Download Plasma Urocortin 1 in Human Heart Failure

Plasma Urocortin 1 in Human Heart Failure
Sue P. Wright, MBChB; Robert N. Doughty, MD, MRCP, FRACP; Chris M. Frampton, PhD;
Greg D. Gamble, MSc; Tim G. Yandle, PhD; A. Mark Richards, MD, PhD, DSc, FRACP, FRSNZ
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 2, 2017
Background—The urocortins are emerging as potentially important contributors to neurohumoral regulation of the
circulation with recent reports attributing a powerful array of hemodynamic, renal, and neurohumoral effects to the
urocortins in cardiac failure. These peptides also seem to have cardioprotective effects in the setting of ischemiareperfusion. Little is known concerning the plasma concentrations of the urocortins in health and disease. We have
investigated plasma urocortin 1 as a potential diagnostic marker of heart failure and documented its relationships to
symptoms, measures of cardiac function, and concurrent levels of other circulating neurohormones.
Methods and Results—In 299 patients with recent onset dyspnea or peripheral edema presenting to primary care, plasma
urocortin 1 and other vasoactive hormones were assayed, and echocardiography was performed. Heart failure was
present in 74 patients (25%) according to predefined diagnostic criteria. Urocortin 1 levels were increased in patients
with heart failure and were related to functional class, clinical signs of heart failure, echocardiographic indicators of left
ventricular dimensions and function, plasma creatinine, and concurrent circulating levels of plasma natriuretic peptides,
adrenomedullin, and endothelin 1.
Conclusions—Plasma urocortin 1 is elevated in heart failure (in proportion to the degree of cardiac dysfunction) in concert
with the generalized neurohormonal activation seen in this condition. Urocortin levels predict heart failure independent
of age, history of previous myocardial infarction, diabetes, hypertension, fractional shortening, and N-terminal
prohormone brain natriuretic peptide levels. (Circ Heart Fail. 2009;2:465-471.)
Key Words: urocortin 䡲 heart failure 䡲 cardiac natriuretic peptides 䡲 adrenomedullin 䡲 endothelin
T
he urocortins are endogenous vasoactive peptides known
to exert powerful beneficial neurohormonal, hemodynamic, and renal actions in experimental heart failure.1–5 In
stable human heart failure infusions of urocortin 1, at the
doses chosen, had little hemodynamic effect, whereas urocortin 2 markedly augmented cardiac output when reducing
systemic vascular resistance.6,7 Urocortins have cardioprotective effects in experimental ischemia/reperfusion.8
Members of the corticotrophin releasing factor (CRF)
family, the urocortins are ancient, with genetic profiling
demonstrating their presence through 550 million years of
evolution.9 The cDNA to urocortin 1 was cloned in 199510
and 2 related peptide sequences, named urocortin 2 and
urocortin 3, were discovered subsequently.11–13 The gene for
urocortin 1 is found in chromosome 2p23-2p21 and encodes
a 123 amino acid polypeptide precursor.14,15
Urocortin 1 expression and peptide release are widespread
in the body; occurring in the central nervous system, gastrointestinal tract, and heart, among other sites.16 –19 Plasma
levels, metabolism, and regulation of the urocortins are still
largely undefined. Urocortin expression and peptide release
from cultured cardiomyocytes is increased by ischemia.20
Secretion in vitro is also enhanced by cytokines including
interleukin-1 ␤, interleukin-6, and tumor necrosis factor-␣21;
factors elevated in heart failure and acute coronary syndromes. Urocortin expression is increased in hypertrophied
rat hearts and in the cardiac ventricles of patients with both
dilated and hypertrophic cardiomyopathy.18,22 Circulating
levels in humans have been reported at ⬇8 pmol/L.23 In
human heart failure, one report indicates levels are elevated
but (perhaps unexpectedly) fall with increasing severity of
symptoms as well as with reducing left ventricular ejection
fraction (LVEF) and increasing age.24 Bioactivity and detectable immunoreactivity may be modulated by a CRF-binding
protein present in humans but not in all mammalian species.
This factor may play an inhibitory role in binding to urocortin
and preventing receptor activation.25,26
Urocortins act predominantly through 2 receptor subtypes,
CRF-1 and CRF-2. Receptors possess 7 transmembrane
domains and are G-protein coupled. CRF-2(a) receptors
constitute the dominant peripheral CRF-2 receptor form,
particularly in the heart and vasculature.15,19 Receptor concentrations are high in the left ventricle and intramyocardial
vessels.19
Received December 1, 2008; accepted June 19, 2009.
From the Department of Medicine (S.P.W., R.N.D., G.D.G.), Faculty of Medicine and Health Sciences, University of Auckland, Auckland, New
Zealand; and Christchurch Cardioendocrine Research Group (C.M.F., T.G.Y., A.M.R.) Department of Medicine, University of Otago, Christchurch, New
Zealand.
Correspondence to Mark Richards, MD, PhD, DSc, FRACP, FRSNZ, Department of Medicine, University of Otago, Christchurch, PO Box 4345,
Christchurch, New Zealand. E-mail [email protected]
© 2009 American Heart Association, Inc.
