Prognostic Value of Growth-Differentiation Factor-15 in Patients With Non–ST-Elevation Acute Coronary Syndrome
Background— Growth-differentiation factor-15 (GDF-15) is a member of the transforming growth factor-β cytokine superfamily that is induced in the heart after ischemia-and-reperfusion injury. Circulating levels of GDF-15 may provide prognostic information in patients with non–ST-elevation acute coronary syndrome.
Methods and Results— Blood samples were obtained on admission from 2081 patients with acute chest pain and either ST-segment depression or troponin elevation who were included in the Global Utilization of Strategies to Open Occluded Arteries (GUSTO)-IV Non–ST-Elevation Acute Coronary Syndrome trial and from a matching cohort of 429 apparently healthy individuals. GDF-15 levels were determined by immunoradiometric assay. Approximately two thirds of patients presented with GDF-15 levels above the upper limit of normal in healthy controls (1200 ng/L); one third presented with levels >1800 ng/L. Increasing tertiles of GDF-15 were associated with an enhanced risk of death at 1 year (1.5%, 5.0%, and 14.1%; P<0.001). By multiple Cox regression analysis, only the levels of GDF-15 and N-terminal pro–B-type natriuretic peptide, together with age and a history of previous myocardial infarction, contributed independently to 1-year mortality risk. Receiver operating characteristic curve analyses further illustrated that GDF-15 is a strong marker of 1-year mortality risk (area under the curve, 0.757; best cutoff, 1808 ng/L). At this cutoff value, GDF-15 added significant prognostic information in patient subgroups defined by age; gender; time from symptom onset to admission; cardiovascular risk factors; previous cardiovascular disease; and the risk markers ST-segment depression, troponin T, N-terminal pro–B-type natriuretic peptide, C-reactive protein, and creatinine clearance.
Conclusions— GDF-15 is a new biomarker of the risk for death in patients with non–ST-elevation acute coronary syndrome that provides prognostic information beyond that provided by established clinical and biochemical markers.
Received July 10, 2006; accepted December 11, 2006.
The clinical manifestations of non–ST-elevation acute coronary syndromes (NSTE-ACS) are based on myocardial ischemia and infarction, which are usually caused by an acute thromboembolic process superimposed on an atherosclerotic plaque fissure in a coronary artery.1 The risk of death and recurrent ischemic events varies considerably across the spectrum of patients with NSTE-ACS. The risk for adverse outcomes can be estimated by clinical background factors; signs of ischemia in the ECG; and by circulating biomarkers of myocardial necrosis (troponin), ventricular wall stress and ischemia (N-terminal pro–B-type natriuretic peptide [NT-proBNP]), inflammatory activity (C-reactive protein [CRP]), and renal dysfunction (creatinine clearance), each of which addresses a separate aspect of NSTE-ACS pathophysiology.2–9
Editorial p 949
Clinical Perspective p 971
Growth-differentiation factor-15 (GDF-15) is a member of the transforming growth factor-β cytokine superfamily that was first identified as macrophage-inhibitory cytokine-1.10 We have recently shown that cultured cardiomyocytes express and secrete GDF-15 via nitrosative stress–dependent signaling pathways when subjected to simulated ischemia and reperfusion.11 Moreover, we have observed in a mouse model of cardiac ischemia-and-reperfusion injury that GDF-15 expression levels rapidly increase in the ischemic area after coronary artery ligation, and remain elevated in the myocardium after reperfusion for several days.11 On the basis of these experimental findings, we hypothesized that circulating GDF-15 levels may offer insight into a distinct pathophysiological axis and provide complementary prognostic information in patients with NSTE-ACS.
Circulating levels of GDF-15 and its relation to other biomarkers and to outcome were therefore studied in patients with NSTE-ACS in the Global Utilization of Strategies to Open Occluded Arteries IV (GUSTO-IV) trial and in a matching cohort of healthy elderly individuals.
