF, worse renal function, higher natriuretic peptide concentrations, and wider QRS durations; and
were more likely to require intravenous inotropes during hospitalization. Lower SBP was associated with increased mortality,
driven by HF and sudden cardiac death, and cardiovascular hospitalization, primarily caused by HHF. After adjusting for
potential confounders, SBP was inversely associated with risk of the coprimary end points both at baseline (ACM: hazard ratio
[HR]/10-mm Hg decrease 1.15, 95% CI1.08-1.22; CVM + HHF: HR 1.09/10-mm Hg decrease, 95% CI 1.04-1.14) and at
the time of discharge/day 7 (ACM: HR 1.15/10-mm Hg decrease, 95% CI 1.08-1.22; CVM + HHF: HR 1.07/10-mm Hg
decrease, 95% CI 1.02-1.13), but the association with inhospital SBP change was not significant.
Conclusion
Systolic blood pressure is an independent clinical predictor of morbidity and mortality after initial therapy
during HHF with reduced EF. (Am Heart J 2013;165:216-25.)
Heart failure (HF) is a public health problem of
pandemic proportions with a prevalence of 5.7 million,
incidence of 670,000 new cases/y, and an annual
hospitalization for HF (HHF) rate in excess of 1 million in
the United States alone. 1,2 Blood pressure (BP) is a
complex and dynamic clinical variable that must be
From the aDepartment of Medicine, Stanford University School of Medicine, Stanford, CA,
b
Department of Medicine, Massachusetts General Hospital, Boston, MA, cDivision of
Cardiology, Duke University Medical Center, Durham, NC, dCenter for Cardiovascular
Randomized controlled trial registration no. NCT00071331.
Javed Butler, MD, MPH, served as guest editor for this article.
Submitted September 27, 2012; accepted November 26, 2012.
Innovation, Northwestern University Feinberg School of Medicine, Chicago, IL, eDivision of
Cardiology, Department of Medicine, Northwestern University Feinberg School of
Medicine, Chicago, IL, fDivision of Cardiology, Department of Medicine, Tufts Medical
Center, Boston, MA, gANMCO Research Center, Florence, Italy, and hDepartment of Molecular
and Clinical Medicine, Sahlgrenska Academy, University of Gothenburg, Göteborg, Sweden.
j
On behalf of the EVEREST trial investigators.
Reprint requests: Mihai Gheorghiade, MD, Center for Cardiovascular Innovation, Northwestern
University Feinberg School of Medicine, 645 North Michigan Ave, Suite 1006, Chicago, IL 60611.
E-mail: [email protected]
0002-8703/$ - see front matter
© 2013, Mosby, Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ahj.2012.11.004
04/04/2014
American Heart Journal
Volume 165, Number 2
Ambrosy et al 217
with the Declaration of Helsinki, the protocol was independently approved by the institutional review board or ethics
committee at each participating center, and written informed
consent was obtained from all participants. The present analysis
includes only those patients randomized to the placebo arm of
the EVEREST trial because of the unclear effects of the study
drug on SBP.
Blood pressure measurements and laboratory testing
rC
DR
Blood pressure measurements were taken in a supine position.
Baseline SBP was defined as SBP at the time of randomization.
For patients with missing baseline SBP measurements, the earliest
measurement recorded up to 4 days postenrollment was substituted
because very little change in SBP occurred after the first day. Patients
with missing baseline measurements (n = 9) were excluded from all
subsequent analyses. Discharge SBP was defined as SBP on
inhospital day 7 or discharge, whichever occurred first. For patients
with missing discharge SBP measurements, the latest inhospital SBP
measurement was substituted. Change in SBP was defined as the
difference between SBP at baseline and discharge/day 7; analyses
excluded patients who died during hospitalization.
Everest end points
da
po
An independent and blinded adjudication committee determined
the cause of all hospitalizations and deaths during follow-up. Rehospitalization was defined as a nonelective hospital admission for
medical therapy with a duration that extended over a change in
calendar date. Hospitalization for HF was defined as admission that
included substantive worsening of symptoms and/or signs resulting
in augmentation of oral medications or new administration of
intravenous therapies or ultrafiltration. Mode of death was
adjudicated as CV, non-CV, or unknown. Cardiovascular deaths
were further classified as sudden cardiac death (SCD), HF death,
acute myocardial infarction (AMI), stroke, or other. Non-CV death
was defined as a death caused by a specific non-CV event, whereas
unknown death was defined as a death for which no information
surrounding the event was available. Prespecified coprimary end
points were all-cause mortality (ACM) and the composite of CV
mortality (CVM) or HHF (CVM + HHF). Each of the coprimary end
points was analyzed as time to first event.
