Growth Differentiation Factor-15 in Idiopathic Pulmonary Arterial Hypertension

Transcription

1 Growth Differentiation Factor-15 in Idiopathic Pulmonary Arterial Hypertension Nils Nickel 1 *, Tibor Kempf 2 *, Heike Tapken 2, Jörn Tongers 2, Florian Laenger 3, Ulrich Lehmann 3, Heiko Golpon 1, Karen Olsson 1, Martin R. Wilkins 4, J. Simon R. Gibbs 5, Marius M. Hoeper 1, and Kai C. Wollert 2 1 Department of Respiratory Medicine, 2 Department of Cardiology and Angiology, and 3 Department of Pathology, Hannover Medical School, Hannover, Germany; 4 Department of Experimental Medicine and Toxicology, Imperial College London, London, United Kingdom; and 5 National Heart and Lung Institute, Hammersmith Campus, Imperial College London, London, United Kingdom Rationale: Growth-differentiation factor (GDF)-15 is a stress-responsive, transforming growth factor-b related cytokine. Circulating levels of GDF-15 provide independent prognostic information in patients with acute pulmonary embolism and chronic left-sided heart failure. Objectives: To assess the prognostic value of GDF-15 in idiopathic pulmonary arterial hypertension. Methods: GDF-15 levels were determined in 76 treatment-naive patients at the time of baseline right heart catheterization. Patients were monitored for a median (range) of 48 (0 101) months (first cohort). Twenty-two additional patients were studied at baseline and 3 to 6 months after initiation of therapy (second cohort). Measurements and Main Results: Fifty-five percent of the patients in the first cohort presented with GDF-15 levels above 1,200 ng/l, the previously defined upper reference limit. The risk of death or transplantation at 3 years was 15 and 44% in patients with GDF-15 levels below or above 1,200 ng/l, respectively (P ). Elevated levels of GDF-15 were associated with increased mean right atrial and pulmonary capillary wedge pressures, a lower mixed venous oxygen saturation (Sv O2 ), and higher levels of uric acid and N-terminal pro-brain natriuretic peptide (NT-proBNP). After adjustment for hemodynamic and biochemical variables, GDF-15 remained an independent predictor of adverse outcomes (P ). GDF-15 provided prognostic information in clinically relevant patient subgroups, and added prognostic information to hemodynamic variables and NT-proBNP. Changes in GDF-15 over time in the second cohort were related to changes in NT-proBNP (P ) and inversely related to changes in Sv O2 (P, 0.001). Conclusions: GDF-15 is a promising new biomarker in idiopathic pulmonary arterial hypertension. Keywords: idiopathic pulmonary artery hypertension; biomarker; risk stratification Growth differentiation factor (GDF)-15 is a distant member of the transforming growth factor-b cytokine superfamily that was originally cloned from activated macrophages (1). In most tissues, GDF-15 is weakly expressed under normal conditions (2). In pathologic conditions involving tissue hypoxia, inflammation, or enhanced oxidative stress, however, expression levels of GDF-15 may sharply increase (3, 4). In mice, GDF-15 expression levels in the heart are significantly induced after (Received in original form February 8, 2008; accepted in final form June 19, 2008) Supported by a grant from the German Ministry of Education and Research (BioChancePlus) to K.C.W. * These authors contributed equally and share first authorship. These authors contributed equally and share senior authorship. Correspondence and requests for reprints should be addressed to Kai C. Wollert, M.D., Klinik für Kardiologie und Angiologie, Medizinische Hochschule Hannover, Carl- Neuberg-Str. 1, Hannover, Germany. wollert.kai@mh-hannover.de Am J Respir Crit Care Med Vol 178. pp , 2008 Originally Published in Press as DOI: /rccm OC on June 19, 2008 Internet address: AT A GLANCE COMMENTARY Scientific Knowledge on the Subject Growth differentiation factor (GDF)-15 is a stress-responsive member of the transforming growth factor-b cytokine superfamily, and an emerging biomarker in patients with cardiovascular disease. Its clinical significance in idiopathic pulmonary arterial hypertension (IPAH) is uncertain. What This Study Adds to the Field Circulating GDF-15 levels are elevated and independently related to the risk of death or transplantation in patients with IPAH. The information provided by GDF-15 is additive to that of established hemodynamic and biochemical markers, including N-terminal pro-brain natriuretic peptide. coronary artery ligation, transverse aortic banding, and in transgenic models of cardiomyopathy (4, 5). Consistent with these experimental findings, circulating levels of GDF-15 are increased in patients with acute coronary syndrome (6, 7) and chronic left-sided heart failure (8). We have recently observed that GDF-15 levels are also elevated in patients with acute pulmonary embolism (9), suggesting that GDF-15 may respond to right ventricular overload as well. In all of these previous studies, the levels of GDF-15 were independently associated