Circ Heart Fail is available at http://circheartfailure.ahajournals.org
465
DOI: 10.1161/CIRCHEARTFAILURE.108.840207
466
Circ Heart Fail
September 2009
In this study, we set out to establish whether or not plasma
urocortin 1 levels were systematically elevated in the presence of symptomatic heart failure, and to establish the
relationship of plasma concentrations of this peptide with
concurrent symptomatic status, echocardiographic measurements of cardiac dimensions and function and with concurrent circulating concentrations of neurohormones known to
be activated in the presence of heart failure.
Methods
Patients
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 2, 2017
Patients were participants in the Natriuretic Peptides in the Community study, a prospective randomized trial of the diagnostic value of
N-terminal prohormone brain natriuretic peptide (NT-proBNP) in
diagnosis of suspected new onset heart failure in primary care.27 The
protocol was approved by an ethics committee of the New Zealand
Ministry of Health. All patients provided written informed consent.
Patients presenting with breathlessness and/or peripheral edema were
referred to the study by their family doctor. Each patient underwent
further cardiological assessment including ECG, chest x-ray, and
echocardiography (ATL HDI 5000). Blood samples for neurohormone assays were collected from seated patients without previous
resting or other special patient preparation. Blood was collected into
EDTA (7.5 mg/mL, Vacutainer; Becton Dickinson, Franklin Lakes,
NJ) and centrifuged within 2 hours of venesection (3000 rpm for 10
minutes) with plasma stored at ⫺80°C before radioimmunoassay.
The definitive diagnosis of heart failure was provided by a panel
of 3 consultants who reviewed each clinical presentation and all test
results (excluding neurohormone levels). The panel decided whether
individual patients met the European Society of Cardiology diagnostic criteria for a diagnosis of heart failure.28 This required the
presence of appropriate symptoms and signs of pulmonary or
peripheral congestion and the presence of objective evidence of
cardiac dysfunction. In the case of doubt, response to treatment was
considered. This standard was based on that used in other recent
studies of the diagnosis of heart failure.29,30
Radioimmunoassays
Blood samples were collected using standard venipuncture technique
into tubes containing EDTA. Samples were centrifuged and frozen at
⫺70°C before immunoassay. Urocortin 1 radioimmunoassay was
performed on plasma samples, as described previously.31 Sample
preparation included steps previously used to extract total CRF
(bound⫹free) from human plasma. One milliliter of plasma was
mixed with 2 mL methanol and centrifuged. The supernatant was
adjusted to contain 0.003% Triton X-100, dried under an air stream
at room temperature, and reconstituted in 0.5 mL assay buffer.
For assay, 100 ␮L plasma extract or standard plus 100 ␮l
antiserum PBL 5779 diluted 1:165 000 were incubated at 4°C for 24
hours before addition of 100 ␮L radiolabel (I125) Tyr-Ucn-1 (prepared by the chloramine T method and purified by reverse phase
high-performance liquid chromatography) containing 10 000 cpm.
The assay was incubated for an additional 24 hours at 4°C after
which bound and free label was separated by a solid phase second
antibody (donkey anti-rabbit; Sac-Cell, IDS, Bolden, United Kingdom). The radioimmunoassay characteristics (mean⫾SD) include
zero binding, 35.5⫾6.3% (n⫽25); standard curve 50% effective
concentration, 414⫾69 pg/mL (88.1⫾14.6 pmol/L); and a detection
limit of 6.6⫾2.7 pg/mL (1.4⫾0.6 pmol/L) in plasma after 2-fold
concentration during extraction (n⫽25). Cross reactivity with urocortins 2 and 3 was ⬍0.004% and ⬍0.002% with human CRF,
human urotensin 2, or the cardiac natriuretic peptides including atrial
natriuretic peptide (ANP), brain natriuretic peptide (BNP), and
NT-proBNP. Within assay coefficients of variation calculated from
variance between assay duplicates were 13.4% over 0 to 52 pg/mL (0
to 11 pmol/L), 10.3% over 56 to 239 pg/mL (12 to 51 pmol/L), and
3.3% over 239 to 1490 pg/mL (51 to 317 pmol/L). Interassay
coefficients of variation were 17.5% at 63 pg/mL (13.5 pmol/L;
n⫽23), 16.2% at 111 pg/mL (23.7 pmol/L; n⫽20), and 13.3% at 145
pg/mL (30.9 pmol/L; n⫽23). Recovery of urocortin 1 added to
plasma was 61.2% at 141 pg/mL (30 pmol/L), 60.9% at 329 pg/mL
(70 pmol/L), and 59.3% at 470 pg/mL (100 pmol/L). Urocortin
results are given without correction for recovery. The urocortin 1
reference range encompassing 95% of results for healthy subjects (2
to 16 pmol/L) was determined in 98 subjects drawn at random from
the Christchurch, New Zealand, electoral roll and matched to the
study population for age and gender.
Immunoassays for NT-proBNP, BNP, ANP, NT-proANP,
C-type natriuretic peptide, adrenomedullin, and endothelin-1
were measured using our previously validated and published
radioimmunoassays.32–37
Statistical Methods
Plasma concentrations of neurohormones are expressed as mean and
standard deviation (when normally distributed) or median and
interquartile range when skewed. Variables with nonnormal distributions were compared using the Wilcoxon rank test. Other continuous variables are compared by Student t tests. Categorical variables
were compared with the Fisher exact test. In univariate analyses,
Spearman rank coefficient was used to determine correlations
between variables, presented as R values. Modeling to investigate
independent determinants of urocortin 1 was performed by means of
multiple linear regression models. Multiple logistic regression with a
forward stepwise selection based on the significance of the Wald
Score if a candidate variable were entered into the regression was
used to test urocortin 1 and other variables for potential independent
association with a diagnosis of heart failure. Urocortin 1 and
NT-proBNP were compared for their ability to discriminate presence
or absence of heart failure by receiver operating characteristic
analysis.38 The authors had full access to the data and take full
responsibility for the integrity of the data. All authors have read and
agree to the manuscript as written.