Study overview
aa
Methods
ut
or
ia
za
interpreted in the context of current disease status (ie,
acute vs chronic), overall progression (ie, early vs endstage), and concurrent medical therapies (ie, vasodilators,
inotropes, etc). 3 In the ambulatory setting, chronically
uncontrolled hypertension has a well-established role in
the development and progression of HF. 4 In contrast,
registries enrolling patients during HHF have demonstrated that approximately 50% of patients have a systolic BP
(SBP) greater than or equal to 140 mm Hg at admission
and higher measurements confer a graded protective
effect on mortality even at extremes of SBP. 5–7 However,
the pathophysiologic significance and prognostic value of
SBP after the initiation of inhospital therapy have not
been well described. In addition, SBP-related differences
in the cause of postdischarge death and cardiovascular
(CV)-related readmissions have not been previously
reported during HHF.
The EVEREST trial provides an opportunity to characterize the baseline clinical features and outcomes by SBP
after initiation of inhospital therapy in a large contemporary cohort of patients enrolled during HHF with reduced
ejection fraction (EF). Specifically, the objectives of this
post hoc analysis of the EVEREST clinical trial database
were to (1) describe the profile of patients enrolled in a
clinical trial presenting with low SBP in the absence of
cardiogenic shock, (2) determine the breakdown of the
mode of death and etiology of CV readmissions, and (3)
evaluate the association of SBP after initial therapy with
subsequent morbidity and mortality.
Co
pi
The study design 8 and primary results 9,10 of the EVEREST trial
have been previously reported. Briefly, EVEREST was a global,
prospective, randomized, double-blind, placebo-controlled trial
designed to examine the short- and long-term efficacy and safety
of tolvaptan, a vasopressin-2 receptor antagonist. Patients ≥18
years of age hospitalized for worsening HF with symptoms
classified as New York Heart Association (NYHA) class III or IV,
with a left ventricular (LV) EF ≤40%, and presenting with 2 or
more signs or symptoms of volume overload (dyspnea, pitting
edema, and jugular venous distension [JVD]) were eligible for
enrollment. Relevant exclusion criteria included supine SBP b90
mm Hg; hemodynamically significant uncorrected primary
cardiac valvular disease; serum creatinine greater than 3.5 mg/
dL or 309.4 μmol/L; subjects currently treated with hemofiltration
or dialysis, or refractory, end-stage HF (ie, defined as candidates
for ventricular assist devices or continuous positive intravenous
inotropic therapy); or a life expectancy of less than 6 months as
determined by the enrolling clinician-investigator.
Patients were randomized within 48 hours of admission to
receive either 30 mg of oral tolvaptan once daily or matching
placebo, both in addition to standard therapy, for a minimum of
60 days. Background HF therapy was left to the discretion of the
treating physician, but guideline-based recommendations for
optimal medical management were included in the study
protocol. The EVEREST trial was conducted in accordance
Statistical analysis
All data were reported as a mean ± SD or a median
(interquartile range). Patients were divided into quartiles (Q14) by SBP at baseline: ≤105, 106 to 119, 120 to 130, and
≥131 mm Hg. Baseline and discharge characteristics as well
as occurrence of clinical events (ie, mode of death and
etiology of CV hospitalizations) were compared across
quartiles of SBP using analysis of variance, Kruskal-Wallis,
and χ 2 tests, as appropriate. The associations between
baseline, discharge, and inhospital change in SBP and ACM
and CVM + HHF were assessed as a continuous variable
using univariate and multivariable Cox proportional hazards
regression models. No evidence of nonlinear effects was
observed by Kolmogorov-type supremum testing for all
analyses. All multivariable models were adjusted for 20
preselected clinical and laboratory variables: age, sex,
geographic region, medical comorbidities (ie, diabetes,
hypertension, and renal insufficiency), history of ventricular
arrhythmias, EF, serum sodium, B-type natriuretic peptide
(BNP)/amino terminal-proBNP, serum blood urea nitrogen
04/04/2014
American Heart Journal
February 2013
218 Ambrosy et al
Co
pi
aa
ut
or
ia
za
da
po
rC
DR
Figure 1
Distribution and descriptive statistics of SBP measurements at baseline (A) and inhospital change (B).
(BUN), QRS duration, NYHA class, baseline medication use
(ie, β-blocker, angiotensin-converting enzyme inhibitors
[ACEIs]/angiotensin receptor blockers [ARBs], mineralocorticoid receptor antagonists [MRAs], digoxin, and inotropes),
and presence of atrial flutter or fibrillation on baseline
electrocardiogram. Hazard ratios (HRs) are presented along
with the 95% CIs per 10-mm Hg decrease in baseline,
discharge, or inhospital change in SBP.