with adverse clinical outcomes; notably, GDF-15 added prognostic information not only to clinical parameters but also to established biomarkers, including N-terminal pro-brain natriuretic peptide (NT-proBNP), suggesting that GDF-15 may provide insight into a distinct pathophysiologic process. Idiopathic pulmonary arterial hypertension (IPAH) is a disease characterized by progressive pulmonary vascular remodeling, which results in right ventricular overload and failure if not treated effectively (10). In addition to clinical and hemodynamic assessment, biomarkers are increasingly used in IPAH as tools to characterize disease severity and response to therapy. Among the biomarkers used predominantly in IPAH are uric acid and NT-proBNP. Circulating levels of these biomarkers correlate with disease severity and have prognostic implications, although the use of these biomarkers to define individual prognosis has not been well established (11 13). Considering that GDF-15 may be a marker of multiple cellular stress pathways in the heart and in extracardiac tissues, and that GDF-15 is emerging as a prognostic biomarker in patients with left and right ventricular pathologies, we hypothesized (1) that GDF-15 may provide prognostic information in IPAH and (2) that a combination of GDF-15 with more established hemodynamic and biochemical markers may enhance risk stratification in these patients. Some of the results of these studies have been previously reported in the form of an abstract (14).

2 Nickel, Kempf, Tapken, et al.: GDF-15 in Idiopathic Pulmonary Arterial Hypertension 535 METHODS Patients The first patient cohort consisted of 76 nonselected, treatment-naive patients with IPAH referred to Hannover Medical School (n 5 67) or to Imperial College London (n 5 9) between 1999 and The second patient cohort included 22 consecutive patients with IPAH studied prospectively at Hannover Medical School between 2007 and In contrast to the first cohort, these patients had systematic follow-up right heart catheterizations at baseline and 3 to 6 months after the introduction of medical therapy. In all patients, the diagnosis of IPAH was based on standard criteria with confirmation by right heart catheterization and exclusion of other forms of pulmonary hypertension by various laboratory studies, echocardiography, pulmonary function testing, chest X-ray, ventilation perfusion scanning, chest computed tomography angiography, and/or pulmonary angiography, if necessary. Baseline blood samples were collected in both cohorts at the time of the initial right heart catheterization before the initiation of any PAH-specific therapy. In the second cohort, an additional blood sample was obtained during right heart catheterization after 3 to 6 months. All blood samples were immediately cooled on ice and centrifuged at 48C. Serum samples were divided into aliquots and stored frozen at 2808C. All patients gave written, informed consent for storage of serum samples and future biomarker analyses at the time when the samples were obtained, an approach that was approved by the institutional review boards of both participating centers. Laboratory Analyses GDF-15 serum concentrations were measured by immunoradiometric assay using a polyclonal GDF-15 affinity-chromatography-purified, goat anti-human GDF-15 IgG antibody (AF957) from R&D Systems (Minneapolis, MN) as recently described (15). The assay has a linear range from 200 to 50,000 ng/l and a detection limit of 20 ng/l (15). NT-proBNP levels were determined by a sandwich immunoassay on an Elecsys 2010 instrument (Roche Diagnostics, Mannheim, Germany). Creatinine and uric acid measurements were performed at the participating study centers using standard laboratory techniques. All biomarker measurements were performed by investigators blinded to patients characteristics and outcome. Follow-up After blood sampling, patients from the first cohort were monitored by regular outpatient assessments for a median of 48 months (range, mo). Survival status was censored on July 30, No patient was lost to follow-up. Death or lung transplantation was defined a priori as the primary composite endpoint of the study. Statistical Analyses Data are presented as absolute numbers, percentages, mean (6SD), or median (interquartile range). Proportions were compared using the x 2 test. The relations between GDF-15 and baseline variables were assessed by Spearman rank correlation coefficients. The Wilcoxon signed rank test was used to compare changes in hemodynamic parameters, sixminute-walk distance, and biomarkers over time. Changes in these parameters in relation to changes in GDF-15 over time were assessed by Spearman rank correlation coefficients. For comparison of the prognostic values of GDF-15, uric acid, NT-proBNP, and selected hemodynamic parameters, receiver operating characteristic (ROC) curves were generated, and the areas under the curves (c-statistics) were calculated. Kaplan-Meier plots were used to