Download Growth-differentiation factor-15 improves risk stratification in ST

Clinical research
European Heart Journal (2007) 28, 2858–2865
doi:10.1093/eurheartj/ehm465
Coronary heart disease
Growth-differentiation factor-15 improves risk
stratification in ST-segment elevation myocardial
infarction
Tibor Kempf1†, Erik Björklund2†, Sylvia Olofsson2, Bertil Lindahl2, Tim Allhoff1, Timo Peter1,
Jörn Tongers1, Kai C. Wollert1*, and Lars Wallentin2*
1
Department of Cardiology and Angiology, Hannover Medical School, Carl-Neuberg Str. 1, 30625 Hannover, Germany; and
Department of Cardiology and Uppsala Clinical Research Center, University of Uppsala, University Hospital, 75185 Uppsala,
Sweden
2
Received 20 March 2007; revised 28 August 2007; accepted 24 September 2007; online publish-ahead-of-print 31 October 2007
KEYWORDS
Biomarker;
Growth-differentiation
factor-15;
STEMI;
Prognosis
Aims Growth-differentiation factor-15 (GDF-15) is a transforming growth factor-b-related cytokine that
is induced in the heart following ischaemia–reperfusion injury. We explored the prognostic utility of
GDF-15 in patients with ST-segment elevation myocardial infarction (STEMI) receiving fibrinolytic
therapy.
Methods and results Circulating levels of GDF-15 were determined by an immunoradiometric assay in
741 STEMI patients who were included in the Assessment of the Safety and Efficacy of a New Thrombolytic (ASSENT)-2 and ASSENT-plus trials. About 72.7% of the patients presented with GDF-15 levels
1200 ng/L, the upper limit of normal in apparently healthy elderly individuals. Increased levels of
GDF-15 were associated with a higher risk of death during 1-year follow-up. Mortality rates at 1 year
were 2.1, 5.0, and 14.0% in patients with GDF-15 levels ,1200, 1200–1800, and .1800 ng/L, respectively (P , 0.001). GDF-15 remained an independent predictor of mortality after adjustment for clinical
variables, troponin T, and N-terminal pro-B-type natriuretic peptide (NT-proBNP). GDF-15 provided
prognostic information in clinically relevant patient subgroups, defined according to age, gender,
cardiovascular risk factors, haemodynamic status, and the TIMI risk score. Moreover, GDF-15 added
prognostic information to the established biomarkers of adverse prognosis in STEMI, troponin T, and
NT-proBNP.
Conclusion GDF-15 is a new biomarker in STEMI that provides prognostic information beyond established
clinical and biochemical markers.
Introduction
In-hospital and long-term mortality rates vary considerably across the spectrum of patients with acute
ST-segment elevation myocardial infarction (STEMI), who
receive fibrinolytic therapy.1 Early and continuous risk
stratification is therefore important to enable informed
decisions about tailoring of treatment and allocation of
clinical resources. Advanced age, lower systolic blood
pressure, elevated heart rate, and higher Killip class are
important baseline factors associated with increased mortality rates.2 These and other clinical variables have been
combined into the TIMI risk score, which may serve as a
tool for the risk assessment at the bedside.3,4 In addition,
elevated biomarker levels at presentation, including
cardiac troponin T (cTnT) and N-terminal pro-B-type
* Corresponding author. Tel: þ46 18 6114953; fax: þ46 18 506638.
E-mail address: [email protected] or Tel: þ49 511 532 4055;
fax: þ49 511 532 5412. E-mail address: [email protected]
†
These authors have contributed equally.
natriuretic peptide (NT-proBNP), have been shown to
provide complementary prognostic information on
STEMI.5–7
GDF-15 is a member of the transforming growth factor-b
(TGF-b) cytokine superfamily.8 GDF-15 has recently been
identified as a gene that is up-regulated in cardiac myocytes
by simulated ischaemia, reactive oxygen species, and
pro-inflammatory cytokines, suggesting that GDF-15 may
be a marker of multiple stress pathways in the heart.9,10
Although GDF-15 is not normally expressed in the adult
myocardium, GDF-15 is strongly induced in the heart
during experimental ischaemia and reperfusion injury.9 We
have recently shown that circulating levels of GDF-15 are
elevated and independently related to all-cause mortality
in patients with non-ST-elevation acute coronary syndrome
(NSTE-ACS).11
To explore whether GDF-15 can provide prognostic information in patients with STEMI, circulating levels of GDF-15
and its relations to clinical characteristics, other biomarkers, and mortality were studied in patients with
Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2007.
For permissions please email: [email protected].
GDF-15 improves risk stratification in STEMI
STEMI from the Assessment of the Safety and Efficacy of a
New Thrombolytic (ASSENT)-2 and ASSENT-plus trials.