Funding and manuscript preparation
Otsuka Inc (Rockville, MD) provided financial and material
support for the EVEREST trial. Database management was
04/04/2014
American Heart Journal
Volume 165, Number 2
Ambrosy et al 219
Table I. Demographics and clinical characteristics by baseline SBP quartile
Q4 131-202 mm
Hg (n = 508)
P
66 ± 12.6
65.4 ± 11.6
66.8 ± 10.9
.007
418
31
28
385
(87.6)
(6.5)
(5.9)
(80.7)
460 (88)
29 (5.5)
34 (6.5)
382 (73.0)
427
48
33
347
.074
159 (33.3)
178 (37.3)
72 (15.1)
68 (14.3)
264 (50.5)
91 (17.4)
108 (20.7)
60 (11.5)
258 (50.8)
132 (26.0)
63 (12.4)
55 (10.8)
b.001
377 (79.4)
238 (50.1)
344 (72.1)
269 (56.4)
106 (22.2)
117 (24.5)
99 (20.8)
24 (5.1)
66 (13.9)
312 (65.4)
254 (53.7)
149 (31.5)
109 (22.9)
92 (19.3)
187 (39.2)
153 (32.1)
54 (11.3)
403 (77.5)
236 (45.5)
373 (71.3)
265 (50.7)
65 (12.4)
82 (15.7)
118 (22.6)
22 (4.3)
59 (11.4)
395 (75.5)
287 (55.1)
105 (20.2)
49 (9.4)
47 (9)
176 (33.7)
107 (20.5)
46 (8.8)
370 (73.3)
230 (45.5)
381 (75.4)
246 (48.6)
68 (13.4)
88 (17.3)
130 (25.6)
24 (4.8)
65 (12.9)
469 (92.3)
251 (49.6)
97 (19.2)
43 (8.5)
40 (7.9)
208 (40.9)
115 (22.6)
42 (8.3)
.001
.223
.001
.082
b.001
b.001
.086
.544
.380
b.001
.283
b.001
b.001
b.001
.088
b.001
.155
68.4 ± 8.9
79.3 ± 15.8
179 (37.5)
427 (91.4)
115 (24.9)
396 (84.6)
245 (51.4)
285 (59.7)
26.9 ± 8
28 (20-37.5)
1.3 (1.1-1.7)
126 (102-157)
139.2 ± 4.4
877.5 (416.8-1817.7)
4906.5 (2202-10,076)
75.7 ± 8.8
80 ± 15
215 (41.1)
472 (90.9)
156 (30.2)
414 (79.8)
310 (59.3)
299 (57.2)
28.6 ± 7.4
25 (20-33)
1.2 (1-1.5)
121 (96-144)
140.4 ± 4.5
609.5 (249-1278)
4648.8 (1971.3-8501)
84.2 ± 11.8
80.4 ± 16.4
202 (37.8)
459 (91.1)
107 (21.4)
411 (81.5)
328 (64.6)
267 (52.6)
30.7 ± 7.4
23 (18-31)
1.2 (1-1.5)
108.5 (93-136)
141.1 ± 4.4
599 (245.2-1272.5)
4416.3 (2285.9-10,457.1)
b.001
.714
.521
.994
.002
.265
b.001
.015
b.001
b.001
b.001
b.001
b.001
b.001
.930
417 (87.4)
389 (81.6)
341 (71.5)
269 (56.4)
234 (49.1)
193 (40.5)
25 (5.2)
16 (3.4)
100 (21)
191 (40)
463 (88.5)
453 (86.6)
363 (69.4)
292 (55.8)
267 (51.1)
222 (42.4)
18 (3.4)
15 (2.9)
95 (18.2)
164 (31.4)
443 (87.2)
436 (85.8)
338 (66.5)
234 (46.1)
224 (44.1)
242 (47.6)
5 (1)
22 (4.3)
57 (11.2)
159 (31.3)
.897
.109
.263
b.001
.142
b.001
b.001
.640
b.001
.002
368 (79.1)
385 (82.8)
343 (73.8)
391 (76.1)
448 (87.2)
381 (74.1)
361 (72.3)
438 (87.8)
380 (76.2)
.092
.007
.760
po
za
ia
or
aa
pi
Co
rC
DR
Q3 120-130 mm
Hg (n = 523)
ut
Demographics
Age (y), mean ± SD
64.3 ± 12.9
Race
White
449 (83.6)
Black
39 (7.3)
Other
49 (9.1)
Male
429 (79.9)
Region
Eastern Europe
124 (23.1)
North America
220 (41.0)
South America
99 (18.4)
Western Europe
94 (17.5)
Medical history
HHF
448 (83.6)
HL
269 (50.1)
CAD
345 (64.2)
MI
270 (50.3)
PTCA
123 (22.9)
CABG
156 (29.1)
PVD
103 (19.3)
TIA
33 (6.2)
Stroke
57 (10.7)
HTN
279 (52)
Atrial arrhythmias
271 (51)
Ventricular Arrhythmias
184 (34.7)
PPM
147 (27.4)
AICD
115 (21.4)
Diabetes mellitus
198 (36.9)
Renal insufficiency
182 (34)
COPD
63 (11.7)
Vital signs and symptoms and laboratory values
DBP, mean ± SD
62.1 ± 8.5
HR, mean ± SD
79.5 ± 15.8
NYHA class IV
224 (41.9)
Dyspnea
480 (91.3)
JVD
160 (30.7)
Rales
430 (81.6)
Edema
258 (48.2)
Murmur
333 (62)
EF, mean ± SD
24 ± 8.4
BUN, median (IQR)
29 (21-45)
sCr, median (IQR)
1.3 (1.1-1.8)
QRS, median (IQR)
130 (104-161)
Na, mean ± SD
137.8 ± 5
BNP, median (IQR)
943 (469-1846)
NT-proBNP, median (IQR)
5075.9 (2307.1-8923)
Admission medications
Furosemide
468 (87.2)
ACEI/ARB
448 (83.4)
β-Blocker
384 (71.5)
MRA
329 (61.3)
Digoxin
264 (49.2)
Nitrates
153 (28.5)
Inotrope
49 (9.1)
Hydralazine