illustrate the timing of events during follow-up in relation to baseline GDF-15 levels, alone or in combination with NT-proBNP, and statistical assessment was performed by the log-rank test. The differences in proportions in outcome events at 1 and 3 years in different strata of GDF-15 levels were judged by Fisher exact test. Simple Cox regression analyses were performed to identify predictors of death or transplantation during follow-up. All variables with a P value of less than 0.05 were then tested in a stepwise forward Cox regression analyses; variables were entered at a P value of less than 0.05 and removed at a P value of greater than GDF-15 and NT-proBNP values were not normally distributed, as shown by the Kolmogorov- Smirnov test, and were transformed to their natural logarithm (ln) before Cox regression analyses. For all analyses, P values less than 0.05 were considered statistically significant. All calculations were performed with StatView (SAS Institute, Cary, NC), MedCalc (Med Calc Software, Mariakerke, Belgium), or SPSS (SPSS, Inc., Chicago, IL). TABLE 1. BASELINE CHARACTERISTICS IN RELATION TO GROWTH DIFFERENTIATION FACTOR-15 LEVELS All Patients (n 5 76) GDF-15,,1,200 ng/l (n 5 34) GDF-15, >1,200 ng/l (n 5 42) P Value Age, yr 52 (44 63) 48 (42 54) 57 (47 70) Female, % Body mass index, kg/m ( ) 24.1 ( ) 26.0 ( ) 0.24 NYHA class II, n III, n IV, n Six-minute-walk distance, m 357 ( ) 416 ( ) 320 ( ) Mean RAP, mm Hg 7 (4 10) 4 (2 7) 10 (7 14),0.001 Mean PAP, mm Hg 55 (48 61) 56 (49 60) 55 (44 61) 0.50 Mean PCWP, mm Hg 7 (5 9) 5 (5 8) 8 (6 9) Cardiac output, L/min 3.6 ( ) 3.6 ( ) 3.6 ( ) 0.64 Cardiac index, L/min/m ( ) 2.2 ( ) 2.1 ( ) 0.24 PVR, dyn s cm 25 1,000 (696 1,237) 980 (821 1,209) 1,028 (638 1,373) 0.42 Sv O2, % 63 (59 69) 67 (62 73) 62 (55 67),0.001 Creatinine, mmol/l 77 (69 96) 73 (65 85) 86 (70 102),0.001 Uric acid, mmol/l 419 ( ) 329 ( ) 448 ( ),0.001 NT-proBNP, ng/l 878 (364 2,071) 573 (219 1,035) 1,312 (516 2,413),0.001 Definition of abbreviations: GDF-15 5 growth differentiation factor-15; NT-proBNP 5 N-terminal pro-brain natriuretic peptide; NYHA 5 New York Heart Association; PAP 5 pulmonary arterial pressure; PCWP 5 pulmonary capillary wedge pressure; PVR 5 pulmonary vascular resistance; RAP 5 right atrial pressure. Data are from the first patient cohort. Except for sex (%) and NYHA class (mean 6 SD), data are shown as median (interquartile range). P values were calculated by Spearman rank correlation for continuous variables, x 2 test for nominal variables, and t test for NYHA class. Data on NYHA class were available from n 5 67 individuals, six-minute-walk distance from n 5 38 individuals, mean RAP from n 5 74 individuals, mean PAP from n 5 75 individuals, PCWP from n 5 70 individuals, cardiac output and cardiac index from n 5 73 individuals, PVR from n 5 69 individuals, and Sv O2 from n 5 73 individuals.

3 536 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL Figure 1. Probability of survival free from lung transplantation in the first patient cohort according to baseline levels of growth differentiation factor (GDF)-15 above or below the upper reference limit (1,200 ng/l). RESULTS GDF-15 Levels and Outcome in IPAH The first patient cohort consisted of 76 individuals (68% females) with a median (interquartile range) age of 52 (44 63) years. The clinical and biochemical characteristics of the patients are summarized in Table 1. GDF-15 levels at baseline ranged from 400 to 9,581 ng/l, with a median (interquartile range) value of 1,481 (771 2,731) ng/l. Forty-two patients (55% of the study population) had GDF-15 levels above 1,200 ng/l, the previously defined upper reference limit in apparently healthy elderly individuals (15). Of the 76 patients, 39 (51%) died and 3 (4%) underwent lung transplantation. Patients who reached the composite endpoint of death or lung transplantation had significantly higher median (interquartile range) levels of GDF-15 at baseline (2,047 [972 3,526] ng/l), as compared with patients without an event (948 [665 2,040] ng/l; P ). The risks of the composite endpoint at 1 year were 3% in patients with GDF-15 levels less than 1,200 ng/l and 19% in patients with GDF-15 levels of 1,200 ng/l or greater (P ). The risks of the composite endpoint at 3 years were 15 and 44% in the two groups, respectively (P ). The timing of events is shown in Figure 1. Association of GDF-15 with Clinical, Biochemical, and Hemodynamic Parameters In the first patient cohort, elevated levels of GDF-15 at presentation (>1,200 ng/l) were associated with older age, male sex, more advanced functional impairment, as evidenced by higher New York Heart Association (NYHA) class and lower sixminute-walk distance, and more elevated levels of creatinine, Figure 2. Relation of growth differentiation factor (GDF)-15 to sixminute-walk distance (A), mean right atrial pressure (B), Sv O2 (C), and the levels of N-terminal probrain natriuretic peptide (NTproBNP) (D) in the first patient cohort.