This patient data was part of the GUSTO-IV trial in NSTE-ACS patients, which was performed in 24 countries between 1999 and 2000. The detailed design and main results of the trial have been previously published.12,13 Eligible patients were ≥21 years of age, with ≥1 episodes of angina that lasted ≥5 minutes within 24 hours of admission and either a positive cardiac troponin test or ≥0.5 mm of ST-segment depression. Patients were randomly assigned to abciximab or placebo infusion for 24 or 48 hours in addition to standard medical treatment. During 30 days of follow-up, all-cause death and the rate of adjudicated myocardial infarctions were recorded. At 1-year follow-up, only all-cause death information was collected. In a substudy of 399 consecutive patients recruited at Swedish sites, serial citrate plasma samples were available on admission (baseline) and at 24, 48, and 72 hours, which were the first to be analyzed in the present study. On the basis of the significant results in this derivation cohort, it was decided to analyze GDF-15 levels in serum samples taken on admission from another 1682 random patients from the international part of the trial to verify the principle hypothesis from the derivation cohort, ie, that increased circulating levels of GDF-15 provide prognostic information in NSTE-ACS. On the basis of very consistent results in the 2 cohorts, the NSTE-ACS patient materials were then combined to investigate the relation between GDF-15 levels and clinical and biochemical factors and to evaluate the prognostic utility of GDF-15 in the context of, and in combination with, other markers of increased mortality risk.
Citrate plasma samples were obtained from apparently healthy subjects included in the Swedish Women and Men and Ischemic Heart Disease (SWISCH) study. This sample consisted of 429 elderly individuals that were matched for age and gender with another contemporary NSTE-ACS sample included in the Fragmin and Fast Revascularization During Instability in Coronary Artery Disease (FRISC) II trial.6 The detailed design and some biomarker results of the SWISCH trial have been published.14 Subjects with an abnormal resting 12-lead ECG, cardiovascular medication, established cardiovascular disease, other chronic disease, or acute illness were excluded from the control sample. All SWISCH participants were required to demonstrate normal levels of creatinine, blood glucose, and hemoglobin, and normal white blood cell and platelet counts.
Serum and plasma samples were stored at −70°C. All subjects provided written informed consent, and the study was approved by the ethics committees of all participating centers.
GDF-15 Immunoradiometric Assay
The concentration of GDF-15 in serum and plasma samples was determined by an immunoradiometric assay with a polyclonal, GDF-15 affinity chromatography-purified, goat anti-human GDF-15 IgG antibody from R&D Systems (Minneapolis, Minn) (AF957).15 In each experiment, a standard curve was generated with recombinant human GDF-15 from R&D Systems (957-GD/CF). The results with new batches of recombinant GDF-15 protein were tested in standard plasma samples, and any deviation >10% was corrected by the introduction of an adjustment factor for this assay. GDF-15 measurements in serum and citrate plasma samples from the same individual yielded virtually identical results.15 The detection limit of the assay was 20 ng/L. All GDF-15 measurements were performed at Hannover Medical School by investigators who were not aware of patients’ characteristics and outcomes.
Troponin T levels were determined by a third-generation assay on an Elecsys 2010 analyzer (Roche Diagnostics, Mannheim, Germany) with a detection limit 0.01 μg/L. NT-proBNP levels were determined with an immunoassay on an Elecsys 2010 analyzer with a detection limit of 20 ng/L. CRP concentrations were measured with a chemiluminescent enzyme–labeled immunometric assay (Immulite CRP, Diagnostic Products Corp, Los Angeles, Calif) with a detection limit of 0.1 mg/L. Creatinine levels were analyzed at the participating study centers, and creatinine clearance was calculated according to the Cockcroft-Gault equation.
Baseline characteristics are presented as numbers and proportions. Continuous data are given as median (25th to 75th percentiles). Comparisons of continuous or categorical variables between cases and controls were evaluated by nonparametric Mann-Whitney U test or Fisher exact test, respectively. To evaluate the relations between the levels of GDF-15 and baseline characteristics, and the levels of troponin T, NT-proBNP, CRP, and creatinine clearance within tertiles of GDF-15, both Spearman’s rank-correlation (ρ) coefficients between GDF-15 level and these factors and the Cochran-Armitage trend test for relations between proportions in several groups were used. ρ Coefficients and the Wilcoxon signed rank test were used to compare the changes of GDF-15 within the patient group over time. Multiple linear regression was used to evaluate the association of GDF-15 as the dependent variable with other predictors. The differences in proportions in outcome events in tertiles of GDF-15 levels were judged by Fisher exact test. The Kaplan-Meier method was used to illustrate the timing of events during follow-up in relation to tertiles of GDF-15, and statistical assessment was performed with the log-rank test. Simple Cox regression analyses were used to identify predictors of death at 1 year. All variables were then tested in a multiple Cox regression analysis. For additional comparison of the prognostic values of GDF-15, troponin T, NT-proBNP, CRP, and creatinine clearance with regard to death at 1 year, receiver operating characteristic (ROC) curves were generated, and the areas under the curves (AUC) were calculated. The Hosmer-Lemeshow test was performed to evaluate the fitness of the models derived from multiple Cox regression and ROC curve analyses; all probability values from these tests were >0.1, which indicates that the models were adequate (data not shown). Data analyses were performed with the SAS 9.1 statistical program (SAS Institute, Inc, Cary, NC).