20 (3.7)
Antiarrhythmics
122 (22.7)
Lipid-lowering agents
208 (38.7)
Discharge medications
Furosemide
394 (77)
ACEI/ARB
416 (81.3)
β-Blocker
376 (73.4)
Q2 106-119 mm
Hg (n = 477)
da
Q1 82-105 mm
Hg (n = 537)
(84.1)
(9.5)
(6.5)
(68.3)
b.001
(continued on next page)
04/04/2014
American Heart Journal
February 2013
220 Ambrosy et al
Table I (continued )
MRA
Digoxin
Nitrates
Inotrope
Hydralazine
Antiarrhythmics
Lipid-lowering agents
Q1 82-105 mm
Hg (n = 537)
Q2 106-119 mm
Hg (n = 477)
Q3 120-130 mm
Hg (n = 523)
Q4 131-202 mm
Hg (n = 508)
P
336 (65.6)
268 (52.3)
116 (22.7)
20 (3.9)
18 (3.5)
114 (22.3)
202 (39.5)
289 (62.2)
231 (49.7)
157 (33.8)
8 (1.7)
18 (3.9)
101 (21.7)
193 (41.5)
314 (61.1)
246 (47.9)
162 (31.5)
8 (1.6)
16 (3.1)
84 (16.3)
180 (35)
277 (55.5)
206 (41.3)
190 (38.1)
0 (0)
23 (4.6)
63 (12.6)
169 (33.9)
.011
.004
b.001
b.001
.636
b.001
.043
HL, Hyperlipidemia; CAD, coronary artery disease; MI, myocardial infarction; PTCA, percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass graft; PVD,
peripheral vascular disease; TIA, transient ischemia attack; HTN, hypertension; PPM, permanent pacemaker; AICD, automated implantable cardioverter-defibrillator; COPD, chronic
obstructive pulmonary disease; DBP, diastolic BP; HR, heart rate; sCr, serum creatinine; NT-proBNP, amino terminal-proBNP.
Clinical and laboratory variables
On examination, patients with a lower SBP were
more likely to exhibit JVD or have an audible murmur
and less likely to have signs of edema, whereas other
signs and symptoms of congestion were comparable
across SBP quartiles. Patients with an SBP in the bottom
quartile had more renal dysfunction and higher
natriuretic peptide concentrations, wider QRS duration,
and a lower EF.
Results
Baseline and discharge medications
At baseline, the use of β-blockers and ACEIs/ARBs was
similar between SBP quartiles. In contrast, MRAs were
more likely to be prescribed to patients with a lower SBP.
Patients in the lowest SBP quartile were more likely to
receive inotropes, whereas those in the highest quartile
tended to more often receive nitrates. Finally, patients in
the bottom quartile of SBP were more likely to be taking
an antiarrhythmic or lipid-lowering agent. Trends in
medication use remained relatively unchanged between
baseline and discharge, with the exception that patients
with a lower SBP were less likely to be prescribed an ACEI
or an ARB at discharge.
Co
pi
aa
ut
or
ia
za
Participants presented with a median (25th, 75th)
baseline SBP of 120 (105, 130) mm Hg ranging from 82 to
202 mm Hg (Figure 1A) and experienced an inhospital
change of −5 (−15, 5) mm Hg ranging from −75 to 60 mm
Hg (Figure 1B). Study participants had a mean age of 65.6
± 12.0 years, were predominantly male, and selfidentified their race as white. Most patients had an
ischemic etiology of HF, and the mean EF was 27.5% ±
8.2%. Approximately 80% of patients had a history of
HHF, and the prevalence of medical comorbidities was
high. Patients were well treated with evidence-based
therapies including β-blockers, ACEIs/ARBs, and MRAs.
da
po
rC
DR
performed by the sponsor according to a prespecified plan. H.S.
conducted all final analyses for this manuscript, with funding
from the Center for Cardiovascular Innovation (Northwestern
University Feinberg School of Medicine, Chicago, IL), using SAS
version 9.3 (Cary, NC). The authors had full access to the data,
take sole responsibility for its integrity, and had complete
control and authority over manuscript preparation and the
decision to publish.