4 Nickel, Kempf, Tapken, et al.: GDF-15 in Idiopathic Pulmonary Arterial Hypertension 537 TABLE 2. RISK OF THE PRIMARY ENDPOINT IN RELATION TO BASELINE RISK MARKER MEASUREMENTS Simple Model Multiple Model HR (95% CI) P Value HR (95% CI) P Value Mean RAP, per 1 mm Hg [ 1.10 (1.05 to 1.16),0.001 Cardiac index, per 0.5 L/min/m 2 Y 1.32 (1.01 to 1.73) PVR, per 100 dyn s cm 25 [ 1.11 (1.04 to 1.17) Sv O2, per 5% Y 1.30 (1.11 to 1.53) Uric acid, per 100 mmol/l [ 1.56 (1.27 to 1.96),0.001 ln NT-proBNP, per 1 SD [ 2.62 (1.78 to 3.86), (1.41 to 4.41) ln GDF-15, per 1 SD [ 2.51 (1.76 to 3.59), (1.43 to 4.76) Definition of abbreviations: CI5 confidence interval; GDF-15 5 growth differentiation factor-15; HR 5 hazard ratio; ln 5 natural logarithm; NT-proBNP 5 N-terminal pro-brain natriuretic peptide; PVR 5 pulmonary vascular resistance; RAP 5 right atrial pressure; [ 5 increase; Y 5 decrease. Data are from the first patient cohort. Estimated HRs, 95% CIs, and P values were calculated by simple and stepwise forward Cox regression analyses. NT-proBNP and GDF-15 were not normally distributed and therefore ln transformed; HRs refer to 1 SD in the ln scale in these variables. uric acid, and NT-proBNP (Table 1). Moreover, patients with elevated levels of GDF-15 had higher mean right atrial and pulmonary capillary wedge pressures and a lower mixed venous oxygen saturation (Sv O2 ). No significant differences were observed with regard to mean pulmonary artery pressure, cardiac output or index, or pulmonary vascular resistance (Table 1). The relations between GDF-15 and six-minute-walk distance (rho , P ), mean right atrial pressure (rho , P, 0.001), Sv O2 (rho , P, 0.001), and NT-proBNP (rho , P, 0.001) are illustrated in Figure 2. GDF-15 in the Context of Other Markers of Adverse Outcome Several biochemical and hemodynamic parameters are indicative of a poor prognosis in patients with IPAH, and simple Cox regression analysis confirmed the prognostic utility of these established markers in the first patient cohort (Table 2): elevated mean right atrial pressure, increased pulmonary vascular resistance, reduced cardiac index or Sv O2, and elevated levels of uric acid, NT-proBNP, and GDF-15 were all associated with an increased risk of death or transplantation. By forward stepwise Cox regression analysis, only NT-proBNP and GDF-15 remained as significant predictors of adverse outcomes (Table 2). Sensitivity analyses confirmed GDF-15 as a significant predictor of death or transplantation when directly compared with each one of the other parameters associated with a poor outcome by simple Cox regression analysis (data not shown). ROC curve analyses further illustrated that GDF-15 is a strong indicator of adverse outcomes in IPAH. The best GDF- 15 cutoff level for predicting outcome at 3 years was 2,097 ng/l (sensitivity, 74%; specificity, 82%). The c-statistic for GDF-15 was 0.80 (95% confidence interval [CI], 0.69 to 0.88), which was similar to the c-statistics of NT-proBNP (0.81; 95% CI, 0.70 to 0.89) or uric acid (0.75; 95% CI, 0.64 to 0.84), but numerically greater than the c-statistics of mean right atrial pressure (0.68; 95% CI, 0.56 to 0.78), Sv O2 (0.68; 95% CI, 0.56 to 0.79), cardiac index (0.64; 95% CI, 0.51 to 0.75), and pulmonary vascular resistance (0.58; 95% CI, 0.45 to 0.70) (Figure 3). Prognostic Utility of GDF-15 across the Clinical Spectrum of IPAH The prognostic value of GDF-15 was assessed in a number of subgroups in the first patient cohort (Figure 4). GDF-15 levels above the ROC curve derived cutoff (2,097 ng/l) were associated with an increased risk of death or transplantation in patient subgroups defined according to age, sex, and body mass index. Furthermore, GDF-15 added significant prognostic information to key hemodynamic parameters, uric acid, and NT-proBNP (Figure 4). GDF-15 also appeared to add prognostic information to six-minute-walk distance, although the number of patients in this subgroup was rather low, and statistical significance was not achieved (Figure 4). The additive prognostic value of GDF-15 and NT-proBNP is further illustrated in Figure 5. Measurement of GDF-15 appeared to be especially useful among patients with elevated levels of NT-proBNP; in that subgroup, high levels of GDF-15 identified a patient population with a dismal prognosis (Figure 5). Longitudinal Changes after the Initiation of Medical Therapy The second cohort included 22 patients (68% females) with a median (interquartile range) age of 57 (50 78) years. The median (interquartile range) level of GDF-15 was 1,298 (860 2,438) ng/l. Baseline characteristics are shown in Table 3. At 3 to 6 months after the initiation of medical therapy, significant decreases of mean pulmonary artery pressure and pulmonary vascular resistance, and significant increases of cardiac output and index, were observed (Table 3). GDF-15 and NT-proBNP levels did not change significantly from baseline to follow-up in Figure 3. Receiver operating characteristic curve analyses relating biomarker levels and hemodynamic parameters to 3-year outcome in the first patient cohort (see text for details). GDF-15 5 growth differentiation factor-15; NT-proBNP 5 N-terminal pro-brain natriuretic peptide; PVR 5 pulmonary vascular resistance; RAP 5 right atrial pressure.

5 538 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL Figure 4. Risk of death or transplantation during follow-up associated with growth differentiation factor (GDF)-15 levels above the receiver operating characteristic curve derived cutoff (2,097 ng/l) in subgroups of the first patient cohort defined according to clinical variables, six-minute-walk distance (6 min WD), hemodynamic parameters, uric acid, and N-terminal pro-brain natriuretic peptide (NT-proBNP). Continuous variables were dichotomized based on their median values to create subgroups of comparable size. BMI 5 body mass index; CI 5 confidence interval; HR 5 hazard ratio; RAP 5 right atrial pressure; PAP 5 pulmonary artery pressure; PCWP 5 pulmonary capillary wedge pressure; PVR 5 pulmonary vascular resistance. the overall patient cohort (Table 3). However, changes in GDF- 15 over time were significantly related to changes in NTproBNP (rho , P ) and inversely related to changes in Sv O2 (rho , P, 0.001) (Table 3 and Figure 6). DISCUSSION A New Biomarker in IPAH The present study identifies GDF-15 as a powerful biomarker of adverse outcome in patients with IPAH. About half of the patients presented with a GDF-15 level above the upper limit of normal (1,200 ng/l), and these individuals had a much higher risk of death or transplantation during follow-up. Differences in outcome were observed early on, and were statistically significant already after 1 year. GDF-15 Levels Integrate Several Markers of Disease Severity GDF-15 levels were significantly related to age, male sex, NYHA class, lower six-minute-walk distances, the levels of creatinine, uric acid, and NT-proBNP, mean right atrial and pulmonary capillary wedge pressures, and a lower Sv O2, indicating that elevated levels of GDF-15 integrate several clinical, functional, hemodynamic, and biochemical indicators of more severe disease and poor prognosis in patients with IPAH. However, it is interesting to note that there was a rather loose association between functional capacity and the levels of GDF-15 (significant overlap in NYHA class and six-minute-walk distances in patients with or without elevated levels of GDF-15 in Table 1), indicating that GDF-15 provides information that cannot be obtained by simple clinical assessment. Pathobiology of GDF-15 in IPAH GDF-15 levels did not correlate with mean pulmonary artery pressure, cardiac output and index, or pulmonary vascular resistance, indicating that GDF-15 is not directly related to, and cannot be used as a diagnostic marker of, pulmonary hypertension. Notably, baseline NT-proBNP levels were significantly

6 Nickel, Kempf, Tapken, et al.: GDF-15 in Idiopathic Pulmonary Arterial Hypertension 539 Figure 5. Probability of survival free from lung transplantation in the first patient cohort according to baseline levels of growth differentiation factor (GDF)-15 and N-terminal pro-brain natriuretic peptide (NTproBNP). Patients were stratified according to the median levels of GDF-15 and NT-proBNP to create subgroups of comparable size. related to mean pulmonary artery pressure, low cardiac output and index, and elevated pulmonary vascular resistance (data not shown), indicating that NT-proBNP and GDF-15 capture distinct aspects of IPAH pathophysiology. Nevertheless, GDF-15 levels were significantly related to NT-proBNP levels at baseline (rho ), suggesting that the levels of GDF-15 may reflect, to some extent, cardiac pathologies. Our observation that changes in GDF-15 levels were significantly related to changes in NT-proBNP levels after initiation of medical therapy supports this conclusion. Similar relations between baseline levels of GDF-15 and NT-proBNP have been observed in patients with chronic left-sided heart failure (8) and acute pulmonary embolism (9). Mechanical stretch triggers GDF-15 expression in isolated cardiomyocytes (16), and acute pressure overload leads to a sharp increase in left ventricular GDF-15 expression levels in mice (5). It is interesting in this regard that patients with pulmonary embolism and acute right ventricular overload have even higher levels of GDF-15 as compared with patients with IPAH (9). Although these considerations point to the pressure-overloaded right ventricle as a significant source of GDF-15 in IPAH, subtle left ventricular abnormalities may contribute as well, as suggested by the significant relation of GDF-15 to mean pulmonary capillary wedge pressure in the present study. Notably, the levels of GDF-15 in IPAH were closely associated with elevated mean right atrial pressures and reduced Sv O2, raising the possibility that GDF-15 is released from peripheral tissues exposed to venous congestion and/or tissue hypoxia. This hypothesis was further supported by the finding that changes in GDF-15 levels were closely and inversely related to changes in Sv O2 after the initiation of medical therapy. The significant relation of GDF-15 to uric acid, which was also observed in patients with left-sided heart failure (8), points in a similar direction. Hyperuricemia has been proposed to reflect tissue hypoxia, impaired oxidative metabolism, and inflammatory cytokine activation in patients with heart failure (17). Previous experimental studies have identified hypoxia and increased oxidative stress as potent inducers of GDF-15 expression in multiple cell types (3, 4, 18 20), thus emphasizing that GDF-15 is not a cardiac-specific cytokine. This apparent lack of cardiac specificity may explain why GDF-15 carries prognostic information beyond that provided by NT-proBNP, which is released primarily from the right ventricle in IPAH. Additive Prognostic Value of GDF-15 and Its Possible Use in Multimarker Strategies Modern therapy for IPAH comprises goal-oriented strategies in which treatment is intensified when predefined goals are not met (21). The ideal treatment goals, however, have not been defined, but it is expected that biomarkers are going to play an increasing role because they can be measured repeatedly and noninvasively. Given the limitations of BNP or NT-proBNP when it comes to predicting individual outcomes (12, 13), the combination of BNP or NT-proBNP with other markers may be expected to facilitate therapeutic decision making. Multimarker strategies combining biomarkers from distinct pathophysiologic axes are increasingly used for risk assessment and for individ- TABLE 3. CHANGES IN FUNCTIONAL VARIABLES AND BIOMARKER LEVELS FROM BASELINE TO 3 6 MONTHS FOLLOW-UP Variables Baseline (n 5 22) Follow-up (n 5 22) P Value (baseline vs. follow-up) DGDF-15/DVariable (baseline vs. follow-up) Rho P Value Mean RAP, mm Hg 7 (4 9) 5 (2 10) Mean PAP, mm Hg 52 (42 61) 46 (38 61) Mean PCWP, mm Hg 6 (5 9) 6 (5 8) Cardiac output, L/min 3.7 ( ) 4.4 ( ) Cardiac index, L/min/m ( ) 2.6 ( ) PVR, dyn s cm (729 1,080) 717 ( ) Sv O2, % 67 (60 70) 67 (60 72) ,0.001 Six-minute-walk distance, m 365 ( ) 420 ( ) NT-proBNP, ng/l 521 (136 1,482) 489 (190 1,798) GDF-15, ng/l 1,298 (860 2,438) 1,387 (946 2,419) 0.32 NC NC Definition of abbreviations: GDF-15 5 growth differentiation factor-15; NC 5 not calculated; NT-proBNP 5 N-terminal pro-brain natriuretic peptide; PAP 5 pulmonary arterial pressure; PCWP 5 pulmonary capillary wedge pressure; PVR 5 pulmonary vascular resistance; RAP 5 right atrial pressure. Data are from the second patient cohort. Data are shown as median (interquartile range). Changes from baseline to 3 6 months follow-up were assessed by Wilcoxon signed rank test. Changes (D) in GDF-15 in relation to changes (D) in baseline variables over time were assessed by Spearman rank correlation coefficients. Data on six-minute-walk distance were available from n 5 17 individuals.