Methods
2859
vectorcardiography or by continuous 12-lead ECG for the assessment
of ST-segment resolution, as described earlier.18 Sixty minutes after
the start of the recordings, patients were classified according to
ST-segment resolution from maximal ST-elevation (,50%
ST-segment resolution vs. 50% ST-segment resolution).18
Patients and study design
The present investigation is a substudy from the ASSENT-2 and
ASSENT-plus trials. ASSENT-2 was a world-wide, multicentre trial
comparing tenecteplase with front-loaded alteplase. ASSENT-plus
recruited patients in Scandinavia and the USA and compared the
efficacy and safety of dalteparin with unfractionated heparin as
an adjunct to alteplase. The protocols and main results of the two
trials have been published.12,13 Inclusion criteria in both trials
were symptoms of acute myocardial infarction within 6 h of onset,
ST-elevation 0.1 mV in two or more limb leads, 0.2 mV in two
or more contiguous precordial leads, or left bundle branch block.
Major exclusion criteria included known cerebrovascular disease,
uncontrolled hypertension, serum creatinine concentration
.150 mmol/L (only in ASSENT-plus), any substantial trauma, or
increased risk of severe bleeding. Plasma samples were available
from patients enrolled at selected Swedish hospitals in ASSENT-2
(n ¼ 546) and ASSENT-plus (n ¼ 195). The primary endpoint of the
present biomarker study was mortality during 1 year, which was
evaluated by patient records and telephone contacts. The TIMI
risk score, which has been developed as a tool for mortality risk prediction in fibrinolytic-eligible patients with STEMI, was calculated as
previously described.3 Instead of ‘history of diabetes, hypertension,
or angina’, which represents one risk feature in the TIMI score, we
used ‘history of diabetes or hypertension’ because information on
previous angina was not available. On the basis of the TIMI score,
patients were classified as low risk (0–3 points) or high risk (4–14
points), as described previously.3
Blood sampling and laboratory analyses
Venous blood samples were collected into citrate tubes immediately
before the start of thrombolytic and anticoagulation treatment and
after 90 min. In ASSENT-plus, additional blood samples were
obtained after 24 h, 52 h, and 5 days. After centrifugation, plasma
samples were stored frozen at 2708C until further analyses. The
concentration of GDF-15 was determined by a new immunoradiometric assay, as described earlier.14 All biomarker measurements
were performed by investigators who were unaware of patients’
characteristics and outcome. GDF-15 levels at individual time
points did not differ significantly between the randomized treatment groups in ASSENT-2 or ASSENT-plus and were therefore
combined. Patients were stratified on the basis of the two prespecified GDF-15 cut-off levels, 1200 and 1800 ng/L. As previously
shown, 1200 ng/L corresponds to the upper limit of normal in apparently healthy, elderly Swedish individuals.14 In our previous study in
patients with NSTE-ACS, 1200 and 1800 ng/L corresponded to the
lower and upper tertile boundaries and allowed for an identification
of patients at low (,1200 ng/L), intermediate (between 1200 and
1800 ng/L), or high risk (.1800 ng/L).11 cTnT was analysed with a
third-generation assay on Elecsys 2010 (Roche Diagnostics), with a
detection limit of 0.01 mg/L. A concentration of 0.1 mg/L was
used for patient stratification based on previous evaluations of
cTnT at presentation and prognosis in STEMI15,16; this cut-off level
should not be confused with the cut-off concentration used for diagnosis of myocardial infarction. NT-proBNP was determined with a
sandwich immunoassay on Elecsys 2010; the cut-off level of
1000 ng/L was used based on the previous evaluations of NT-proBNP
for risk stratification in STEMI.17
ST-segment resolution
A subgroup of patients (n ¼ 470) without bundle branch block
was monitored for 24 h after presentation by continuous
Statistical analyses
Data are presented as percentages or median (interquartile range),
as appropriate. Spearman’s correlation coefficients were calculated
to evaluate the relations between the levels of GDF-15 and baseline
characteristics. Multiple linear regression was used to evaluate the
association of GDF-15 as the dependent variable with other predictors. The Kaplan–Meier method was used to illustrate the timing of
events during 1 year in relation to the pre-specified strata of
GDF-15 (,1200, 1200–1800 and .1800 ng/L), and statistical assessment was performed using the log-rank test. The differences in proportions in outcome events in the different strata of GDF-15 levels
were judged by Fisher’s exact test. Simple Cox-regression analysis
was used to identify predictors of death during 1 year. All variables
were then tested in a multiple Cox-regression analysis. The Wilcoxon signed-rank test was used to compare the changes in
GDF-15 within the patient group over time. All P-values are
reported two-sided. The alpha-level is 0.05. No adjustments for
multiplicity were made, as the results are to be considered exploratory. All data analyses were performed using the SAS 9.1 statistical
program.