Demographics
Patients with a lower SBP were younger and tended to
be male (Table I). Although there was no association
between baseline SBP and self-identified race, patients
with an SBP in the lowest quartile were more frequently
enrolled in North America and Western Europe, whereas
patients with an SBP in the highest quartile were more
commonly enrolled in Eastern Europe.
Medical history
Patients with a lower SBP were more likely to report a
history of HHF. Of note, the prevalence of percutaneous
transluminal coronary angioplasty and coronary artery
bypass graft was almost twice as high in the bottom
quartile compared with the top quartile. History of
ventricular arrhythmias and cardioverter-defibrillator
implantation was highest in the patient with the
lowest SBP.
Cause of death and rehospitalization
During a median follow-up of 9.9 months, the mortality
rate was 26.4% (n = 543) and the CV hospitalization rate
was 38.9% (n = 772) for the entire cohort (Table II).
There were significant differences in mortality (Q1 38.9%
vs Q4 15.9%) and CV hospitalization (Q1 49.9% vs Q4
29.7%) by SBP. Systolic BP–related disparities in mortality
were driven by HF and SCD, whereas HHF was the only
subset of CV hospitalizations occurring with greater
frequency in patients presenting with lower SBP.
Predictive value of SBP
Kaplan-Meier curves showed increased rates of both
coprimary end points, ACM and CVM + HHF, in the lower
SBP quartiles at baseline (log-rank test, P b .001; Figure 2)
and discharge (log-rank test, P b .001; Figure 3). After
adjusting for potential confounders, SBP as a continuous
04/04/2014
American Heart Journal
Volume 165, Number 2
Ambrosy et al 221
Table II. Coprimary end points, mode of death, and etiology of CV hospitalization by baseline SBP quartile
ACM
CVM + HHF
Mortality
CV
HF
AMI
SCD
Stroke
Other CV
Non-CV
Unknown
CV hospitalizations
CV
HHF
AMI
Arrhythmia
Stroke
Other CV
Q2 106-119 mm
Hg (n = 477)
Q3 120-130 mm
Hg (n = 523)
Q4 131-202 mm
Hg (n = 508)
Total (n = 2045)
209 (38.9)
293 (54.6)
135 (28.3)
215 (45.1)
115 (22.0)
176 (33.7)
81 (15.9)
138 (27.2)
540 (26.4)
822 (40.2)
b.001
b.001
145 (28.3)
82 (16.0)
1 (0.2)
53 (10.3)
2 (0.4)
7 (1.4)
22 (4.3)
18 (3.5)
85 (18.3)
43 (9.3)
6 (1.3)
24 (5.2)
3 (0.7)
9 (1.9)
20 (4.3)
17 (3.7)
76 (14.9)
33 (6.5)
5 (1.0)
30 (5.9)
0 (0.0)
8 (1.6)
16 (3.1)
10 (2.0)
49 (9.8)
22 (4.4)
2 (0.4)
21 (4.2)
2 (0.4)
2 (0.4)
13 (2.6)
10 (2.0)
355 (17.9)
180 (9.1)
14 (0.7)
128 (6.4)
7 (0.4)
26 (1.3)
71 (3.6)
55 (2.8)
b.001
b.001
.136
b.001
.366
.133
.372
.192
256 (49.9)
195 (38.0)
9 (1.8)
16 (3.1)
4 (0.8)
32 (6.2)
201 (43.3)
146 (31.5)
7 (1.5)
20 (4.3)
5 (1.1)
23 (5.0)
167 (32.8)
118 (23.1)
6 (1.2)
14 (2.8)
3 (0.6)
26 (5.1)
148 (29.7)
96 (19.2)
8 (1.6)
7 (1.4)
7 (1.4)
30 (6.0)
772 (38.9)
555 (27.9)
30 (1.5)
57 (2.9)
19 (1.0)
111 (5.6)
b.001
b.001
.893
.059
.581
.762
dyssynchrony, a finding that is particularly relevant to
patients presenting with low SBP, and prolonged QRS
duration is an independent predictor of postdischarge
morbidity and mortality. 11 Similarly, only 20% of the
patients in the bottom quartile of SBP had an ICD at the
time of enrollment, although nearly all these patients met
the guideline-recommended criteria for primary prevention. In addition, the prevalence of ventricular arrhythmias was close to 35%, and the incidence of SCD was
approximately 10% during follow-up. These findings
suggest that LV dyssynchrony and ventricular arrhythmias
may be undertreated in this high-risk population.