7 540 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL Figure 6. Changes (D) in growth differentiation factor (GDF)-15 levels from baseline to 3 to 6 months in relation to changes in Sv O2 (A) and changes in N-terminal pro-brain natriuretic peptide (NT-proBNP) (B) in the second patient cohort. ualizing therapy in patients with acute coronary syndromes (22). It has been shown in this clinical scenario that GDF-15, in combination with markers of myocardial necrosis and the electrocardiogram, may help to identify patients who will derive the greatest benefit from early revascularization (23). In the present study, GDF-15 added substantial prognostic information to established biochemical and hemodynamic parameters currently used for risk assessment in IPAH. The observation that a combined assessment of GDF-15 and NT-proBNP at baseline allows for an identification of a patient subgroup with an extremely poor prognosis may be of particular interest. Due to the limited number of patients with available information on six-minute-walk distance in our study, more work is needed to clarify whether GDF-15 can also add prognostic information to functional capacity. The rather loose association between six-minute-walk distance and GDF-15 (rho ) indicates that GDF-15 carries information beyond functional capacity. Notably, median GDF-15 levels did not change significantly in the overall patient population 3 to 6 months after the initiation of medical therapy. However, the significant relation between changes in GDF-15 and changes in NT-proBNP and the significant inverse relation between changes in GDF-15 and changes in mixed Sv O2 suggest that repeat measurements of GDF-15 may help to identify patients with a favorable response to contemporary treatment regimes. Conclusions GDF-15 emerges as a new and promising biomarker of the risk of death or transplantation in patients with IPAH. Measurement of GDF-15 in treatment-naive patients adds substantial prognostic information to that provided by right heart catheterization and more established biomarkers, including NTproBNP. On the basis of these findings, we believe that the prognostic value of GDF-15 in this setting deserves further investigation. Appropriate cutoff levels of GDF-15 and the use of GDF-15 in multimarker strategies need to be validated prospectively. From a therapeutic perspective, it will be important to explore whether GDF-15 can be used to guide therapeutic decisions and monitor response to therapy. Conflict of Interest Statement: N.N. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. T.K. (with K.C.W.) has filed a patent and has a contract with Roche Diagnostics to develop an assay for GDF-15 used for diagnosis and prognosis in cardiovascular disease. H.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. F.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. U.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. H.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.O. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. M.R.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.S.R.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. M.M.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. K.C.W. (with T.K.) has filed a patent and has a contract with Roche Diagnostics to develop an assay for GDF-15 used for diagnosis and prognosis in cardiovascular disease. References 1. Bootcov MR, Bauskin AR, Valenzuela SM, Moore AG, Bansal M, He XY, Zhang HP, Donnellan M, Mahler S, Pryor K, et al. MIC-1, a novel macrophage inhibitory cytokine, is a divergent member of the TGF-b superfamily. Proc Natl Acad Sci USA 1997;94: Su AI, Wiltshire T, Batalov S, Lapp H, Ching KA, Block D, Zhang J, Soden R, Hayakawa M, Kreiman G, et al. A gene atlas of the mouse

8 Nickel, Kempf, Tapken, et al.: GDF-15 in Idiopathic Pulmonary Arterial Hypertension 541 and human protein-encoding transcriptomes. Proc Natl Acad Sci USA 2004;101: Schlittenhardt D, Schober A, Strelau J, Bonaterra GA, Schmiedt W, Unsicker K, Metz J, Kinscherf R. Involvement of growth differentiation factor-15/macrophage inhibitory cytokine-1 (GDF-15/MIC-1) in oxldl-induced apoptosis of human macrophages in vitro and in arteriosclerotic lesions. Cell Tissue Res 2004;318: Kempf T, Eden M, Strelau J, Naguib M, Willenbockel C, Tongers J, Heineke J, Kotlarz D, Xu J, Molkentin JD, et al. The transforming growth factor-b superfamily member growth-differentiation factor-15 protects the