Results
Patient Characteristics
Two hundred ninety-nine patients with a median age of 72
years (range 40 to 95 years), 65% women, were included in
the study (Table 1). Forty-nine percent of patients presented
with dyspnea alone, 12% with edema alone, and 39% with
both symptoms. There were 80, 109, 84, and 26 patients in
New York Heart Association (NYHA) functional classes I, II,
III, and IV, respectively. Seventy-four of 299 (25%) were
diagnosed as having heart failure by the panel (8, 22, 27, and
17 in NYHA classes I, II, III, and IV, respectively). Underlying etiologies included ischemic heart disease in 27 (36%)
and was multifactorial with contributions from 2 or more of
ischemia, diabetes, hypertension, and/or atrial fibrillation in
23 (31%). Valvular heart disease was present in 5 patients
(7%), dilated cardiomyopathy in 5 (7%), isolated diastolic
heart failure in 11 (15%), and with cause unknown in a
further 6 patients (8%).
Common non– heart failure conditions leading to presentation among the other 225 patients included respiratory
disease (exacerbation of chronic obstructive pulmonary disease and malignancies), myocardial ischemia with exertional
dyspnea, and obesity. Other less frequent causes included
anxiety with hyperventilation, thoracic deformity, and
chronic pulmonary venous thromboembolic disease. Hypertension was similarly frequent in both heart failure (48%) and
non– heart failure (53%) groups. However, diabetes (28%
versus 10%; P⬍0.001), previous myocardial infarction (34%
versus 8%; P⬍0.001) and prescription of both loop diuretics
Wright et al
A
Heart Failure
(n⫽74)
Age, y
74 (11)
72 (11)
0.16
Gender, female
39 (51)
159 (70)
0.004
Atrial fibrillation
23 (30)
12 (5)
⬍0.0001
Heart rate, bpm
73 (16)
74 (15)
0.75
Systolic BP, mm Hg
140 (24)
148 (22)
0.007
Diastolic BP, mm Hg
76 (17)
80 (14)
0.024
0.114 (0.04)
0.09 (0.03)
⬍0.0001
58.6 (28.1)
69.6 (28.6)
0.02
N-CNP, pmol/L
32.4 (25.4, 39.4)
24.7 (21.0, 29.5)
⬍0.0001
CNP, pmol/L
1.85 (1.35, 2.15)
1.45 (1.13, 1.67)
⬍0.0001
Serum creatinine, mmol/L
Creatinine clearance,
mL/min
Not Heart
Failure (n⫽225)
10000
Clinical, Echo, and
Hormone Variables
P
Neurohormones
ANP, pmol/L
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 2, 2017
NT-proANP, nmol/L
32 (26, 45)
1.77 (1.36, 2.44)
17 (12, 25)
0.77 (0.49, 1.17)
⬍0.0001
⬍0.0001
30 (16, 45)
8 (5, 14)
⬍0.0001
NT-proBNP, pmol/L
183 (103, 319)
38 (20, 76)
⬍0.0001
Endothelin-1, pmol/L
1.7 (1.4, 2.1)
1.0 (2.6, 1.2)
⬍0.0001
Adrenomedullin, pmol/L
7.7 (6.5, 9.7)
7.4 (6.0, 8.8)
0.0126
BNP, pmol/L
Echocardiographic variables
LVEDD, mm
57.5 (10)
50.2 (5.8)
⬍0.0001
LVESD, mm
44.7 (11.5)
33.3 (6.3)
⬍0.0001
LV mass, g
229.7 (90.4)
177.6 (56.7)
⬍0.0001
Fractional shortening, %
23.9 (9.7)
34.3 (7.7)
⬍0.0001
LVEF, %
44.1 (14.7)
60.7 (9.9)
⬍0.0001
LA area, cm
26.2 (5.9)
20.7 (4.8)
⬍0.0001
E-wave velocity, cm/s
63.0 (22.5)
52.0 (17.1)
⬍0.0001
A-wave velocity, cm/s
58.3 (24.7)
69.1 (17.2)
⬍0.0001
E/A ratio
1.29 (0.84)
0.77 (0.30)
0.0003
4.9 (2.1)
6.0 (1.8)
0.0002
E⬘
E/E⬘ ratio
13.6 (6.7)
9.2 (3.9)
0.0014
Deceleration time, ms
175 (75)
215 (170)
⬍0.0001
Data are presented as mean (SD), n (%), or median (25th, 75th percentile).
BP indicates blood pressurel LVEDD, left ventricular end diastolic diameter;
LVESD, left ventricular end-systolic diameter.
(38% versus 18%; P⬍0.001), and angiotensin-converting
enzyme inhibitors (47% versus 20%; P⬍0.001) were more
frequent in patients with heart failure.
Plasma Urocortin 1 Concentrations
Mean plasma urocortin 1 in normal control subjects at
7.2⫾2.9 pmol/L (n⫽98) was significantly lower than levels
recorded in the 225 non– heart failure symptomatic patients
(11.1⫾3.2 pmolL), which in turn were lower than in the 74
patients with a final diagnosis of heart failure (13.6⫾4.1
pmol/L; P⬍0.001 for all intergroup comparisons). Levels of
urocortin 1 and NT-proBNP in symptomatic subjects with
and without a final diagnosis of heart failure are compared in
Figure 1. Urocortin levels fell within a much tighter distribution and heart failure induced lesser absolute and proportional
increments in urocortin 1 than in NT-proBNP.
The significant intergroup differences in plasma urocortin
1 were independent of age and gender. Levels were significantly higher in men than in women in normal subjects and in
the study population, whether the study group was considered
B
p<0.001
Urocortin I (pmol/L)
Patient Characteristics
NT-proBNP (pmol/L)
Table 1.
Plasma Urocortin 1 in Human Heart Failure
1000
100
10
1
467
10000
1000
100
p<0.001
10
1
N
n=225
HF
n=74
N
n=225
HF
n=74
Figure 1. Plasma NT-proBNP (A) and urocortin 1 (B) box plots
showing median, 25th and 75th percentiles, and smallest and
largest values. Identical axes are used to display both
skewed NT-proBNP data and normally distributed urocortin 1
data to facilitate direct comparison. For both peptides, levels
in heart failure are significantly higher than in patients with
symptoms from other causes (P⬍0.001 for both urocortin 1
and NT-proBNP).
in totality or separately for those with and without a final
diagnosis of heart failure (Table 2). No significant relationship to age was seen in the study group, the normal reference
range population, or in any subgroup. We also observed the
expected increases in other neurohormones in heart failure
(Table 1).
Urocortin 1 levels were related to symptoms, exhibiting
stepwise increases from 10.6⫾1.3 to 10.8⫾1.3, 11.5⫾1.4,
and 13.3⫾1.5 pmol/L in NYHA functional classes I through
IV, respectively (P⫽0.007) for the group overall. A similar
trend was seen in the 74 patients with a final diagnosis of
heart failure (11.5⫾3.5; 12.4⫾2.5; 14.3⫾4.3; and 14.6⫾5.0
pmol/L, respectively). Levels from patients in NYHA classes
I and II combined were compared with those in NYHA
classes III and IV combined. Severe symptoms (NYHA
III/IV, n⫽44) were associated with significantly higher mean
plasma urocortin 1 levels than mild (NYHAI/II, n⫽30) symptoms (14.4⫾3.9 pmol/L versus 12.1⫾3.9 pmol/L; P⫽0.017).
Table 2. Plasma Urocortin 1 Levels in Normal, Heart Failure,
and Non–Heart Failure Groups
Plasma Urocortin 1, pmol/L
n
(Mean⫾SD)
Normal
Total
98
7.2⫾2.9
M
33
9.1⫾2.7
F
65
6.4⫾2.6
Heart failure
Total
74
13.6⫾4.1
M
37
14.6⫾4.5
F
37
12.5⫾2.6
Non–heart failure illness
Total
225
11.1⫾3.2
M
68
12.1⫾3.3
F
157
10.7⫾3.1
All 3 groups total values differ (P⬍0.001 for all pair-wise comparisons).
Mean values in males (M) are greater than in females (F) in all 3 groups
(P⬍0.001 for all).
Circ Heart Fail
NT-proBNP (pmol/L)
468
September 2009
Table 3. Multivariable Analysis for Independent Predictors of
Heart Failure
1000
n=299
r=0.293 (0.19-0.39)
p<0.001
10
1
BNP (pmol/L)
1000000
n=294
r=0.311(0.20-0.41)
p<0.001
100
1
10
n=286
r=0.384(0.28-0.48)
p<0.001
1
0.1
10000
ADM (pmol/L)
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 2, 2017
CNP (pmol/L)
10000
1000
n=298
r=0.331(0.23-0.43)
p<0.001
100
10
100
ET (pmol/L)
Hazard
Ratio
95% CI
Urocortin 1, median
2.97
1.39 – 6.38
NTproBNP, median
7.12
2.71–18.70
Age, per 10-y increment
1.28
0.78–2.13
0.328
LVEF, per 1% fall
1.08
1.04–1.11
⬍0.001
Prior MI
2.27
0.95–5.92
0.063
Diabetes
3.80
1.39–10.42
0.009
Estimated GFR, per 40
mL/min decline
1.58
0.71–3.53
0.261
100
10
n=292
r=0.244(0.13-0.35)
p<0.001
1
0.1
0
10
20
30
Urocortin I (pmol/L)
Figure 2. Plasma urocortin 1 (in patients presenting with dyspnea and/or edema) plotted against concurrent levels of
NT-proBNP, BNP, C-type natriuretic peptide, adrenomedullin,
and endothelin 1 levels. R values are given with 95% CIs.
Peptide levels were also significantly related to jugular
venous pressure (measured by standard clinical technique as
the distance of the highest excursion of internal jugular
venous waves in centimeters above the sternal angle with
torso at 45° and legs horizontal) in those with heart failure
(R⫽0.309, n⫽72, P⫽0.008) but not in those without
(R⫽0.05, n⫽222, P⫽0.495).