The authors had full access to and take responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
GDF-15 Levels in Apparently Healthy Elderly Control Subjects
The control group comprised 288 men (67.1%) and 141 women (32.9%) with a median age of 65 years (25th to 75th percentiles: 59 to 71 years), of whom 14.5% were current smokers. The median GDF-15 level in this sample was 762 (600 to 959) ng/L; 1191 ng/L marked the 90th percentile, and the upper limit of normal was therefore rounded to 1200 ng/L.
GDF-15 Levels on Admission in Patients With NSTE-ACS (Derivation Cohort)
The derivation cohort comprised 258 men (64.7%) and 141 women (35.3%) with a median age of 69 (59 to 76) years. The baseline characteristics of this patient sample are presented in Table 1. The NSTE-ACS patients displayed significantly (P<0.001) higher GDF-15 levels as compared with the healthy controls, the median was 1499 (1151 to 2203) ng/L, and 1238 ng/L and 1896 ng/L marked the 33rd and 66th percentiles, respectively. Accordingly, about two thirds of the patients had GDF-15 levels above the upper limit of normal (1200 ng/L) in healthy controls. Because this upper limit of normal corresponded to the lower tertile in patients, the patient data were stratified in tertiles (rounded cutoff limits, 1200 and 1800 ng/L) when related to outcome.
Temporal Evolution of GDF-15 Levels in NSTE-ACS (Derivation Cohort)
Serum samples were available on admission and after 24, 48, and 72 hours in the derivation cohort, which allowed us to study the temporal evolution of circulating GDF-15 levels during an episode of NSTE-ACS. Although there was a slight increase, GDF-15 levels stayed within the same range during this time interval (Table 2). There was a very limited intra-individual variation over time, as shown by the close correlation between the GDF-15 levels on admission and at later time points (Table 2); 69.4%, 70.4%, 72.2%, and 71.9% of the patients had GDF-15 levels above the upper limit of normal on admission and at 24, 48, and 72 hours, respectively. GDF-15 levels on admission and increases in GDF-15 levels during the first 72 hours did not differ significantly between patients randomized to abciximab or placebo (not shown).
GDF-15 Levels and Death in NSTE-ACS (Derivation Cohort)
The risk of death in patients from the derivation cohort increased markedly with increasing levels of GDF-15 on admission, with 30-day mortality rates of 0% (0 of 122), 0.8% (1 of 127), and 5.3% (8 of 150) (P=0.005) and 1-year mortality rates of 0.8% (1 of 122), 2.4% (3 of 127), and 12.7% (19 of 150) (P<0.001) in the respective tertiles. Notably, GDF-15 levels measured on admission or at later time points provided similar prognostic information on 1-year mortality risk, although it seemed that the GDF-15 levels within the initial 24 hours had the highest predictive value (Table 3).
GDF-15 Levels on Admission and Death in NSTE-ACS (Validation Cohort)
The validation cohort comprised 1056 men (62.8%) and 626 women (37.2%) with a median age of 66 (56 to 74) years. Baseline characteristics are presented in Table 1 and show some differences between the derivation cohort from 1 country and the validation cohort from a multitude of countries. The median level of GDF-15 in the validation cohort was 1434 (1035 to 2078) ng/L, which was significantly higher than in the control cohort (P<0.001) and somewhat lower than in the derivation cohort (P=0.025) (Table 1). With the tertile boundaries established in the derivation cohort, there was a similar distribution, with 35.1%, 31.9%, and 32.9% of the patients in the validation cohort presenting with GDF-15 levels <1200 ng/L, between 1200 and 1800 ng/L, and >1800 ng/L, respectively (Table 1). Also in this cohort, increasing levels of GDF-15 were strongly related to death, with 30-day mortality rates of 0.7% (4 of 591), 2.2% (12 of 537), and 4.0% (22 of 554) (P<0.001) and 1-year mortality rates of 1.7% (10 of 591), 5.6% (30 of 537), and 14.4% (80 of 554) in the respective tertiles (P<0.001).