Results
Baseline characteristics and growth-differentiation
factor-15 levels at presentation
The combined patient sample consisted of 221 women
(29.8%) and 520 men (70.2%), with a median age of 67
years (interquartile range, 58–74). Baseline characteristics
are presented in Table 1. The median GDF-15 level at
presentation was 1635 ng/L (interquartile range, 1164–
2309). Two hundred and two patients (27.3%) presented
with GDF-15 levels ,1200 ng/L, the upper limit of normal
in apparently healthy elderly individuals14; GDF-15 levels
were moderately elevated (between 1200 and 1800 ng/L)
in 222 patients (30.0%) and highly elevated (.1800 ng/L)
in 317 patients (42.8%).
Relations between growth-differentiation
factor-15 levels and clinical and biochemical
factors
Increasing levels of GDF-15 at presentation were associated
with increased age, diabetes, and a history of previous myocardial infarction (Table 2). GDF-15 levels were also related
to reduced systolic blood pressure, elevated heart rate,
Killip class .1, NT-proBNP levels, and a cTnT level at presentation .0.1 mg/L (Table 2). By multiple regression analysis that included all patients’ characteristics shown in
Table 2 and trial (ASSENT-2 vs. ASSENT-plus), and used the
natural logarithm of GDF-15 as the dependent variable,
GDF-15 was independently associated with age (P ,
0.001), male gender (P ¼ 0.027), diabetes (P , 0.001),
current smoking (P , 0.001), reduced systolic blood
pressure (P , 0.001), elevated heart rate (P , 0.001),
Killip class .1 (P ¼ 0.013), NT-proBNP (P , 0.001), and a
cTnT level at presentation .0.1 mg/L (P ¼ 0.011). R 2
value of the model was 0.32.
2860
T. Kempf et al.
Table 1 Baseline characteristics
Age (years)
Female gender (%)
Delay time (h)
Current smoking (%)
Hypertension (%)a
Diabetes mellitus (%)
Previous MI (%)
Systolic BP (mmHg)
Heart rate (min21)
Killip class .1 (%)
NT-proBNP (ng/L)
cTnT .0.1 mg/L (%)
GDF-15 , 1200 ng/L (%)
GDF-15 1200–1800 ng/L (%)
GDF-15 . 1800 ng/L (%)
ASSENT-2 (n ¼ 546)
ASSENT-plus (n¼195)
Combined (n ¼ 741)
68 (58–75)
30.4
2.7 (1.9–4.0)
32.4
26.7
10.4
13.6
140 (128–158)
70 (60–81)
13.6
236 (86–686)
23.8
25.6
29.9
44.5
66 (57–73)
28.2
2.6 (1.9–3.6)
28.7
33.8
13.3
17.9
140 (122–154)
70 (60–80)
11.8
169 (72–345)
16.9
31.8
30.3
37.9
67 (58–74)
29.8
2.7 (1.9–3.9)
31.4
28.6
11.2
14.7
140 (125–156)
70 (60–80)
13.1
209 (83–588)
22.0
27.3
30.0
42.8
Data are presented as percentages or median (interquartile range). MI, myocardial infarction; BP, blood pressure.
a
History of hypertension requiring medical therapy.
Table 2 Patients’ characteristics according to growth-differentiation factor-15 levels at presentation
GDF-15 level
Age (years)
Female gender (%)
Delay time (h)
Current smoking (%)
Hypertension (%)a
Diabetes mellitus (%)
Previous MI (%)
Systolic BP (mmHg)
Heart rate (min21)
Killip class .1 (%)
NT-proBNP (ng/L)
cTnT .0.1 mg/L (%)
Spearman’s
correlation
,1200 ng/L (n ¼ 202)
1200–1800 ng/L (n ¼ 222)
.1800 ng/L (n ¼ 317)
r
58 (53–66)
25.7
2.7 (1.7–4.0)
28.7
27.2
4.0
8.4
141 (128–156)
70 (58–80)
5.4
117 (48–307)
17.3
69 (58–74)
32.0
2.6 (1.9–3.8)
34.2
27.9
9.9
12.6
140 (130–160)
66 (59–78)
10.4
163 (78–399)
21.6
72 (64–78)
30.9
2.7 (2.0–3.9)
31.2
30.0
16.7
20.2
140 (122–155)
74 (61–84)
19.9
331 (151–1223)
25.2
0.42
0.05
0.05
0.02
0.03
0.18
0.13
20.09
0.18
0.21
0.37
0.15
P-value
,0.001
0.21
0.16
0.56
0.47
,0.001
,0.001
0.014
,0.001
,0.001
,0.001
,0.001
Data are presented as percentages or median (interquartile range). MI, myocardial infarction; BP, blood pressure.
a
History of hypertension requiring medical therapy.
Growth-differentiation factor-15 levels and
mortality
Growth-differentiation factor-15 in the context of
other markers of increased mortality
GDF-15 levels at presentation were significantly related to
the risk of death during 1-year follow-up (Figure 1). Patients
with GDF-15 levels ,1200 ng/L had a low 1-year mortality
rate of 2.1%. Patients with GDF-15 levels between 1200
and 1800 ng/L had an intermediate 1-year mortality rate
of 5.0%, whereas those with GDF-15 levels .1800 ng/L had
a very high 1-year mortality rate of 14.0% (P , 0.001).
Differences in mortality were observed early after the
index event (Figure 1) and were highly significant after 30
days (Table 3). Receiver operating characteristic curve
analysis further illustrated that GDF-15 is a strong indicator
of mortality with a c-statistic of 0.75 [95% confidence interval (CI), 0.68–0.83]. For comparison, the c-statistics for
NT-proBNP and cTnT were 0.79 (95% CI, 0.71–0.87) and
0.67 (95% CI, 0.59–0.75), respectively.
Several established markers of increased patient risk,
including reduced systolic blood pressure, elevated heart
rate, Killip class .1, NT-proBNP, and a cTnT level at presentation .0.1 mg/L, were associated with an increased
mortality during 1 year by simple Cox-regression analysis
(Table 4). Using multiple Cox-regression analysis, only
GDF-15 (P ¼ 0.005), reduced systolic blood pressure (P ¼
0.007), and age [estimated hazard ratio per 5-year increase,
1.35 (95% CI, 1.08–1.70); P ¼ 0.009] emerged as independent predictors of mortality (Table 4). Inclusion of peak
CK-MB as an additional variable into the multiple Coxregression analysis did not change the overall results (i.e.
GDF-15, reduced systolic blood pressure, and age emerged
as the only independent predictors). GDF-15 remained independently related to mortality when including in-hospital
GDF-15 improves risk stratification in STEMI
2861
associated with a 3.1-fold and a 5.0-fold higher risk of
death during 1 year in patients presenting with a cTnT
level 0.1 or .0.1 mg/L, respectively. Moreover, a GDF-15
level .1800 ng/L was associated with a 4.9-fold higher
risk of death during 1 year in patients with a TIMI score
.3. Patients presenting with a TIMI score 0–3 had a low
1-year mortality rate of 3.0% (13 deaths in 431 patients),
and GDF-15 did not add significant prognostic information
in the low-risk subgroup (Figure 2).
Temporal evolution of growth-differentiation
factor-15 levels in ST-segment elevation myocardial
infarction
Figure 1 Cumulative probability of death according to levels of growthdifferentiation factor-15 at presentation in 730 patients with ST-segment
elevation myocardial infarction enrolled in ASSENT-2 or ASSENT-plus
(P , 0.001 by log-rank test).
Table 3 Mortality rates during follow-up according to
growth-differentiation factor-15 levels at presentation
Mortality [% (n)]
GDF-15 level
,1200 ng/L
1200–1800 ng/L
.1800 ng/L
P-value
24 h
0.5 (1)
0.9 (2)
2.5 (8)
0.16
30 days
1.0 (2)
3.6 (8)
9.8 (31)
,0.001
1 year
2.1 (4)
5.0 (11)
14.0 (44)
,0.001
Patients with STEMI (n ¼ 730) were stratified according to their GDF-15
levels at presentation and followed up to 1 year. Per cent mortality is
shown (patient numbers in parentheses).
P-values were calculated with Fisher’s exact test.
b-blocker and ACE-inhibitor therapy as additional variables
into the multiple Cox-regression analysis (data on in-hospital
treatment were available only in ASSENT-2).
Prognostic utility of growth-differentiation
factor-15 in different patient subgroups
The prognostic utility of GDF-15 was explored across the
clinical spectrum of patients by using the pre-specified
GDF-15 cut-off level of 1800 ng/L. As shown in Figure 2, a
GDF-15 level .1800 ng/L was associated with a significantly
increased risk of death in most investigated subgroups.
Growth-differentiation factor-15 adds prognostic
information to cardiac troponin T, N-terminal
pro-B-type natriuretic peptide, and the TIMI risk
score
A GDF-15 level .1800 ng/L was associated with a 2.2-fold
and a 4.8-fold higher risk of death during 1 year in patients
with an NT-proBNP level 1000 or .1000 ng/L, respectively
(Figure 2). Similarly, a GDF-15 level .1800 ng/L was
After presentation, GDF-15 levels remained elevated for at
least 5 days (Figure 3). In the combined patient population,
72.7, 88.2, 76.0, 76.8, and 69.0% of the patients had GDF-15
levels above the upper limit of normal (1200 ng/L) at presentation, after 90 min, 24 h, 52 h, and after 5 days, respectively. GDF-15 levels measured 90 min after the start of
randomized treatment were significantly higher when compared with the values at presentation; the levels increased
from 1661 ng/L (interquartile range, 1187–2367) to
2179 ng/L (1470–3262) in ASSENT-2 and from 1563 ng/L
(1105–2148) to 3009 ng/L (1789–4554) in ASSENT-plus
(Figure 3). As shown in ASSENT-plus, GDF-15 levels at 24
and 52 h were still slightly higher when compared with the
values at presentation (Figure 3).