However, studies examining medical device therapy in
HF have been exclusively conducted in the ambulatory
setting in stable patients with chronic HF on maximal
medical therapy, and existing guidelines do not discuss
the application of these potentially lifesaving treatments
in HHF. 12 Nevertheless, some experts have advocated
further research to clarify the optimal timing and
application of electrophysiology diagnostics and therapeutics during HHF. 13–15
It is also notable that upward of 10% of patients in the
bottom quartile of SBP required inotropic support at some
point during hospitalization. Although inotropes may be
clinically indicated for acute stabilization, even short-term
use has been associated with increased inhospital and
postdischarge mortality. 16 However, it is presently
unclear whether SBP itself is simply a marker of poor
prognosis or a potential targetable mediator of outcomes.
Discussion
Co
pi
aa
ut
or
da
ia
za
Inhospital change in SBP
Inhospital change in SBP was a univariate predictor of
both ACM (HR 0.93/10-mm Hg decrease, 95% CI 0.880.98) and CVM + HHF (HR 0.93/10-mm Hg decrease, 95%
CI 0.89-0.97). In contrast, the change in SBP from baseline
to discharge was not associated with either ACM (P = .26)
or CVM + HHF (P = .61) after accounting for prespecified
covariates (Table III). In addition, the association
between change in SBP and the coprimary end points
remained statistically nonsignificant when patients were
divided by SBP quartile at baseline.
P
po
variable was directly associated with risk of the coprimary
end points both at baseline and at discharge/day 7 (Table III).
rC
DR
Q1 82-105 mm
Hg (n = 537)
The present analysis contributes several additions to the
existing literature on SBP during HHF. First, the distribution
of SBP measurements after initial therapy was narrow.
Second, the low SBP clinical profile included young men
who had a higher prevalence of revascularization, ventricular arrhythmias, and ICD implantation; had a lower EF and
worse laboratory profile; and were more likely to require
intravenous inotropes during hospitalization. Lower SBP
was associated with increased mortality, driven by
progressive HF and SCD and CV hospitalizations, because
of an excess of HHF. Finally, low SBP remained a poor
prognostic indicator after the initiation of standard therapy
and the resolution of the “acute” phase of hospitalization.
Identifying targets for treatment
It has already been reported that more than half of the
patients without any device therapy enrolled in the
EVEREST trial had electrocardiographic evidence of LV
Mode of death and etiology of CV rehospitalization
Despite trends toward improvement in per-capita riskadjusted HHF rates in the United States, 17,18 postdischarge mortality and readmission, respectively, may
be as high as 15% and 30% within 60 to 90 days. 5
04/04/2014
American Heart Journal
February 2013
222 Ambrosy et al
Co
pi
aa
ut
or
ia
za
da
po
rC
DR
Figure 2
Kaplan-Meier curves for ACM (A) and composite of CVM + HHF (B) by baseline SBP quartile (note: Q1 ≤105 mm Hg, Q2 106-119 mm Hg, Q3
120-130 mm Hg, and Q4 ≥131 mm Hg).
Although it is well established that admission SBP is an
important predictor of mortality, 5,19,20 this study establishes that SBP-related disparities in postdischarge mortality is principally caused by progressive HF and SCD. In
contrast to the data from the Organized Program to
Initiate Lifesaving Treatment in Hospitalized Patients with
Heart Failure (OPTIMIZE-HF) registry, which found allcause hospitalization to be comparable across quartiles of
04/04/2014
American Heart Journal
Volume 165, Number 2
Ambrosy et al 223
Co
pi
aa
ut
or
ia
za
da
po
rC
DR
Figure 3
Kaplan-Meier curves for ACM (A) and composite of CVM + HHF (B) by discharge SBP quartile (note: Q1 ≤ 100 mm Hg, Q2 = 101-111 mm Hg,
Q3 = 112-125 mm Hg, and Q4 ≥126 mm Hg).
SBP, the present analysis found lower SBP to be
associated with higher CV readmission rates as a result
of a greater incidence of HHF. However, it should be
noted that in the EVEREST trial, approximately half of the
hospitalizations occurring during follow-up were adjudicated as non-CV in etiology, highlighting the difficulty in
interpreting these seemingly discrepant findings based on
all-cause vs CV hospitalizations. 21
04/04/2014
American Heart Journal
February 2013
224 Ambrosy et al
Table III. Hazard ratios and 95% CI for risk of ACM and CVM +
HHF per 10-mm Hg decrease in baseline, discharge, or inhospital
change in SBP
define a “reference range” and systematically study
inhospital changes in SBP.