Mortality rates in the combined data of 2081 NSTE-ACS patients are illustrated in Figure 1. Mortality curves showed an early separation, with mortality rates of 0.6% (4 of 713), 2.0% (13 of 664), and 4.3% (30 of 704) at 30 days (P<0.001) and 1.5% (11 of 713), 5.0% (33 of 664), and 14.1% (99 of 704) at 1 year (P<0.001) in the respective tertiles.
GDF-15 Levels and the Risk of Recurrent Myocardial Infarction in NSTE-ACS
The risk of the composite end point of death or recurrent myocardial infarction at 30 days in the respective tertiles was 7.4% (9 of 122), 9.4% (12 of 127), and 12.0% (18 of 150) in the derivation cohort (P=0.46); 4.6% (27 of 591), 6.3% (34 of 537), and 10.5% (58 of 554) in the validation cohort (P<0.001); and 5.0% (36 of 713), 6.9% (46 of 664), and 10.8% (76 of 704) in the combined patient data (P<0.001). This relationship was mainly driven by the association of GDF-15 with death. The risk of a subsequent myocardial infarction within 30 days in the respective tertiles was 7.4% (9 of 122), 8.7% (11 of 127), and 8.7% (13 of 150) in the derivation cohort (P=0.92); 4.2% (25 of 591), 4.8% (26 of 537), and 6.9% (38 of 554) in the validation cohort (P=0.12); and 4.8% (34 of 713), 5.6% (37 of 664), and 7.2% (51 of 704) in the combined data (P=0.14).
Relation Between GDF-15 Levels and Clinical and Biochemical Factors in NSTE-ACS
In the combined sample of 2081 patients with NSTE-ACS, increasing tertiles of GDF-15 on admission were positively associated with age, female gender, history of hypertension, and diabetes (Table 4). GDF-15 levels were also associated with previous manifestations of cardiac disease; current angiotensin-converting enzyme inhibitor therapy; and markers of ongoing ischemia and necrosis, myocardial dysfunction, and inflammation, as indicated by ST-segment depression and the levels of troponin T, NT-proBNP, and CRP. GDF-15 levels were inversely related to current smoking and creatinine clearance (Table 4). In a multiple linear regression analysis that used the natural logarithm of GDF-15 as the dependent variable, the following factors were significantly associated with GDF-15: age (P<0.001), male gender (P<0.001), time from symptom onset to admission (P=0.006; inverse relation), current smoking (P<0.001), diabetes (P<0.001), history of heart failure (P<0.001), ST-segment depression (P=0.050), NT-proBNP (P<0.001), CRP (P<0.001), and creatinine clearance (P<0.001; inverse relation). The R2 value of this model was 0.43. There was no independent relation to the level of troponin T on admission (P=0.44).
GDF-15 in the Context of Other Biomarkers of Increased Mortality Risk in NSTE-ACS
By simple Cox regression analysis, the levels of troponin T, NT-proBNP, CRP, creatinine clearance, and GDF-15 were all related to 1-year mortality in the combined NSTE-ACS sample (Table 5). With multiple Cox regression analysis, GDF-15 (P<0.001) and NT-proBNP (P=0.009) emerged as the only independent predictors of 1-year mortality after adjustment for demographic and clinical variables (Table 5). Besides GDF-15 and NT-proBNP, only age (hazard ratio per 5-year increase, 1.22; 95% confidence interval [CI], 1.06 to 1.40; P=0.006) and a history of myocardial infarction (hazard ratio, 1.71; 95% CI, 1.14 to 2.56; P=0.009) emerged as independent predictors in the multiple model. Among these independent risk indicators, GDF-15 was the strongest predictor of death. The results were unchanged when a backward stepwise approach was used or when stratification of GDF-15 levels and the other biomarkers in tertiles was performed. Also, when added one at a time in stepwise selection, the only variables that remained significant were GDF-15, NT-proBNP, age, and previous myocardial infarction.
ROC Curve Analyses of GDF-15 and Other Biomarkers
ROC curve analyses of GDF-15 and other individual biomarkers further illustrated that GDF-15 is a strong biochemical indicator of mortality risk with an AUC of 0.757, as compared with NT-proBNP (AUC=0.735), creatinine clearance (AUC=0.728), CRP (AUC=0.629), and troponin T (AUC=0.620) (Figure 2). The best GDF-15 level for the prediction of 1-year mortality was 1808 ng/L (sensitivity, 68.5%; specificity, 68.8%).
Combination of GDF-15 With Other Markers of Prognosis in NSTE-ACS
The multiple Cox regression analysis showed that GDF-15 and NT-proBNP were the only biomarkers with an independent prognostic importance for death. Tertiles of these markers were therefore combined and used on top of the independent clinical predictor of death, previous myocardial infarction (Figure 3). The results illustrate that a combination of increasing tertiles of GDF-15 and NT-proBNP levels provided additive prognostic information and identified strata of patients with 1-year mortality rates that ranged from 0.3% to 13.7% in patients without a history of a previous myocardial infarction and from 4.7% to 23.6% in patients with a previous myocardial infarction (Figure 3). These relationships were further modified only by the age of the patient, with a 22% relative change in mortality risk for every increase in 5 years of age (see above).
To investigate the prognostic utility of GDF-15 across the spectrum of patients with NSTE-ACS and to explore whether GDF-15 adds prognostic information to that provided by established markers of mortality risk, the prognostic value of GDF-15 was assessed in several patient subgroups (Figure 4). A GDF-15 level >1808 ng/L (best cutoff value determined by ROC curve analysis) was associated with an increased risk of death in patients ≤66 years (median) and >66 years of age. GDF-15 was prognostic in males and in females; in current smokers and nonsmokers; and in patients with or without hypertension, hypercholesterolemia, diabetes, previous myocardial infarction, or a history of heart failure. GDF-15 added prognostic information in patients assigned to abciximab or placebo. The prognostic utility of GDF-15 was maintained in patients with a delay time ≤9.8 hours (median) or >9.8 hours, in patients with or without ST-segment depression ≥0.5 mm, in patients with troponin T levels on admission ≤0.01 μg/L or >0.01 μg/L, in patients with a creatinine clearance ≤63 mL/min (median) or >63 mL/min, in patients with CRP levels ≤10 mg/L or >10 mg/L, and in patients with NT-proBNP levels ≤1000 ng/L or >1000 ng/L (Figure 4).
This study demonstrates for the first time that GDF-15 is a powerful biomarker of cardiovascular disease and increased risk of death in patients with NSTE-ACS. Measurement of GDF-15 adds substantial and independent prognostic information beyond that provided by established clinical and biochemical markers such as troponin T, NT-proBNP, CRP, and renal function.
Circulating GDF-15 levels were significantly elevated in patients with NSTE-ACS and closely related to 1-year mortality after the acute event. This was first observed in a derivation cohort of 399 patients from Sweden and subsequently verified in an independent international cohort of 1682 patients. About two thirds of NSTE-ACS patients in our study had levels >1200 ng/L, the upper limit of normal in apparently healthy elderly subjects, and about one third had levels >1800 ng/L. Among patients within the normal range of GDF-15, the 1-year mortality rate was low at 1.5%. In patients with a moderate elevation of GDF-15 (1200 to 1800 ng/L), the 1-year mortality rate was intermediate at 5.0%, and in those patients with a marked elevation (>1800 ng/L), the 1-year mortality rate was very high at 14.1%. The differences in mortality were observed early after the index event and were highly significant after 1 month.
Although the median GDF-15 level increased slightly after admission, it remained remarkably stable within each patient during 72 hours of observation. GDF-15 levels determined in individual patients at different occasions were closely correlated and gave similar prognostic information about mortality risk, which shows that a single measurement of GDF-15, obtained at any time point during the first 24 hours and maybe even 72 hours after admission, will provide comparable prognostic information in patients with NSTE-ACS.
GDF-15 levels on admission were strongly and independently related to age, male gender, current smoking, diabetes, history of heart failure, cardiac dysfunction (NT-proBNP), inflammatory activity (CRP), and renal dysfunction (creatinine clearance), which indicates that elevated levels of GDF-15 integrate several important clinical and biochemical indicators of more severe cardiovascular disease and/or poor prognosis in patients with NSTE-ACS. The inverse relation of the GDF-15 level to the time from symptom onset to admission (delay time) might be related to the association between GDF-15 and the severity of disease/symptoms, which would result in a shorter delay in patients with higher GDF-15 levels. Together, these factors explained only part of the variation in the GDF-15 levels on admission (R2 value 0.43), which indicates that additional, as-yet-undefined factors have an impact on circulating GDF-15 levels.