To explore whether the peak in circulating GDF-15 concentrations at 90 min was related to successful reperfusion,
the rise in GDF-15 from baseline to 90 min was re-evaluated
in 462 patients, in whom paired baseline and 90 min plasma
samples and continuous vectorcardiography or 12-lead
ECG recordings were available. However, GDF-15 levels
increased to a similar extent after 90 min in patients with
or without ST-segment resolution 50% at 60 min (P ¼
0.28), arguing against a reperfusion-induced rise in circulating GDF-15 levels.
GDF-15 levels at 90 min (n ¼ 711) provided significant
prognostic information, with 1-year mortality rates of 1.2
and 8.4% in patients with GDF-15 levels ,1200 or
1200 ng/L, respectively (P ¼ 0.014). Similar trends were
observed at 24 h, 52 h, or after 5 days, although the
numbers of available samples at these later time points
were too small to allow definitive conclusions (data not
shown).
Discussion
This study demonstrates, for the first time, that GDF-15 is a
biomarker of raised mortality in patients with acute STEMI
that adds independent prognostic information to established
clinical and biochemical markers. Our data show that
patients who present with a GDF-15 concentration within
the normal range (,1200 ng/L) have a rather good prognosis
with a 1-year mortality rate of 2.1%. Conversely, elevated
levels of GDF-15 are indicative of a worse prognosis, with
1-year mortality rates of 5.0 and 14.0% in patients with moderately elevated (between 1200 and 1800 ng/L) or markedly
elevated levels of GDF-15 (.1800 ng/L), respectively. Mortality curves showed a very early separation, indicating
that measurement of GDF-15 may enable risk stratification
2862
T. Kempf et al.
Table 4 Cox-regression analyses for mortality during 1 year in relation to risk markers at presentation
Simple model
Systolic BP (per 10 mmHg decrease)
Heart rate (per 10 b.p.m.)
Killip class .1
NT-proBNPa,b
cTnT (. vs. 0.1 mg/L)
GDF-15a,b
Multiple model
Estimated hazard ratio (95% CI)
P-value
Estimated hazard ratio (95% CI)
P-value
1.22
1.37
2.65
2.89
3.13
2.14
0.001
,0.001
0.001
,0.001
,0.001
,0.001
1.22
1.14
1.21
1.35
1.55
1.55
0.007
0.13
0.61
0.15
0.23
0.005
(1.09–1.37)
(1.23–1.53)
(1.49–4.72)
(2.23–3.75)
(1.88–5.23)
(1.82–2.52)
(1.06–1.40)
(0.96–1.35)
(0.58–2.51)
(0.90–2.01)
(0.76–3.18)
(1.14–2.11)
The multiple model is adjusted for age, gender, delay time, current smoking, hypertension, diabetes mellitus, history of myocardial infarction, and trial
(ASSENT-2 vs. ASSENT-plus).
CI, confidence interval; BP, blood pressure.
a
These variables were not normally distributed and transformed to their natural logarithms before analysis.
b
Hazard ratios refer to 1 SD in the natural logarithmic scale.
for a time window when important management decisions
are made.
Data on the cause of death were not available in our
patient population. However, as previous investigations
have shown that 85% of the mortality during the first year
after STEMI is due to ischaemic heart disease events,19 it
can safely be concluded that the increase in the mortality
in patients with elevated levels of GDF-15 in our study is
related to an increase in the cardiovascular mortality.
Although our data suggest that 1200 and 1800 ng/L are
useful cut-points for risk stratification, this does not imply
that mortality risk is homogeneous within each of these
three strata. In fact, there appeared to be a continuous
relationship between the GDF-15 level and mortality risk
as shown by the multiple Cox-regression analysis.
GDF-15 provided prognostic information on clinically relevant patient subgroups defined according to age, gender,
delay time, cardiovascular risk factors, haemodynamic
status, and TIMI risk score at presentation. Highlighting an
even broader role for GDF-15 as a prognosticator in acute
coronary syndrome, we have recently identified GDF-15 as
an independent prognostic marker also in patients with
NSTE-ACS.11 Patients with STEMI in the present study and
patients with NSTE-ACS in our previous report presented
with similar levels of GDF-15 (the median GDF-15 level
was 13% higher in STEMI when compared with NSTEACS).11 Accordingly, the present study verified prospectively
the same cut-off levels as being useful for risk stratification
in both ST-elevation and non-ST-elevation ACS.
Troponin levels on admission are often used in clinical
practice for risk assessment in patients with STEMI, as
patients presenting already with elevated levels of troponin
are known to have a 2–3-fold higher risk of death during
follow-up.6,7 As shown in the present study, risk stratification based on cTnT on admission was significantly improved
by adding GDF-15. In STEMI, cTnT levels are related to
symptom duration, and most patients have elevated levels
after 6 h.6,20 In contrast, GDF-15 levels at presentation
were not related to symptom duration (delay time) in the
present study. Moreover, GDF-15 carried similar prognostic
information when determined at presentation, 90 min
after admission, and possibly even at later time points.
This parallels our findings in NSTE-ACS, in which GDF-15
levels determined on admission and up to 72 h later
provided similar prognostic information.11 NT-proBNP (or
BNP) levels at presentation are related to long-term mortality in STEMI.5,7,21–23 As shown in the present study, combination of NT-proBNP with GDF-15 allowed us to identify
patients with NT-proBNP levels .1000 ng/L at very high or
relatively low risk. As the information provided by GDF-15
was independent of cTnT and NT-proBNP on admission, and
also independent of peak CK-MB, combination of GDF-15
with these established biomarkers may allow for a more
refined risk stratification.
Among the three biomarkers examined here (cTnT,
NT-proBNP, and GDF-15), GDF-15 emerged as the only
marker that was independently related to mortality by the
multiple Cox-regression analysis. Although our data also
indicate that GDF-15 adds prognostic information to cTnT
and NT-proBNP, it should be pointed out that cTnT and
NT-proBNP appeared to provide better prognostic information when compared with GDF-15, when the three
markers were interpreted individually (as evidenced by the
simple Cox-regression analysis).
GDF-15 levels at presentation were independently related
to age, male gender, diabetes, current smoking, reduced
systolic blood pressure, elevated heart rate, Killip class
.1, and the levels of NT-proBNP and cTnT, suggesting that
GDF-15 carries at least some of the prognostic information
contained in these clinical and biochemical markers of
more severe cardiovascular disease and/or poor prognosis
in patients with STEMI. However, these factors explained
only part of the variation in the GDF-15 levels (R 2 value,
0.32), indicating that additional, as yet undefined, factors
have an impact on the GDF-15 level. In experimental
studies, cardiac GDF-15 expression levels are controlled by
oxidative/nitrosative stress and pro-inflammatory cytokinemediated pathways.9 Circulating levels of GDF-15 may
therefore provide insight into a distinct pathophysiological
process in patients with STEMI.
Circulating levels of GDF-15 remained increased for at
least 5 days after presentation in our study, which is consistent with studies in mice showing that GDF-15 expression
levels in the heart remain elevated for at least 1 week
after an episode of myocardial ischaemia.9 There was a transient peak in GDF-15 levels 90 min after the start of randomized treatment, which was observed in patients with or
without successful reperfusion as judged by ST-segment
GDF-15 improves risk stratification in STEMI
2863
Figure 2 Risk of death during 1 year associated with a growth-differentiation factor-15 level .1800 ng/L at presentation in subgroups of 730 patients with
ST-segment elevation myocardial infarction from ASSENT-2 or ASSENT-plus. The number of patients, mortality rates [% (n)], estimated hazard ratios, and 95%
confidence interval are shown. P-values refer to the comparison between the event rates in patients with growth-differentiation factor-15 levels .1800 vs.
1800 ng/L in each stratum. Data have been adjusted for trial (ASSENT-2 vs. ASSENT-plus). For analyses according to age, delay time, systolic blood pressure
(BP), and heart rate, the median values of the entire study sample were used to create subgroups of comparable size. *Hazard ratios could not be calculated
in the ASSENT-plus cohort (there were zero deaths in 121 patients with GDF-15 levels 1800 ng/L and eight deaths in 74 patients with GDF-15 levels
.1800 ng/L) and in diabetic patients (there were zero deaths in 29 patients with GDF-15 levels 1800 ng/L and nine deaths in 52 patients with GDF-15
levels .1800 ng/L).
resolution at 60 min.24,25 Washout of GDF-15 from the reperfused myocardium is therefore unlikely to explain this peak.
The early rise in circulating GDF-15 levels may be related to
increased synthesis of GDF-15 during the early hours of a
myocardial infarction, or, possibly, to thrombolytic therapy
and plasmin-mediated cleavage and release of GDF-15
from extracellular proGDF-15 stores, a mechanism that is
operative in the processing of TGF-b.26,27 In any case,
measurement of GDF-15 levels at 90 min provided prognostic
information, similar to the levels at presentation.
In conclusion, GDF-15 is a new and independent biomarker
of the risk of death in patients with STEMI. As this is the first
presentation of results on the prognostic importance of
GDF-15 in STEMI, these results need to be further explored
in forthcoming studies before considering implementation in
clinical practice. These studies need to define the prognostic
value and clinical utility of GDF-15 when compared with more
established biomarkers, such as the cardiac troponins and
NT-proBNP. It will be especially important to see whether
GDF-15 levels can help with patient management decisions.
2864
Figure 3 Temporal evolution of growth-differentiation factor-15 levels
during an episode of ST-segment elevation myocardial infarction. Data are
presented as box (25th percentile, median, and 75th percentile) and
whisker (10th and 90th percentiles) plots. The previously defined upper
limit of normal (ULN, 1200 ng/L) is indicated by a dotted line.
Conflict of interest: K.C.W., T.K., and L.W. have filed a patent and
have a contract with Roche Diagnostics to develop a commercial
assay for GDF-15 used for diagnosis and prognosis in cardiovascular
disease.
Funding
This work was supported by grants from the Swedish Heart-Lung
Foundation to L.W. and the German Ministry of Education and
Research to K.C.W. (BMBF, BioChancePlus Program).
References
1. Indications for fibrinolytic therapy in suspected acute myocardial
infarction: collaborative overview of early mortality and major
morbidity results from all randomised trials of more than 1000 patients.
Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group. Lancet 1994;
343:311–322.
2. Lee KL, Woodlief LH, Topol EJ, Weaver WD, Betriu A, Col J, Simoons M,
Aylward P, Van de Werf F, Califf RM. Predictors of 30-day mortality in
the era of reperfusion for acute myocardial infarction. Results from an
international trial of 41,021 patients. GUSTO-I Investigators. Circulation
1995;91:1659–1668.
3. Morrow DA, Antman EM, Charlesworth A, Cairns R, Murphy SA, de
Lemos JA, Giugliano RP, McCabe CH, Braunwald E. TIMI risk score for
ST-elevation myocardial infarction: a convenient, bedside, clinical
score for risk assessment at presentation: An intravenous nPA for treatment of infarcting myocardium early II trial substudy. Circulation 2000;
102:2031–2037.
4. Morrow DA, Antman EM, Parsons L, de Lemos JA, Cannon CP, Giugliano RP,
McCabe CH, Barron HV, Braunwald E. Application of the TIMI risk score for
ST-elevation MI in the National Registry of Myocardial Infarction 3. JAMA
2001;286:1356–1359.
5. Omland T, Aakvaag A, Bonarjee VV, Caidahl K, Lie RT, Nilsen DW,
Sundsfjord JA, Dickstein K. Plasma brain natriuretic peptide as an indicator of left ventricular systolic function and long-term survival after
acute myocardial infarction. Comparison with plasma atrial natriuretic
peptide and N-terminal proatrial natriuretic peptide. Circulation 1996;
93:1963–1969.
6. Ohman EM, Armstrong PW, White HD, Granger CB, Wilcox RG,
Weaver WD, Gibler WB, Stebbins AL, Cianciolo C, Califf RM, Topol EJ.
Risk stratification with a point-of-care cardiac troponin T test in acute
myocardial infarction. GUSTOIII Investigators. Global use of strategies
to open occluded coronary arteries. Am J Cardiol 1999;84:1281–1286.
T. Kempf et al.
7. Bjorklund E, Jernberg T, Johanson P, Venge P, Dellborg M, Wallentin L,
Lindahl B. Admission N-terminal pro-brain natriuretic peptide and its
interaction with admission troponin T and ST segment resolution for
early risk stratification in ST elevation myocardial infarction. Heart
2006;92:735–740.
8. Bootcov MR, Bauskin AR, Valenzuela SM, Moore AG, Bansal M, He XY,
Zhang HP, Donnellan M, Mahler S, Pryor K, Walsh BJ, Nicholson RC,
Fairlie WD, Por SB, Robbins JM, Breit SN. MIC-1, a novel macrophage
inhibitory cytokine, is a divergent member of the TGF-beta superfamily.
Proc Natl Acad Sci USA 1997;94:11514–11519.
9. Kempf T, Eden M, Strelau J, Naguib M, Willenbockel C, Tongers J,
Heineke J, Kotlarz D, Xu J, Molkentin JD, Niessen HW, Drexler H,
Wollert KC. The transforming growth factor-beta superfamily member
growth-differentiation factor-15 protects the heart from ischemia/reperfusion injury. Circ Res 2006;98:351–360.
10. Xu J, Kimball TR, Lorenz JN, Brown DA, Bauskin AR, Klevitsky R,
Hewett TE, Breit SN, Molkentin JD. GDF15/MIC-1 functions as a protective antihypertrophic factor released from the myocardium in association
with SMAD protein activation. Circ Res 2006;98:342–350.
11. Wollert KC, Kempf T, Peter T, Olofsson S, James S, Johnston N, Lindahl B,
Horn-Wichmann R, Brabant G, Simoons ML, Armstrong PW, Califf RM,
Drexler H, Wallentin L. Prognostic value of growth-differentiation
factor-15 in patients with non-ST-segment elevation acute coronary syndrome. Circulation 2007;115:962–971.
12. Van De Werf F, Adgey J, Ardissino D, Armstrong PW, Aylward P, Barbash G,
Betriu A, Binbrek AS, Califf R, Diaz R, Fanebust R, Fox K, Granger C,
Heikkila J, Husted S, Jansky P, Langer A, Lupi E, Maseri A, Meyer J,
Mlczoch J, Mocceti D, Myburgh D, Oto A, Paolasso E, Pehrsson K,
Seabra-Gomes R, Soares-Piegas L, Sugrue D, Tendera M, Topol E,
Toutouzas P, Vahanian A, Verheugt F, Wallentin L, White H. Single-bolus
tenecteplase compared with front-loaded alteplase in acute myocardial
infarction: the ASSENT-2 double-blind randomised trial. Lancet 1999;
354:716–722.
13. Wallentin L, Bergstrand L, Dellborg M, Fellenius C, Granger CB, Lindahl B,
Lins LE, Nilsson T, Pehrsson K, Siegbahn A, Swahn E. Low molecular
weight heparin (dalteparin) compared to unfractionated heparin as an
adjunct to rt-PA (alteplase) for improvement of coronary artery
patency in acute myocardial infarction-the ASSENT Plus study. Eur
Heart J 2003;24:897–908.
14. Kempf T, Horn-Wichmann R, Brabant G, Peter T, Allhoff T, Klein G,
Drexler H, Johnston N, Wallentin L, Wollert KC. Circulating concentrations of growth-differentiation factor 15 in apparently healthy
elderly individuals and patients with chronic heart failure as assessed
by a new immunoradiometric sandwich assay. Clin Chem 2007;53:
284–291.
15. Bjorklund E, Lindahl B, Johanson P, Jernberg T, Svensson AM, Venge P,
Wallentin L, Dellborg M. Admission troponin T and measurement of
ST-segment resolution at 60 min improve early risk stratification in
ST-elevation myocardial infarction. Eur Heart J 2004;25:113–120.
16. Stubbs P, Collinson P, Moseley D, Greenwood T, Noble M. Prognostic
significance of admission troponin T concentrations in patients with myocardial infarction. Circulation 1996;94:1291–1297.
17. Jernberg T, James S, Lindahl B, Johnston N, Stridsberg M, Venge P,
Wallentin L. Natriuretic peptides in unstable coronary artery disease.
Eur Heart J 2004;25:1486–1493.
18. Johanson P, Jernberg T, Gunnarsson G, Lindahl B, Wallentin L, Dellborg M.
Prognostic value of ST-segment resolution—when and what to measure.
Eur Heart J 2003;24:337–345.
19. Stenestrand U, Lindback J, Wallentin L. Long-term outcome of primary
percutaneous coronary intervention vs prehospital and in-hospital thrombolysis for patients with ST-elevation myocardial infarction. JAMA 2006;
296:1749–1756.
20. Katus HA, Remppis A, Neumann FJ, Scheffold T, Diederich KW, Vinar G,
Noe A, Matern G, Kuebler W. Diagnostic efficiency of troponin T measurements in acute myocardial infarction. Circulation 1991;83:902–912.
21. Arakawa N, Nakamura M, Aoki H, Hiramori K. Plasma brain natriuretic
peptide concentrations predict survival after acute myocardial infarction. J Am Coll Cardiol 1996;27:1656–1661.
22. Richards AM, Nicholls MG, Yandle TG, Frampton C, Espiner EA, Turner JG,
Buttimore RC, Lainchbury JG, Elliott JM, Ikram H, Crozier IG, Smyth DW.
Plasma N-terminal pro-brain natriuretic peptide and adrenomedullin:
new neurohormonal predictors of left ventricular function and prognosis
after myocardial infarction. Circulation 1998;97:1921–1929.
23. Mega JL, Morrow DA, De Lemos JA, Sabatine MS, Murphy SA, Rifai N,
Gibson CM, Antman EM, Braunwald E. B-type natriuretic peptide at presentation and prognosis in patients with ST-segment elevation myocardial
GDF-15 improves risk stratification in STEMI
infarction: an ENTIRE-TIMI-23 substudy. J Am Coll Cardiol 2004;44:
335–339.
24. Klootwijk P, Langer A, Meij S, Green C, Veldkamp RF, Ross AM,
Armstrong PW, Simoons ML. Non-invasive prediction of reperfusion and
coronary artery patency by continuous ST segment monitoring in the
GUSTO-I trial. Eur Heart J 1996;17:689–698.
25. Zeymer U, Schroder R, Tebbe U, Molhoek GP, Wegscheider K, Neuhaus KL.
Non-invasive detection of early infarct vessel patency by resolution
of ST-segment elevation in patients with thrombolysis for acute myocardial infarction; results of the angiographic substudy of the Hirudin
2865
for Improvement of Thrombolysis (HIT)-4 trial. Eur Heart J 2001;22:
769–775.
26. Bauskin AR, Brown DA, Junankar S, Rasiah KK, Eggleton S, Hunter M, Liu T,
Smith D, Kuffner T, Pankhurst GJ, Johnen H, Russell PJ, Barret W,
Stricker PD, Grygiel JJ, Kench JG, Henshall SM, Sutherland RL,
Breit SN. The propeptide mediates formation of stromal stores of
PROMIC-1: role in determining prostate cancer outcome. Cancer Res
2005;65:2330–2336.
27. Annes JP, Munger JS, Rifkin DB. Making sense of latent TGFb activation.
J Cell Sci 2003;116:217–224.