Urocortin was related to left ventricular dimensions and
function. Correlations with left ventricular end-diastolic dimension, left ventricular end-systolic dimension, fractional
shortening, and LVEF were significant (R⫽0.190, 0.228,
⫺0.291, and ⫺0.187, respectively with sample number varying between 212 for calculated left ventricular mass to 259 for
left ventricular end-diastolic dimension and corresponding P
values varying between 0.002 and ⬍0.0001). R values tended
Variable
Significance
(P )
0.005
⬍0.001
MI indicates myocardical infraction; GFR, glomerular filtration rate.
to strengthen when analyses was confined to the 74 patients
with heart failure (R⫽0.192, 0.266, ⫺0.313 and ⫺0.238,
respectively) and lost significance when only the 255 patients
without heart failure were considered (P⬎0.05 for all).
Significant positive relationships were observed between
plasma urocortin 1 levels and levels of the other neurohormones measured, including ANP, NT-proANP, BNP, NTproBNP, C-type natriuretic peptide, adrenomedullin, and
endothelin 1 (R values varying between 0.210 for endothelin
1 and 0.398 for C-type natriuretic peptide levels, sample size
varying between 286 and 298 with P values ⬍0.0001 for all
relationships; Figure 2).
An inverse relationship with renal function was observed.
Urocortin was directly related to plasma creatinine (R⫽0.249,
n⫽294, P⬍0.0001) and inversely related to estimated glomerular filtration rate (R⫽⫺0.107, n⫽286, P⫽0.072). By
multiple linear regression analysis, only gender (P⬍0.0001),
NT-proBNP levels (P⫽0.002), and a diagnosis of heart
failure (P⫽0.034) remained independently predictive of increased plasma urocortin levels (R2⫽0.16). Other variables,
which were significantly associated in univariate analysis but
not in this multivariate analysis included plasma creatinine,
all neurohormones with the exception of NT-proBNP, and the
different measures of cardiac function.
With respect to predictors of heart failure, univariate
relationships were observed for fractional shortening, creatinine clearance, NT-proBNP, and plasma urocortin 1.
By multivariate analysis, using stepwise multiple logistic
regression, urocortin remained an independent predictor of
heart failure (P⫽0.005 Table 3) independent of age, history
of previous myocardial infarction, diabetes, LVEF, NTproBNP, and calculated creatinine clearance.
The likelihood of a final diagnosis of heart failure rose
from 15% to 20% to 39% in first, second, and third tertiles of
plasma urocortin 1, respectively. In comparison, for tertiles of
plasma NT-proBNP the corresponding figure were 5%, 12%,
and 58%. In patients falling within both the lowest tertile of
urocortin 1 and the lowest tertile of NT-proBNP, 5% had
heart failure. Conversely, in those in the upper tertile plasma
concentrations of both peptides, the risk of heart failure was
75% and 49% of all patients with heart failure fell within this
subgroup (Figure 3).
Wright et al
60
469
1.00
36
.75
Urocortin I
(tertiles)
10
11
40
Low
Mid
High
7
20
3
2
2
Sensitivity
Heart Failure Patients (%)
80
Plasma Urocortin 1 in Human Heart Failure
Reference Line
.25
2
1
0.00
0
0.00
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 2, 2017
Low
n=100
NTproBNP
Urocortin 1
.50
Mid
n=101
High
n=98
NT-proBNP (tertiles)
Figure 3. Distribution of patients with heart failure (n⫽74)
according to tertiles (low, mid, and high) of plasma NT-proBNP
and urocortin 1 levels derived from the total study group
(n⫽299). Forty-nine percent (36/74) of patients with a final diagnosis of heart failure exhibit upper tertile levels of both peptides.
Urocortin 1 and NT-proBNP were assessed for their ability
to predict the presence or absence of heart failure by receiver
operating characteristic analysis. Urocortin 1 yielded an area
under the curve (AUC) of 0.68 (95% CI, 0.61 to 0.75,
P⬍0.001). The best-performing value of urocortin 1 (12.3
pmol/L) was associated with sensitivity, specificity, and
positive and negative predictive values of 61%, 67%, 38%,
and 84%, respectively. This was clearly a weaker performance than that observed for NT-proBNP with an AUC of
0.85 (95% CI, 0.80 to 0.90, P⬍0.0001), sensitivity of 80%,
specificity of 81%, positive predictive value of 58%, and
negative predictive value of 93%. Area under the receiver
operating characteristic was significantly less for urocortin 1
when compared with NT-proBNP (P⬍0.01; Figure 4).
Discussion
In patients with a final diagnosis of heart failure, plasma
urocortin levels were double those in the normal reference
range population. Symptomatic patients without a final diagnosis of heart failure also had elevated urocortin at intermediate levels ⬇50% higher than those observed in the normal
range. Plasma urocortin levels were related to symptoms
(NYHA class), signs (jugular venous pressure), concurrent
levels of other neurohormones known to be activated in heart
failure, and echocardiographic measures of cardiac dimensions and function. Urocortin was independently predictive
of heart failure. Urocortin 1 is clearly not as powerful a
stand-alone indicator of cardiac function or as good a marker
of the presence of heart failure as the natriuretic peptides
(Figure 4). However, it does add additional and independent
information beyond that offered by established predictors as
reflected in our multivariable analysis. Intuitively, this is best
appreciated from our tertile analysis (Figure 3). When NT-
.25
.50
.75
1.00
1 - Specificity
Figure 4. Receiver operating characteristic, AUC (95% CI) for
discrimination of heart failure by urocortin 1 (AUC, 0.68; [0.61 to
0.75]) and NT-proBNP (AUC, 0.87; [0.82 to 0.92[). AUCs differed
significantly (P⬍0.001).
proBNP results were considered alone plasma levels in the
upper tertile carried a 58% chance of a final diagnosis of heart
failure; when both NT-proBNP and urocortin levels fell in the
upper tertile the likelihood rose to 75% and half of all patients
with heart failure exhibited top tertile levels for both peptides.
Even in symptomatic patients without heart failure, urocortin 1 was increased above normal. The reasons for this are
uncertain, but it is possible a burden of cardiovascular disease
within the non– heart failure study group produced increased
urocortin 1 at an intermediate level between truly healthy
subjects and those with symptomatic or overt heart failure.
Furthermore, urocortin may also be stimulated by noncardiovascular illness. The stimulation of urocortin secretion in
vitro by interleukin-1, interleukin-6, and tumor necrosis
factor-␣21 raises the possibility that an array of inflammatory
states, with attendant cytokine activation, may result in some
increase in plasma urocortin 1. Notably, the R2 value (0.16)
from the multivariate model assessing independent determinants of plasma urocortin 1 only reflects 16% of the interindividual variation in plasma urocortin 1 levels. This suggests
other factors, which influence plasma urocortin1 levels remain to be found. Further studies in health and disease should
establish the spectrum of stimuli, which trigger increases in
circulating urocortin 1.
Ng et al24 have also reported elevated levels of urocortin in
heart failure. In contrast to this previous report, we observed
an inverse relationship with LVEF rather than the positive
correlation reported previously. In addition, we observed a
continuous increase in plasma urocortin levels across NYHA
functional classes I to IV in contradistinction to the report
from Ng et al in which levels seem to peak in NYHA classes
I and II and were somewhat less in the NYHA classes III and
IV. We observed no relationship with age rather than the
inverse relationship reported previously.24
From this data set, it is not possible to determine why our
results differ from the previous report in such clear-cut
fashion. Answers may lie within the differing characteristics
470
Circ Heart Fail
September 2009
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 2, 2017
of the 2 immunoassays used. They are likely to detect
differing epitopes and/or partial degradation products of
urocortin. The potential effects of corticotrophin binding
protein on assay performance may also differ between assays.
The assay used in this report exhibits no significant crossreactivity with relevant alternative peptides and has been
fully validated.31 Notably, in this report, plasma urocortin
levels follow the common pattern of neurohormones known
to be activated in heart failure. That is, they increase with
increasing severity of symptoms and cardiac dysfunction and
rise in parallel with other neurohormonal markers including
the natriuretic peptides, adrenomedullin, and endothelin 1.
Clearly, further corroborating data are required.
In contrast to the cardiac peptides, plasma urocortin 1 is
higher in men than in women. This pattern was apparent in
the symptomatic cohort overall and in age and gender
matched healthy subjects. The underlying reasons for the
gender difference are unknown. Levels of urocortin are also
mildly inversely related to renal function by univariate
analysis, although in multivariate tests, this relationship to
creatinine clearance was not independent of other predictors
of elevated urocortin 1.
In summary, plasma urocortin 1 levels are elevated in
human heart failure with significant positive relationships to
other circulating neurohormones known to be activated in this
condition. They are also related to symptoms, signs (ie,
jugular venous pressure), and echocardiographic measures of
cardiac structure and function. In contradistinction to a
previous report,24 we found an inverse rather than positive
relationship to LVEF. In addition, the relationship between
urocortin 1 and symptomatic status was continuous with
levels showing a continuous increment between NYHA
classes I–IV.
Our findings suggest that plasma urocortin levels are
unlikely to assume a role as a diagnostic marker in heart
failure. However, they do add to our overall understanding of
the pathophysiological response to heart failure. In experimental heart failure, deliberate antagonism of endogenous
urocortins clearly exacerbates hemodynamic, neurohormonal,
and renal impairment.3 Taken together with previous studies
defining the bioactivity of the urocortins in experimental
heart failure,1–5 it seems likely that elevated plasma urocortin
1 in human heart failure reflects a beneficial compensatory
response to this condition.
Sources of Funding
Support for this work was provided by the Health Research Council
of New Zealand and the National Heart Foundation of New Zealand.
Marguerite Hunter provided secretarial support.
Disclosures
Dr Richards holds the National Heart Foundation of New Zealand
Chair of Cardiovascular Studies.
References
1. Rademaker MT, Charles CJ, Espiner EA, Fisher S, Frampton CM, Kirkpatrick CMJ, Lainchbury JG, Nicholls MG, Richards AM, Vale WW.
Beneficial hemodynamic, endocrine, and renal effects of urocortin in
experimental heart failure: comparison with normal sheep. J Am Coll
Cardiol. 2002;40:1495–1505.
2. Rademaker MT, Cameron VA, Charles CJ, Richards AM. Integrated
hemodynamic, hormonal, and renal actions of urocortin 2 in normal and
paced sheep: beneficial effects in heart failure. Circulation. 2005;112:
3624 –3632.
3. Rademaker MT, Charles CJ, Espiner EA, Frampton CM, Lainchbury JG,
Richards AM. Endogenous urocortins reduce vascular tone and renin-aldosterone/endothelin activity in experimental heart failure. Eur Heart J.
2005;26:2046 –2054.
4. Rademaker MT, Charles CJ, Espiner EA, Frampton CM, Lainchbury JG,
Richards AM. Four-day urocortin-I administration has sustained beneficial haemodynamic, hormonal, and renal effects in experimental heart
failure. Eur Heart J. 2005;26:2055–2062.
5. Rademaker MT, Charles CJ, Richards AM. Urocortin 1 administration
from onset of rapid left ventricular pacing represses progression to overt
heart failure. Am J Physiol Heart Circ Physiol. 2007;293:H1536 –H1544.
6. Davis ME, Pemberton CJ, Yandle TG, Lainchbury JG, Rademaker MT,
Nicholls MG, Frampton CM, Richards AM. Effect of urocortin-1 infusion
in humans with stable congestive cardiac failure. Clin Sci. 2005;109:
381–388.
7. Davis ME, Pemberton CJ, Yandle TG, Fisher SF, Lainchbury JG,
Frampton CM, Rademaker MT, Richards M. Urocortin 2 infusion in
human heart failure. Eur Heart J. 2007;28:2589 –2597.
8. Brar BK, Jonassen AK, Egorina EM, Chen A, Negro A, Perrin MH, Mjos
OD, Latchman DS, Lee K-F, Vale W. Urocortin-II and urocortin-III are
cardioprotective against ischemia reperfusion injury: an essential endogenous cardioprotective role for corticotropin releasing factor receptor type
2 in the murine heart. Endocrinology. 2004;145:24 –35.
9. Chang CL, Hsu SYT. Ancient evolution of stress-regulating peptides in
vertebrates. Peptides. 2004;25:1681–1688.
10. Vaughan J, Donaldson C, Bittencourt J, Perrin MH, Lewis K, Sutton S,
Chan R, Turnbull AV, Lovejoy D, Rivier C, Rivier J, Sawchenko PE,
Vale W. Urocortin, a mammalian neuropeptide related to fish urotensin I
and to corticotropin-releasing factor. Nature. 1995;378:287–292.
11. Hsu SY, Hsueh AJW. Human stresscopin and stresscopin-related peptide
are selective ligands for the type 2 corticotropin-releasing hormone
receptor. Nat Med. 2001;7:605– 611.
12. Imaki T, Katsumata H, Miyata M, Naruse M, Imaki J, Minami S.
Expression of corticotropin releasing factor (CRF), urocortin and CRF
type 1 receptors in hypothalamic-hypophyseal systems under osmotic
stimulation. J Neuroendocrinol. 2001;13:328 –338.
13. Lewis K, Li C, Perrin MH, Blount A, Kunitake K, Donaldson C, Vaughan
J, Reyes TM, Gulyas J, Fischer W, Bilezikjian L, Rivier J, Sawchenko
PE, Vale WW. Identification of urocortin III, an additional member of the
corticotropin-releasing factor (CRF) family with high affinity for the
CRF2 receptor. Proc Natl Acad Sci USA. 2001;98:7570 –7575.
14. Scarabelli T, Knight R. Urocortins: take them to heart. Curr Med Chem
Cardiovasc Hematol Agents. 2004;2:335–342.
15. Eckart K, Radulovic J, Radulovic M, Jahn O, Blank T, Stiedl O, Spiess
J. Actions of CRF and its analogs. Curr Med Chem. 1999;6:1035–1053.
16. Iino K, Sasano H, Oki Y, Andoh N, Shin R-W, Kitamoto T, Takahashi K,
Suzuki H, Tezuka F, Yoshimi T, Nagura H. Urocortin expression in the
human central nervous system. Clin Endocrinol. 1999;50:107–114.
17. Muramatsu Y, Fukushima K, Iino K, Totsune K, Takahashi K, Suzuki T,
Hirasawa G, Takeyama J, Ito M, Nose M, Tashiro A, Hongo M, Oki Y,
Nagura H, Sasano H. Urocortin and corticotropin-releasing factor
receptor expression in the human colonic mucosa. Peptides. 2000;21:
1799 –1809.
18. Nishikimi T, Miyata A, Horio T, Yoshihara F, Nagaya N, Takishita S,
Yutani C, Matsuo H, Matsuoka H, Kangawa K. Urocortin, a member of
the corticotropin-releasing factor family, in normal and diseased heart.
Am J Physiol Heart Circ Physiol. 2000;279:H3031–H3039.
19. Wiley KE, Davenport AP. CRF2 receptors are highly expressed in the
human cardiovascular system and their cognate ligands urocortins 2 and
3 are potent vasodilators. Br J Pharmacol. 2004;143:508 –514.
20. Brar BK, Stephanou A, Okosi A, Lawrence KM, Knight RA, Marber MS,
Latchman DS. CRH-like peptides protect cardiac myocytes from lethal
ischaemic injury. Mol Cell Endocrinol. 1999;158:55– 63.
21. Ikeda K, Tojo K, Tokudome G, Hosoya T, Tajima N, Oki Y, Asaba K,
Hashimoto K, Harade M, Nakao K. Urocortin is an endogenous cardioprotective factor against negative effects of cytokines in the heart.
Abstract of Paper Presented at the Endocrine Society’s 83rd Annual
Meeting, 2001.
22. Ikeda K, Tojo K, Tokudome G, Ohta M, Sugimoto K, Tamura T, Tajima
N, Mochizuki S, Kawakami M, Hosoya T. Cardiac expression of urocortin (Ucn) in diseased heart; preliminary results on possible
Wright et al
23.
24.
25.
26.
27.
28.
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 2, 2017
29.
30.
involvement of Ucn in pathophysiology of cardiac diseases. Mol Cell
Biochem. 2003;252:25–32.
Watanabe F, Oki Y, Ozawa M, Masuzawa M, Iwabuchi M, Yoshimi T,
Nishiguchi T, Iino K, Sasano H. Urocortin in human placenta and
maternal plasma. Peptides. 1999;20:205–209.
Ng LL, Loke IW, O’Brien RJ, Squire IB, Davies JE. Plasma urocortin in
human systolic heart failure. Clin Sci. 2004;106:383–388.
Behan DP, De Souza EB, Lowry PJ, Potter E, Sawchenko P, Vale WW.
Corticotropin releasing factor (CRF) binding protein: a novel regulator of
CRF and related peptides. Front Neuroendocrinol. 1995;16:362–382.
Seasholtz AF, Valverde RA, Denver RJ. Corticotropin-releasing
hormone-binding protein: biochemistry and function from fishes to
mammals. J Endocrinol. 2002;175:89 –97.
Wright SP, Doughty RN, Pearl A, Gamble GD, Whalley GA, Walsh HJ,
Gordon G, Bagg W, Oxenham H, Yandle T, Richards M, Sharpe N.
Plasma amino-terminal pro-brain natriuretic peptide and accuracy of
heart-failure diagnosis in primary care: A randomized, controlled trial.
J Am Coll Cardiol. 2003;42:1793–1800.
Guidelines for the diagnosis of heart failure. The Task Force on Heart
Failure of the European Society of Cardiology. Eur Heart J. 1995;16:
741–751.
Cowie MR, Struthers AD, Wood DA, Coats AJS, Thompson SG, PooleWilson PA, Sutton GC. Value of natriuretic peptides in assessment of
patients with possible new heart failure in primary care. The Lancet.
1997;350:1349 –1353.
Maisel AS, Krishnaswamy P, Nowak RM, McCord J, Hollander JE, Duc
P, Omland T, Storrow AB, Abraham WT, Wu AHB, Clopton P, Steg PG,
Westheim A, Knudsen CW, Perez A, Kazanegra R, Herrmann HC,
McCullough PA; the Breathing Not Properly Multinational Study Inves-
Plasma Urocortin 1 in Human Heart Failure
31.
32.
33.
34.
35.
36.
37.
38.
471
tigators. Rapid measurement of B-type natriuretic peptide in the
emergency diagnosis of heart failure. N Engl J Med. 2002;347:161–167.
Davis ME, Pemberton CJ, Yandle TG, Lainchbury JG, Rademaker MT,
Nicholls MG, Frampton CM, Richards AM. Urocortin-1 infusion in
normal humans. J Clin Endocrinol Metab. 2004;89:1402–1409.
Hunt P, Richards A, Nicholls M, Yandle T, Doughty R, Espiner E.
Immunoreactive amino-terminal pro-brain natriuretic peptide (NTPROBNP): a new marker of cardiac impairment. Clin Endocrinol. 1997;
47:287–296.
Yandle TG, Richards AM, Gilbert A, Fisher S, Holmes S, Espiner EA.
Assay of brain natriuretic peptide (BNP) in human plasma: evidence for
high molecular weight BNP as a major plasma component in heart failure.
J Clin Endocrinol Metab. 1993;76:832– 838.
Yandle TG, Espiner EA, Nicholls MG, Duff H. Radioimmunoassay and
characterization of atrial natriuretic peptide in human plasma. J Clin
Endocrinol Metab. 1986;63:72–79.
Prickett TCR, Yandle TG, Nicholls MG, Espiner EA, Richards AM.
Identification of amino-terminal pro-c-type natriuretic peptide in human
plasma. Biochem Biophys Res Commun. 2001;286:513–517.
Lewis LK, Smith MW, Yandle TG, Richards AM, Nicholls MG.
Adrenomedullin(1–52) measured in human plasma by radioimmunoassay: plasma concentration, adsorption, and storage. Clin Chem. 1998;
44:571–577.
Kitamura K, Tanaka T, Kato J, Eto T, Tanaka K. Regional distribution of
immunoreactive endothelin in porcine tissue: Abundance in inner medulla
of kidney. Biochem Biophys Res Commun. 1989;161:348 –352.
Hanley JA, McNeil BJ. A method of comparing the areas under receiver
operating characteristic curves derived from the same cases. Radiology.
1983;148:839 – 843.
Plasma Urocortin 1 in Human Heart Failure
Sue P. Wright, Robert N. Doughty, Chris M. Frampton, Greg D. Gamble, Tim G. Yandle and A.
Mark Richards
Downloaded from http://circheartfailure.ahajournals.org/ by guest on May 2, 2017
Circ Heart Fail. 2009;2:465-471; originally published online July 8, 2009;
doi: 10.1161/CIRCHEARTFAILURE.108.840207
Circulation: Heart Failure is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX
75231
Copyright © 2009 American Heart Association, Inc. All rights reserved.
Print ISSN: 1941-3289. Online ISSN: 1941-3297
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circheartfailure.ahajournals.org/content/2/5/465
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation: Heart Failure 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: Heart Failure is online at:
http://circheartfailure.ahajournals.org//subscriptions/