Of all these interrelated variables, GDF-15 emerged as the strongest predictor of raised 1-year mortality risk. Beyond GDF-15, only age, previous myocardial infarction, and the levels of NT-proBNP added independent prognostic information. The combined information from these 4 variables supplanted the prognostic contribution from other markers of raised mortality risk in NSTE-ACS, including indicators such as diabetes,5,9 history of heart failure,3,5 inflammatory activity (CRP),4,9 renal dysfunction,5,7 ST-segment depression,8,9 and troponin T.2,6,8,9 ROC curve analyses further illustrated that GDF-15 is a strong biochemical marker of 1-year mortality risk, with an AUC of 0.757. The best cutoff value for the identification of patients with a high 1-year mortality risk was 1808 ng/L, which was almost identical to the upper tertile boundary. At this cutoff, GDF-15 provided prognostic information in clinically relevant patient subgroups defined according to age, gender, underlying cardiovascular risk factors, and history of previous cardiovascular disease. GDF-15 provided prognostic information in patients who presented early or late after symptom onset, and in patients randomized to abciximab or placebo. Notably, GDF-15 added prognostic information to several established indicators of raised mortality risk in patients with NSTE-ACS, including ST-segment depression ≥0.5 mm,8,9 a troponin T level >0.01 μg/L on admission,6,8,9 renal dysfunction,5,7 CRP levels >10 mg/L,9 and NT-proBNP levels >1000 ng/L.16 Besides the ROC curve analysis–derived cutoff value, which identified high-risk patients, a cutoff value <1200 ng/L (the upper limit of normal) was useful for identification of low-risk patients in the aforementioned patient subgroups (not shown).
Despite its strong and independent relation to 1-year mortality risk, GDF-15 was not related to the risk of early reinfarction in this cohort of patients with NSTE-ACS. However, it should be emphasized that the follow-up for reinfarction was rather short in GUSTO-IV (30 days). Indeed, in our ongoing studies in other NSTE-ACS patient cohorts with longer follow-up, GDF-15 also emerges as an independent predictor of the risk of reinfarction.
The GDF-15 level was only weakly related to the magnitude of ongoing ischemia and infarction, as evidenced by the weak independent relation to ST-segment depression and lack of independent relation to the level of troponin T in the acute phase of NSTE-ACS. Also, there was only a slight increase of GDF-15 levels during the acute phase of NSTE-ACS, which suggests that the prognostic value of GDF-15 might be unrelated to the extent of myocardial damage. These findings are consistent with studies in mice that show that GDF-15 expression levels increase to a similar extent in the ischemic myocardium after transient coronary artery ligation (limited damage) or permanent coronary ligation (extensive necrosis) and that GDF-15 levels remain elevated in the heart for at least 1 week after an ischemic insult.11 Notably, GDF-15 appears to be actively involved in the pathophysiology of cardiac ischemia–reperfusion injury by promotion of antiapoptotic effects in cardiomyocytes.11 In cultured cardiomyocytes, GDF-15 expression is strongly upregulated by several stressors, including nitric oxide, nitrosative stress, and inflammatory cytokines,11 all of which are putative signaling mediators in the ischemic and reperfused heart.17 Together, these experimental data suggest that GDF-15 may provide insight into a distinct pathophysiological axis in ACS.
In conclusion, GDF-15 is a new biomarker of the risk of death in patients with NSTE-ACS that provides prognostic information beyond that provided by established clinical and biochemical markers. Future studies need to explore whether GDF-15 can help with therapeutic decision making in NSTE-ACS.
Sources of Funding
This work was supported by grants from the German Research Foundation to Dr Wollert (WO 552/2–4) and the Swedish Heart-Lung Foundation to Dr Wallentin. The biobank of the GUSTO-IV trial was established by a grant from Centocor Inc. to Dr Wallentin.
Drs Wollert, Kempf, Drexler, and Wallentin have filed a patent and have a contract with Roche Diagnostics (Mannheim, Germany) to develop a commercial assay for GDF-15 used for diagnosis and prognosis in cardiovascular disease. The other authors report no conflicts.
Ohman EM, Armstrong PW, Christenson RH, Granger CB, Katus HA, Hamm CW, O’Hanesian MA, Wagner GS, Kleiman NS, Harrell FE Jr, Califf RM, Topol EJ. Cardiac troponin T levels for risk stratification in acute myocardial ischemia. GUSTO IIA Investigators. N Engl J Med. 1996; 335: 1333–1341.
James SK, Lindahl B, Siegbahn A, Stridsberg M, Venge P, Armstrong P, Barnathan ES, Califf R, Topol EJ, Simoons ML, Wallentin L. N-terminal pro-brain natriuretic peptide and other risk markers for the separate prediction of mortality and subsequent myocardial infarction in patients with unstable coronary artery disease: a Global Utilization of Strategies To Open occluded arteries (GUSTO)-IV substudy. Circulation. 2003; 108: 275–281.
Lindahl B, Toss H, Siegbahn A, Venge P, Wallentin L. Markers of myocardial damage and inflammation in relation to long-term mortality in unstable coronary artery disease. FRISC Study Group. Fragmin during Instability in Coronary Artery Disease. N Engl J Med. 2000; 343: 1139–1147.
Anavekar NS, McMurray JJ, Velazquez EJ, Solomon SD, Kober L, Rouleau JL, White HD, Nordlander R, Maggioni A, Dickstein K, Zelenkofske S, Leimberger JD, Califf RM, Pfeffer MA. Relation between renal dysfunction and cardiovascular outcomes after myocardial infarction. N Engl J Med. 2004; 351: 1285–1295.
Cannon CP, Weintraub WS, Demopoulos LA, Vicari R, Frey MJ, Lakkis N, Neumann FJ, Robertson DH, DeLucca PT, DiBattiste PM, Gibson CM, Braunwald E. Comparison of early invasive and conservative strategies in patients with unstable coronary syndromes treated with the glycoprotein IIb/IIIa inhibitor tirofiban. N Engl J Med. 2001; 344: 1879–1887.
Lagerqvist B, Diderholm E, Lindahl B, Husted S, Kontny F, Stahle E, Swahn E, Venge P, Siegbahn A, Wallentin L. FRISC score for selection of patients for an early invasive treatment strategy in unstable coronary artery disease. Heart. 2005; 91: 1047–1052.
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-β superfamily. Proc Natl Acad Sci U S A. 1997; 94: 11514–11519.
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-β superfamily member growth-differentiation factor-15 protects the heart from ischemia/reperfusion injury. Circ Res. 2006; 98: 351–360.
Ottervanger JP, Armstrong P, Barnathan ES, Boersma E, Cooper JS, Ohman EM, James S, Topol E, Wallentin L, Simoons ML. Long-term results after the glycoprotein IIb/IIIa inhibitor abciximab in unstable angina: one-year survival in the GUSTO IV-ACS (Global Use of Strategies To Open Occluded Coronary Arteries IV-Acute Coronary Syndrome) Trial. Circulation. 2003; 107: 437–442.
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.
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.
The risk of early and late death varies considerably across the spectrum of patients with non–ST-elevation acute coronary syndrome, so risk stratification is a critical first step in the evaluation of these patients. Here we report that growth-differentiation factor-15 (GDF-15), a distant member of the transforming growth factor-β cytokine superfamily, is a new biomarker of the risk of death in non–ST-elevation acute coronary syndrome. We measured circulating GDF-15 levels on admission in 2081 patients from the Global Utilization of Strategies to Open Occluded Arteries IV (GUSTO-IV) trial and observed that two thirds of non–ST-elevation acute coronary syndrome patients presented with GDF-15 levels above the upper limit of normal in a matching control cohort. In patients within the normal range of GDF-15 (<1200 ng/L), 1-year mortality rate was low at 1.5%. In patients with a moderate elevation of GDF-15 (1200 to 1800 ng/L), 1-year mortality rate was intermediate at 5.0%, whereas in patients with a marked elevation (>1800 ng/L), 1-year mortality rate was very high at 14.1%. The differences in mortality rates were observed early after the index event and were highly significant after 1 month. In multiple Cox regression analyses, GDF-15 emerged as the strongest independent predictor of 1-year death. GDF-15 provided prognostic information in clinically relevant patient subgroups defined by age, gender, underlying cardiovascular risk factors, or previous cardiovascular disease. More importantly, GDF-15 added prognostic information to that provided by established indicators of increased mortality risk, such as ST-segment depression, troponin T, renal dysfunction, C-reactive protein, and N-terminal pro–B-type natriuretic peptide. Thus, GDF-15 appears to be a promising new tool for early risk stratification in patients with non–ST-elevation acute coronary syndrome.