ACM
CVM + HHF
1.15 (1.08-1.22)
1.15 (1.08-1.22)
0.99 (0.94-1.05)
1.09 (1.04-1.14)
1.07 (1.02-1.13)
0.98 (0.94-1.03)
Limitations
There are several limitations to the data. First, this
study was performed post hoc and is subject to the
potential biases inherent to exploratory analyses of
observational data. Nonetheless, data were prospectively
collected, and events were adjudicated by a blinded,
independent events committee. Second, enrollment was
permitted up to 48 hours after admission, whereas
standard therapy was likely started in the emergency
department shortly after initial presentation. Third,
patients with a baseline SBP b90 mm Hg and hemodynamically significant uncorrected primary cardiac valvular disease were excluded, greatly limiting the
applicability of this study's findings to cardiogenic
shock and other low output states. Finally, these data
were collected in the context of a clinical trial; patients
were well treated with background evidenced-based
therapies and had prespecified postdischarge follow-up,
potentially restricting the generalizability of this analysis.
Baseline
Discharge
Inhospital change
po
da
Conclusions
Patients hospitalized for HF with low SBP tend to
consist of young men with a low EF, ischemic etiology of
HF, and a high-risk laboratory and neurohormonal profile.
These patients may manifest comparable congestive signs
and symptoms at initial presentation, respond favorably
to standard therapy, but experience a dramatically higher
postdischarge event rate driven by death from progressive HF and SCD and HF-related readmissions. Candidate
targets for intervention include LV dyssynchrony and
ventricular arrhythmias, which require prospective
validation in the acute setting. Importantly, the prognostic value of SBP is not attenuated after the resolution of
the acute phase and remains an independent clinical
predictor of morbidity and mortality throughout admission in HHF with reduced EF.
Co
pi
aa
ut
or
ia
za
Pathophysiologic significance and prognostic value
Although extremes of SBP were recorded, the later
enrollment (ie, up to 48 hours after admission) after the
very acute phase and initiation of standard therapy, as well
as restriction of the EVEREST population to patients with
reduced EF, led to a more narrow SBP distribution at
enrollment than has been described in reports of registries
of patients enrolled during HHF. Thus, it is notable that
small increases in SBP still conferred an appreciable
protective effect. Elevated BP in the acute setting is a
result of high sympathetic tone, termed reactive hypertension, indicating the presence of functional cardiac
reserve in the face of an acute physiologic stressor(s) (ie,
pulmonary congestion). 3 In contrast, low or even normal
BP at presentation, which may be the goal of treatment in
the ambulatory setting, may be a more ominous finding,
reflecting a low cardiac output and suboptimal or
inadequate end-organ perfusion. Of note, the association
of SBP increases with reduced mortality reported in the
present analysis is comparable in magnitude with those
reported in the primary results of OPTIMIZE-HF registry 5
and other studies, 19,20 suggesting that these pathophysiologic principles apply to SBP increases reported beyond
initial presentation in the emergency department up until
the time of discharge.
Interestingly, after adjusting for potential confounders,
inhospital change in SBP from baseline to discharge was
not associated with outcomes for the overall cohort or
within SBP quartile at baseline. This finding is in contrast
to the hypothesis that SBP normalization might identify
treatment “responders” and be associated with a decreased incidence of adverse outcomes. There are a
couple of possible explanations for this neutral finding.
First, the greatest change in SBP occurred within the first
24 hours of enrollment, and perhaps, a discernible effect
may have been detected had study participants been
enrolled closer to the time of initial presentation. Second,
the pathophysiologic significance of inhospital change in
SBP likely depends on starting SBP, which is why
inhospital change in SBP was analyzed within each
quartile of SBP at baseline. However, there are no obvious
absolute physiologic “cutoffs,” making it challenging to
rC
DR
Multivariate models are adjusted for age, sex, geographic region, medical
comorbidities (ie, diabetes, hypertension, and renal insufficiency), history of
ventricular arrhythmias, EF, serum sodium, BNP/amino terminal-proBNP, serum
BUN, QRS duration, NYHA class, baseline medication use (ie, β-blocker, ACEIs/
ARBs, MRAs, digoxin, and inotropes), and presence of atrial flutter or fibrillation on
baseline electrocardiogram. Values are presented as HR (95% CI).
Disclosures
Mihai Gheorghiade is a consultant for and/or has
received research support from Abbott Labs, Astellas,
AstraZeneca, Bayer Schering PharmaAG, CorThera Inc,
Cytokinetics Inc, DebioPharm SA, ErrekappaTerapeutici
(Milan, Italy), Glaxo Smith Kline, JNJ, Medtronic, Novartis
Pharma AG, Otsuka, Sigma Tau, Solvay Pharmaceuticals,
and Pericor Therapeutics. The other authors declare no
conflicts of interest.
References
1. Roger VL, Go AS, Lloyd-Jones DM, et al. Executive summary: heart
disease and stroke statistics—2012 update: a report from the
American Heart Association. Circulation 2012;125(1):188-97.
04/04/2014
American Heart Journal
Volume 165, Number 2
Ambrosy et al 225
da
po
rC
DR
12. Hunt SA, Abraham WT, Chin MH, et al. 2009 Focused update
incorporated into the ACC/AHA 2005 Guidelines for the Diagnosis
and Management of Heart Failure in Adults: a report of the American
College of Cardiology Foundation/American Heart Association Task
Force on Practice Guidelines: developed in collaboration with the
International Society for Heart and Lung Transplantation. Circulation
2009;119(14):e391-479.
13. Wang NC, Fonarow GC, Gheorghiade M. The potential application
of electrophysiology diagnostics and therapeutics in acute heart
failure syndromes. Heart Fail Rev 2011;16(5):437-9.
14. Wang NC, Bhattacharya S, Gheorghiade M. The potential role of
cardiac resynchronization therapy in acute heart failure syndromes.
Heart Fail Rev 2011;16(5):481-90.
15. Wang NC, Piccini JP, Gheorghiade M. Implantable cardioverterdefibrillator implementation in acute heart failure syndromes:
unanswered questions. Heart Fail Rev 2011;16(5):477-80.
16. Abraham WT, Adams KF, Fonarow GC, et al. In-hospital mortality in
patients with acute decompensated heart failure requiring intravenous vasoactive medications: an analysis from the Acute Decompensated Heart Failure National Registry (ADHERE). J Am Coll
Cardiol 2005;46(1):57-64.
17. Gheorghiade M, Braunwald E. Hospitalizations for heart failure in
the United States—a sign of hope. JAMA 2011;306(15):1705-6.
18. Chen J, Normand SL, Wang Y, et al. National and regional trends in
heart failure hospitalization and mortality rates for Medicare
beneficiaries, 1998-2008. JAMA 2011;306(15):1669-78.
19. Fonarow GC, Adams Jr KF, Abraham WT, et al. Risk stratification for
in-hospital mortality in acutely decompensated heart failure:
classification and regression tree analysis. JAMA 2005;293(5):
572-80.
20. Lee DS, Austin PC, Rouleau JL, et al. Predicting mortality among
patients hospitalized for heart failure: derivation and validation of a
clinical model. JAMA 2003;290(19):2581-7.
21. O'Connor CM, Miller AB, Blair JE, et al. Causes of death and
rehospitalization in patients hospitalized with worsening heart failure
and reduced left ventricular ejection fraction: results from Efficacy of
Vasopressin Antagonism in Heart Failure Outcome Study with
Tolvaptan (EVEREST) program. Am Heart J 2010;159(5):841 e1-9 e1.
Co
pi
aa
ut
or
ia
za
2. Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke
statistics—2012 update: a report from the American Heart Association. Circulation 2012;125(1):e2-e220.
3. Gheorghiade M, Vaduganathan M, Ambrosy A, et al. Current
management and future directions for the treatment of patients
hospitalized for heart failure with low blood pressure. Heart Fail Rev
2012.
4. Harinstein ME, Flaherty JD, Fonarow GC, et al. Clinical assessment of
acute heart failure syndromes: emergency department through the
early post-discharge period. Heart 2011;97(19):1607-18.
5. Gheorghiade M, Abraham WT, Albert NM, et al. Systolic blood
pressure at admission, clinical characteristics, and outcomes in
patients hospitalized with acute heart failure. JAMA 2006;296(18):
2217-26.
6. Adams Jr KF, Fonarow GC, Emerman CL, et al. Characteristics and
outcomes of patients hospitalized for heart failure in the United States:
rationale, design, and preliminary observations from the first
100,000 cases in the Acute Decompensated Heart Failure National
Registry (ADHERE). Am Heart J 2005;149(2):209-16.
7. Bhatia RS, Tu JV, Lee DS, et al. Outcome of heart failure with
preserved ejection fraction in a population-based study. N Engl J Med
2006;355(3):260-9.
8. Gheorghiade M, Orlandi C, Burnett JC, et al. Rationale and design of
the multicenter, randomized, double-blind, placebo-controlled study
to evaluate the Efficacy of Vasopressin antagonism in Heart Failure:
Outcome Study with Tolvaptan (EVEREST). J Card Fail 2005;11(4):
260-9.
9. Gheorghiade M, Konstam MA, Burnett Jr JC, et al. Short-term clinical
effects of tolvaptan, an oral vasopressin antagonist, in patients
hospitalized for heart failure: the EVEREST Clinical Status Trials.
JAMA 2007;297(12):1332-43.
10. Konstam MA, Gheorghiade M, Burnett Jr JC, et al. Effects of oral
tolvaptan in patients hospitalized for worsening heart failure: the
EVEREST Outcome Trial. JAMA 2007;297(12):1319-31.
11. Wang NC, Maggioni AP, Konstam MA, et al. Clinical implications of
QRS duration in patients hospitalized with worsening heart failure
and reduced left ventricular ejection fraction. JAMA 2008;299(22):
2656-66.
04/04/2014