heart from ischemia/reperfusion injury. Circ Res 2006;98: 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 and antihypertrophic factor released from the myocardium in association with SMAD protein activation. Circ Res 2006;98: Wollert KC, Kempf T, Peter T, Olofsson S, James S, Johnston N, Lindahl B, Horn-Wichmann R, Brabant G, Simoons ML, et al. Prognostic value of growth-differentiation factor-15 in patients with non-st-segment elevation acute coronary syndrome. Circulation 2007;115: Kempf T, Bjorklund E, Olofsson S, Lindahl B, Allhoff T, Peter T, Tongers J, Wollert KC, Wallentin L. Growth-differentiation factor-15 improves risk stratification in ST-segment elevation myocardial infarction. Eur Heart J 2007;28: Kempf T, von Haehling S, Peter T, Allhoff T, Cicoira M, Doehner W, Ponikowski P, Filippatos GS, Rozentryt P, Drexler H, et al. Prognostic utility of growth differentiation factor-15 in patients with chronic heart failure. J Am Coll Cardiol 2007;50: Lankeit M, Kempf T, Dellas C, Cuny M, Tapken H, Peter T, Olschewski M, Konstantinides S, Wollert KC. Growth differentiation factor-15 for prognostic assessment of patients with acute pulmonary embolism. Am J Respir Crit Care Med 2008;177: Humbert M, Sitbon O, Simonneau G. Treatment of pulmonary arterial hypertension. N Engl J Med 2004;351: Hoeper MM, Hohlfeld JM, Fabel H. Hyperuricaemia in patients with right or left heart failure. Eur Respir J 1999;13: Nagaya N, Nishikimi T, Uematsu M, Satoh T, Kyotani S, Sakamaki F, Kakishita M, Fukushima K, Okano Y, Nakanishi N, et al. Plasma brain natriuretic peptide as a prognostic indicator in patients with primary pulmonary hypertension. Circulation 2000;102: Fijalkowska A, Kurzyna M, Torbicki A, Szewczyk G, Florczyk M, Pruszczyk P, Szturmowicz M. Serum N-terminal brain natriuretic peptide as a prognostic parameter in patients with pulmonary hypertension. Chest 2006;129: Nickel N, Kempf T, Golpon H, Wollert KC, Wilkins M, Gibbs SJ, Hoeper MM. Growth-differentiation factor-15 (GDF-15) as a prognostic marker in idiopathic pulmonary arterial hypertension (IPAH) [abstract]. Proc Am Thorac Soc 2008;5:A 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: Frank D, Kuhn C, Brors B, Hanselmann C, Ludde M, Katus HA, Frey N. Gene expression pattern in biomechanically stretched cardiomyocytes: evidence for a stretch-specific gene program. Hypertension 2008;51: Anker SD, Doehner W, Rauchhaus M, Sharma R, Francis D, Knosalla C, Davos CH, Cicoira M, Shamim W, Kemp M, et al. Uric acid and survival in chronic heart failure: validation and application in metabolic, functional, and hemodynamic staging. Circulation 2003;107: Hsiao EC, Koniaris LG, Zimmers-Koniaris T, Sebald SM, Huynh TV, Lee SJ. Characterization of growth-differentiation factor 15, a transforming growth factor b superfamily member induced following liver injury. Mol Cell Biol 2000;20: Zimmers TA, Jin X, Hsiao EC, McGrath SA, Esquela AF, Koniaris LG. Growth differentiation factor-15/macrophage inhibitory cytokine-1 induction after kidney and lung injury. Shock 2005;23: Zimmers TA, Jin X, Hsiao EC, Perez EA, Pierce RH, Chavin KD, Koniaris LG. Growth differentiation factor-15: induction in liver injury through p53 and tumor necrosis factor-independent mechanisms. J Surg Res 2006;130: Hoeper MM, Markevych I, Spiekerkoetter E, Welte T, Niedermeyer J. Goal-oriented treatment and combination therapy for pulmonary arterial hypertension. Eur Respir J 2005;26: Anderson JL, Adams CD, Antman EM, Bridges CR, Califf RM, Casey DE Jr, Chavey WE II, Fesmire FM, Hochman JS, Levin TN, et al. ACC/AHA 2007 guidelines for the management of patients with unstable angina/non ST-elevation myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 2002 Guidelines for the Management of Patients with Unstable Angina/Non ST-Elevation Myocardial Infarction): developed in collaboration with the American College of Emergency Physicians, the Society for Cardiovascular Angiography and Interventions, and the Society of Thoracic Surgeons: endorsed by the American Association of Cardiovascular and Pulmonary Rehabilitation and the Society for Academic Emergency Medicine. Circulation 2007;116:e148 e Wollert KC, Kempf T, Lagerqvist B, Lindahl B, Olofsson S, Allhoff T, Peter T, Siegbahn A, Venge P, Drexler H, et al. Growth-differentiation factor 15 for risk stratification and selection of an invasive treatment strategy in non-st-elevation acute coronary syndrome. Circulation 2007;116: