Figures
Abstract
Objective
Growth differentiation factor-15 (GDF-15) has been identified as a strong marker of cardiovascular disease; however, no data are available concerning the role of GDF-15 in the occurrence of organ dysfunction during coronary artery bypass grafting (CABG) associated with cardiopulmonary bypass (CPB).
Methods
Five arterial blood samples were taken sequentially in 34 patients from anesthesia induction (IND) until 24 h after arrival at the intensive care unit (ICU). Plasma levels of GDF-15, follistatin-like 1 (FLST1), myeloperoxidases (MPO), hydroperoxides and plasma antioxidant status (PAS) were measured at each time-point. Markers of cardiac (cardiac-troponin I, cTnI) and renal dysfunction (neutrophil gelatinase-associated lipocalin, NGAL) and other classical biological factors and clinical data were measured.
Results
Plasma GDF-15 levels increased gradually during and after surgery, reaching nearly three times the IND levels in the ICU (3,075±284 ng/L vs. 1,061±90 ng/L, p<0.001). Plasma MPO levels increased dramatically during surgery, attaining their highest level after unclamping (UNCLAMP) (49±11 ng/mL vs. 1,679±153 ng/mL, p<0.001) while PAS significantly decreased between IND and UNCLAMP (p<0.05), confirming the high oxidative status induced by this surgical procedure. ICU levels of GDF-15 correlated positively with cTnI and NGAL (p = 0.006 and p = 0.036, respectively), and also with hemoglobin and estimated glomerular filtration rate (eGFR). Among all the post-operative biomarkers available, only eGFR, NGAL and GDF-15 measured at ICU arrival were significantly associated with the onset of acute kidney injury (AKI). Patients with a EuroSCORE >3 were shown to have higher GDF-15 levels.
Citation: Kahli A, Guenancia C, Zeller M, Grosjean S, Stamboul K, Rochette L, et al. (2014) Growth Differentiation Factor-15 (GDF-15) Levels Are Associated with Cardiac and Renal Injury in Patients Undergoing Coronary Artery Bypass Grafting with Cardiopulmonary Bypass. PLoS ONE 9(8): e105759. https://doi.org/10.1371/journal.pone.0105759
Editor: Jaap A. Joles, University Medical Center Utrecht, Netherlands
Received: April 17, 2014; Accepted: July 23, 2014; Published: August 29, 2014
Copyright: © 2014 Kahli et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by grants from the French Ministry of Research, from the Institut National de la Santé et de la Recherche Médicale (INSERM)and from the Regional Council of Burgundy. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
During cardiopulmonary bypass (CPB), pro-inflammatory mediators activate leucocytes, vascular endothelial cells and platelets, thus producing a systemic inflammatory response that results in organ dysfunction, affecting the heart, brain, lungs and kidneys. Several factors, including operative trauma, the contact of blood components with the artificial surface of the circuit, cardioplegic techniques and allogenic blood transfusion account for this phenomenon. The cellular response is mediated by the main effectors of the inflammatory response, including neutrophils, which degranulate and release cytotoxic molecules such as elastase, myeloperoxidase (MPO) and free radicals[1]. Several therapeutic strategies have thus been developed to minimize this inflammation and lessen organ dysfunction[2]. In a previous study carried out in our laboratory, we measured systemic and coronary sinus free-radical release during and following removal of the cross-clamp[3]. We observed that radical production was significantly greater during and after the CPB procedure, and that circulating free-radical levels correlated with serum creatine kinase-MB (CK-MB). Various biological and hemodynamic markers are measured to estimate the pre-and postoperative risk of developing complications; however, no data exist to assess the role and impact of growth differentiation factor-15 (GDF-15) in the occurrence of organ dysfunction during coronary artery bypass grafting (CABG) associated with CPB.
GDF-15 (also known as NAG-1 and MIC-1) is a cytokine related to the superfamily of transforming growth factor-β (TGF-β), and is weakly expressed or not expressed at all under physiological conditions[4]. The plasma concentration of GDF-15 increases under pathological conditions such as hypoxia, inflammation or oxidative stress and is closely associated with all-cause mortality[5]. The expression of GDF-15 has been induced in cultured cardiomyocytes during experimental ischemia/reperfusion (IR), and in cardiomyocytes subjected to nitrosative stress and stimulation with proinflammatory cytokines and interferon-γ (IFN-γ)[6]. In the cardiovascular system, GDF-15 is now identified as a strong prognosis marker in patients with cardiovascular disease, including coronary artery disease (CAD), acute coronary syndromes (ACS)[7], [8] and heart failure (HF)[9]. GDF-15 is associated with reduced endothelium-dependent vasodilation, the risk of atherosclerotic plaque rupture and reduced left ventricular ejection fraction (LVEF)[10]. Cardiomyocytes have been identified as the main source of GDF-15 in patients with ST-elevation myocardial infarction (STEMI) or HF[6]. Recent findings also suggested that the secreted protein FSTL-1 could be an upstream inducer of GDF15 production and an independent prognostic biomarker in acute coronary syndromes[11]. Finally, it was suggested that GDF-15 could represent a novel risk marker in association with the EuroSCORE for risk stratification in cardiac surgery patients[12]. In addition, GDF-15 is also a novel independent serum marker of mortality in chronic kidney disease (CKD), and combining this marker with other established predictors of mortality may help to identify individuals at high risk for developing CKD [13], [14].
The present study was designed to evaluate the kinetics of plasma GDF-15 levels in patients with significant CAD undergoing cardiac surgery associated with CPB, and to relate GDF-15 to inflammatory/ oxidative stress status and to organ dysfunction.
Methods
Study patients
Thirty-four consecutive patients operated on for CABG under CPB at Dijon University Hospital, Bocage Central, Dijon, France were included in this prospective study between April 26th 2012 and October 11th 2012. The study protocol complied with the Declaration of Helsinki and was approved by the regional ethics committee (Espace éthique de Bourgogne/Franche-Comté, CHU de Besançon, Hôpital St Jacques, Besançon, France; ARTICLE Protocol, 2012-A00184-39). Informed written consent was obtained from all of the patients. Patients with stable CAD were included. The following criteria led to the exclusion of patients: surgical emergencies, aortic valve replacement, ACS reported within 30 days before the surgery, LVEF <30%, chronic inflammatory pathologies, infectious or malignant diseases, chronic renal failure, transplant patients and patients treated with corticosteroids.
Clinical postoperative events
Clinical events were defined as hospital stay>10 days, intensive care unit (ICU) stay>5 days, the use of vasopressor therapy and clinical complications (blood transfusion, infection, acute kidney injury and atrial fibrillation).
The association between these events and biomarkers was also analyzed.
Anesthesia and heart surgery procedure
Patients were pre-medicated with midazolam orally plus hydroxyzine 90 min before anesthesia. Routine cardiac medications were continued until the morning of the surgery, except for Clopidogrel, which was stopped at least 5 days earlier. Before the induction of anesthesia, a complete hemodynamic monitoring system was set up in the operating room. Anesthesia was induced with intravenous midazolam (0.02 mg.kg−1), sufentanil (0.2 to 0.5 µg.kg−1.h−1), and propofol (1.5 to 2.5 mg.kg−1). After verifying that manual ventilation was satisfactory, cisatracurium dibesylate (0.06 mg.kg−1.h−1) was injected. Patients were orally intubated and ventilated with FiO2: 0.4. Anesthesia was maintained with sufentanil and cisatracurium as required and inhaled desflurane.
Surgery was performed according to the following protocol: sternotomy, harvesting of the right and left internal mammary arteries and saphenous vein, and aortic and cavoatrial cannulation, followed by the implementation of CPB and myocardial protection by anterograde cardioplegia. Moderate body hypothermia (32°C) was used. The time from removal of the aortic cross-clamp until the discontinuation of CPB was measured. After completion of the CABG (from 1 to 5 CABG in the patients studied), CPB was discontinued and protamine was given for heparin reversal. Patients were given a blood transfusion in the case of severe anemia (blood hematocrit, 22% during CPB or, 26% after CPB). After closure of the sternum, the patients were transferred to the post-operative ICU, and finally to the surgery ward. Clinical and biological parameters were regularly recorded. Complications such as arrhythmia, blood loss, infection, and organ failure were recorded. ICU length of stay was noted.
Study protocol
Blood samples were taken in heparinized tubes as shown in Figure 1. Arterial samples were taken from the arterial catheter or from the CPB pump: after anesthesia (IND), just before starting CPB (pre-CPB), just after removing the cross clamp (UNCLAMP), after surgery (Post-SURG) upon arrival and after 24 hours at the ICU. Blood samples were immediately centrifuged after collection and the plasma was immediately frozen in liquid nitrogen and stored at −80°C until analysis.
Arterial blood samples were harvested after anesthesia (IND), just before starting CPB (pre-CPB), just after removing the cross clamp (UNCLAMP), after surgery upon arrival in the ICU (Post-SURG) and after 24 hours at the cardiovascular intensive care unit (ICU). Markers of oxidative stress, inflammation, organ failure and clinical criteria were collected.
Plasma levels of GDF-15 and follistatin like-1 during cardiac surgery with CPB
Plasma GDF-15 concentrations were measured by quantitative sandwich enzyme immunoassay (Human GDF-15, Quantikine, R&D Systems Europe, Lille, France) with a linear range from 200 to 50,000 ng/L. The color intensity, relative to GDF-15 concentration, was measured at 450 nm with a spectrophotometer (VictorV3, Perkin Elmer, Courtaboeuf, France). A GDF-15 value of 1,200 ng/L corresponds to the upper limit of normal in healthy elderly individuals, whereas GDF-15 levels of 1,200 and 1,800 ng/L allowed the identification of patients at low (<1,200 ng/L), intermediate (between 1,200 and 1,800 ng/L), or high risk (>1,800 ng/L)[7], [15]. Plasma follistatin like-1 protein (FSTL-1) concentrations were measured by a quantitative sandwich enzyme immunoassay technique (ELISA kit for FSTL-1, Uscn, Life Science Inc). In order to account for hemodilution, plasma GDF-15 and FSTL-1 levels were corrected using proteinemia.
Oxidative stress/inflammation analysis
The plasma antioxidant status (PAS) was determined by measuring the Oxygen Radical Absorbance Capacity (ORAC), as detailed elsewhere[3], [16]. The results were expressed in ORAC units, where 1 ORAC unit equals the net protection provided by 1 µM Trolox. Plasma antioxidant status is a dynamic measurement and does not need to be corrected for hemodilution. Plasma MPO concentrations were measured by a quantitative sandwich enzyme immunoassay technique (Human Myeloperoxidase Immunoassay, Quantikine ELISA, R&D Systems Europe, Lille, France). Plasma concentrations of hydroperoxides were determined by a colorimetric assay using Free Oxygen Radical Testing (FORT, FORM-PLUS-3000, Optimabio, Ollioules, France), as previously described[17]. The hydroperoxide concentrations were corrected for hemodilution.
Plasma protein concentrations (expressed as grams per liter of plasma) were evaluated by the spectrophotometric method according to Lowry et al.
Markers of organ dysfunction
Neutrophil gelatinase-associated lipocalin (NGAL) is considered an early marker of acute kidney injury (AKI)[18], [19]. Plasma NGAL concentrations were measured at IND, post-SURG and ICU by a quantitative sandwich enzyme immunoassay technique (Human NGAL Immunoassay, Quantikine ELISA, R&D Systems Europe, Lille, France). Standard blood count, platelet count, renal enzymes, blood gases, pH, lactic acid, cardiac troponin-I (cTnI) were measured by the hospital’s medical analysis laboratory the day before (DAY-1), before (PRE-CPB) and after CPB (post-SURG) and at the first postoperative day (ICU). The following patients’ data were collected: age, sex, and cardiovascular risk factors (hypertension, atrial fibrillation, body mass index (BMI), dyslipidemia, diabetes mellitus and renal failure), current smoker and chronic medications before admission. Hemodynamic parameters on admission (heart rate, systolic blood pressure) were recorded. LVEF was measured by echocardiography at admission and during the postoperative hospital stay. The estimated glomerular filtration rate (eGFR) was calculated from pre-operative plasma creatinine by the abbreviated Modifications of Diet in Renal Disease equation[20]. AKI was defined in accordance with the AKI network (AKIN) criteria as an abrupt decrease (in 48 h) in renal function, defined by an increase in absolute serum creatinine (SCr) of at least 26.5 µmol/L (0.3 mg/dL) or by a percentage increase in SCr ≥50% (1.5× baseline value), or by a decrease in the urinary output (documented oliguria <0.5 mL/kg/h for more than 6 h)[21]. The additive EuroSCORE was calculated, and a cut-off at 3 was used for more clinical relevance.
Statistical analyses
Continuous variables are presented as mean ± SEM or median (IQR) as appropriate, and discrete variables as absolute number and percentages.
Given time points were analyzed by Student’s un-paired t-test or the Mann-Whitney test for normally or not normally distributed data, respectively. Results are expressed as box plots in which the boxes represent the 25th and 75th percentiles, or as means+SEM.
One Way Repeated Measures Analysis of Variance or Friedman Repeated Measures Analysis of Variance on Ranks was used to analyze differences within groups over time, and corrected for multiple comparisons by the Tukey test. A probability value of <0.05 was considered statistically significant.
Associations between laboratory parameters and GDF-15 were tested using Spearman’s correlation rank test.
To determine the incremental diagnostic value of post-operative GDF-15 in the early detection of AKI compared with conventional biomarkers, a stepwise analysis was performed given the limited number of events. We calculated the improvement in global χ2 after the addition of GDF-15 to a model that included post-operative eGFR and eGFR plus NGAL.
A probability value <0.05 was considered statistically significant.
Statistical analyses were performed using SigmaPlot version 12 (Systat Software, Inc.) and SPSS 20.0 0 (SPSS, Inc., Chicago, IL, USA).
Results
Perioperative clinical and laboratory information is shown in Table 1 (left part). The mean age of patients was 65±2 years. Preoperative LVEF was 58±2% and the mean EuroSCORE was 2.9. Initial plasma GDF-15 levels (IND) ranged from 456 to 2,832 ng/L with a mean value of 1,061±91 ng/L.
Intraoperative data and postoperative outcomes are shown in Table 2. The majority of patients (n = 21) underwent three bypass grafts. Nine patients with LVEF<40% received an infusion of Glucose-Insulin–Potassium (GIK) (potassium 40 mEq and regular insulin 300 IU in 500 ml of 30% glucose) one hour before aortic clamping. All of the patients were still alive on 31st October 2013.
Plasma levels of GDF-15 and FSTL-1 during cardiac surgery with CPB
Compared with IND levels, plasma GDF-15 concentrations rose significantly (p<0.05) after removal of the cross-clamp and during the post-surgery period in the ICU, reaching nearly three times the induction level (Figure 2A). Plasma FSTL-1 levels were not modified during the surgical procedure. However, a slight rise in FSTL-1 was observed in the ICU (Figure 2B).
Values, corrected for hemodilution, are mean ± SEM, *: significantly different (p<0.05) between time-point and IND levels (One way repeated measures ANOVA, followed by pair-wise comparison of mean levels with Tukey test).
Clinical and biological variables associated with IND plasma GDF-15 concentrations are shown in Table 1 (right part). GDF-15 levels were significantly associated with age (r = 0.477, p = 0.005). A medical history of diabetes mellitus was significantly associated with higher GDF-15 levels (p = 0.011). Moreover, associations between GDF-15 concentrations and blood glucose levels were significant (r = 0.464, p = 0.006). Concerning the cardiac status, plasma levels of GDF-15 were significantly associated with those of N-terminal pro-brain natriuretic peptide (NT-proBNP) (r = 0.347, p = 0.044) but not with plasma levels of creatine kinase.
Although lower GFR values and higher plasma creatinine levels were associated with higher GDF-15 levels (respectively, r = −0.465, p = 0.006 and r = 0.398, p = 0.019), neither urea nor NGAL concentrations were significantly associated with GDF-15 levels. Regarding the inflammatory and oxidative stress status, the highest GDF-15 levels were associated with CRP (p = 0.041), but not with higher hydroperoxide and MPO concentrations or PAS. Higher preoperative GDF-15 levels were significantly associated with higher EuroSCORE values (r = 0.585, p = 0.0003).
The association of clinical and biological variables with peak plasma GDF-15 concentrations, observed one day after surgery in ICU, was also investigated. Patients with moderate or high operative risk (EuroSCORE > 3) had higher plasma GDF-15 concentrations than did patients with low operative risk (EuroSCORE ≤3) (p = 0.007) (Figure 3A). GDF-15 levels at the ICU were significantly higher in patients who presented anemia during the postoperative period (p = 0.010) (Figure 3B); moreover, higher GDF-15 concentrations correlated with lower ICU plasma hemoglobin (r = −0.456, p = 0.007) (Figure 3C). Patients who developed postoperative renal failure presented significantly higher GDF-15 levels at the ICU (p = 0.0004) (Figure 3D). Furthermore, ICU levels of GDF-15 correlated significantly with a worse eGFR (r = −0.640, p<0.0001) (Figure 3E).
(Student’s test, Spearman’s correlation).
Evolution of oxidative stress/inflammation
The PAS decreased significantly (p<0.05) during the CPB procedure (Figure 4A) while plasma MPO levels increased dramatically (p<0.05) during the surgical procedure, attaining their highest levels just after unclamping and then decreasing rapidly (Figure 4B), confirming the high inflammatory status induced by this surgical procedure. Whereas plasma hydroperoxide levels did not fluctuate significantly during and after the surgery (Figure 4C), ICU hydroperoxide plasma levels were associated with longer ICU and hospital stays (p = 0.027 and p = 0.044 respectively).
GDF-15 and Markers of tissue injury
Cardiac Troponin I (cTnI) was significantly higher (p = 0.008) on arrival at the ICU (Figure 5A) and remained high the following day. Concerning the early stages of kidney damage, plasma NGAL levels increased significantly (p<0.001) after surgery (Figure 5C) and remained high in the ICU. Moreover, the day after surgery, plasma GDF-15 levels correlated positively and significantly with cTnI and NGAL ICU levels (Figure 5B, D). Among all of the post-operative biomarkers available, only eGFR, NGAL and GDF-15 measured at ICU arrival were significantly associated with the onset of AKI. Given the small size of the population, we performed stepwise multivariate regression beginning with eGFR, and we tested the incremental performance of the model when the two other biomarkers were added (Figure 6). The addition of NGAL to eGFR significantly improved the model’s performance (p = 0.02), as did the addition of GDF-15 levels to eGFR (p = 0.03). When GDF-15 levels were added to the combined model of eGFR+NGAL, there was a trend towards an improvement in the global Chi-square coefficient of the model (p = 0.06).
(One way repeated measures ANOVA followed by pair-wise comparison of mean levels with Tukey test; Spearman’s correlation, n = 34).
Bar graph illustrating the change in global χ2 value by the addition of post-operative GDF-15 and/or NGAL to post-operative eGFR. All models were significantly associated with AKI (p<0.05). * = p<0.05 vs eGFR model. § = p = 0.06 vs eGFR + NGAL model.
Discussion
During CABG, ischemia-reperfusion injury may occur because of elevated intracellular levels of reactive oxygen species and calcium. CPB is associated with an inflammatory response that can generate systemic oxidative stress, which, in turn, causes further damage not only to the heart, but also to other organs. Therefore, it is necessary to identify new biomarkers of inflammation and oxidative stress during this surgical procedure, in order to predict the severity of damage to vital organs.
Determinants of GDF-15 at baseline
Our data showed that pre-operative GDF-15 levels in patients undergoing bypass surgery were positively associated with age, a history of diabetes, chronic renal failure, high NT-proBNP and plasma CRP levels. Clinical studies have shown that GDF-15 concentrations correlated strongly with age in both healthy and unhealthy adults[15]. Ho et al. found that GDF-15 concentrations in apparently healthy elderly adults were higher than those in younger adults [22]. More recently, it was reported that these changes could reflect both cardiovascular and renal perturbations, inflammation and other independent pathophysiological processes[23]. The association between elevated GDF-15 levels and diabetes has also been reported. In obese non-diabetic individuals, GDF-15 could predict future insulin resistance and impaired glucose control[24], and in patients free of clinically-overt cardiovascular disease, GDF-15 was positively associated with age, diabetes, hypertension, worse kidney function[22] and NT-proBNP levels[23]. Indeed, several studies have also reported higher plasma GDF-15 levels in patients with cardiovascular pathologies such as CAD [25], ACS[7], [8], [26] or chronic heart failure[9], [27]. Furthermore, in patients with CAD, GDF-15 concentrations correlated with other markers of inflammation[28], and elevated GDF-15 levels are now considered a strong biomarker of risk progression and mortality in cardiovascular disease[29]–[31]. GDF-15 expression is up-regulated in chronic kidney disease and represents a novel independent serum marker of mortality[13].
Evolution of GDF-15 during CABG
This is the first report of the time course of GDF-15 during cardiac surgery. Our study demonstrated that plasma GDF-15 levels in humans increased during cardiac surgery associated with CPB. Concerning myocardial IR injury during CABG, the increase in plasma GDF-15 levels that we report is in accordance with experimental findings[6]. Rat cardiomyocytes subjected to simulated IR secreted GDF-15, but only after 3 hours of IR. In an in vivo mouse model of cardiac IR injury, myocardial GDF-15 mRNA levels rapidly increased in the ischemic area after both permanent and transient coronary artery ligation. In our study, the median duration of ischemia (cross-clamp time) was 61.7±3.7 minutes, and we examined the levels of GDF-15 after unclamping and after 1 to 2 hours of reperfusion; showing that, in humans, plasma GDF-15 levels increased significantly after a period of myocardial IR.
Heart surgery with CPB is associated with an acute inflammatory response, which has implications for postoperative recovery and myocardial function[32]. Recent findings also suggested that the secreted protein FSTL-1 could be an upstream inducer of GDF15 production and an independent prognostic biomarker in acute coronary syndromes[11]. FSTL1 is a cardiokine, whose cardiac expression levels are upregulated in hypertrophic hearts of mice or in myocardial infarction models[33]. FSTL1 can serve as a useful biomarker of cardiovascular disease given that elevated circulating levels of FSTL1 are found in patients with ACS and associated with chronic HF[34]. In our study performed in patients undergoing bypass surgery, although IND plasma levels of FSTL1 were significantly associated with LVEF (r = −0.385, p = 0.025) and with NT-proBNP concentrations (r = 0.350, p = 0.043), we were unable to demonstrate an association between GDF-15 levels and FSTL-1, at any time during the surgical procedure or the recovery period.
To summarize, our data emphasize the role of GDF-15 in inflammatory situations, and, as suggested by the rapid increase in levels of GDF-15, showed that this induction occurs quickly in cardiac surgery associated with CPB.
CABG and oxidative stress
Increases in various markers of systemic oxidative stress and in inflammatory products have been shown to occur during CPB[3]. Neutrophils are thought to be the primary source of systemic radical oxygen species (ROS) during heart surgery on CPB [3]. MPO is one of the major neutrophil effector proteins and levels of MPO have been associated with a variety of clinical conditions including systemic inflammation, risk of cardiovascular events, vascular endothelial dysfunction and oxidative stress[35]_ENREF_63. MPO produces hypochlorous acid from hydrogen peroxide and chloride anion during the neutrophil's respiratory burst. In the present work, circulating hydroperoxides did not increase throughout the surgical procedure, even though the plasma antioxidant status significantly decreased and plasma MPO levels increased significantly just after aortic cross clamping. In a previous study, we observed that the PAS decreased during CPB[3], reflecting the reduced capacity of plasma to protect its environment from free-radical aggression; this phenomenon is probably triggered to a large extent by hemodilution. Additionally, it should be noted that mannitol, a well-known non-specific antioxidant, administered just before myocardial reperfusion, could have influenced the level of oxidative stress in these patients. Kempf et al.[6] reported that nitro-oxidative stress induced by peroxynitrite was able to induce the expression of GDF-15 in cultured rat neonatal cardiomyocytes. However, at this point in our study, it seems difficult to draw any conclusions on the possible interaction between circulating markers of oxidative stress and GDF-15 levels in the clinical setting of cardiac surgery associated with CPB.
GDF-15 and markers of renal and cardiac injury during CABG
We did find a close association between high GDF-15 levels and worse eGFR. As our population was small, few adverse events were recorded. There were only 9 AKI events, and the results of the multivariate analysis therefore need to be interpreted with caution. In particular, we were unable to show that the prognostic value of the early measurement of GDF-15 was better than a model that associated two powerful biomarkers of the AKIN. Nonetheless, this result supports our findings that GDF-15 is closely associated with renal function and, in combination with known biomarkers, such as NGAL and eGFR, could be of use for the early detection of AKI. This is in agreement with previous results showing that GDF-15 was a novel independent serum marker of mortality in chronic kidney disease[13]. Correlations between GDF-15 levels and renal function biomarkers, such as creatinine, urea or NGAL levels, became significant during the post-surgery period and the following day, suggesting that GDF-15 possibly originates in the kidney or is influenced by renal function. Elevated urinary levels of GDF-15 were found in type-2 diabetic patients with subnormal eGFR, and were associated with proximal tubule injury[36]. NGAL was originally identified in neutrophil granules but is also expressed in the kidney and liver, and its synthesis is induced in response to inflammation, infection, ischemia, and acute kidney injury[18]. Several studies investigating cardiac surgery patients have shown that NGAL levels in both the blood and urine could be a useful predictive marker of acute kidney injury[19].
While there was no correlation between preoperative GDF-15 and cTnI and CK, we found a significant positive association between GDF-15 and NT-proBNP during the post-operative reperfusion period, and there was a significant positive correlation between ICU GDF-15 and cTnI. In experimental conditions[6], IR injury was shown to induce GDF-15 in cardiomyocytes and mouse hearts. Several clinical studies have clearly demonstrated that GDF-15 was a strong biomarker of cardiovascular disease, and a prognosis marker of fatal events in patients with myocardial infarction[7], [8], [30] and chronic HF[9], [15]. Indeed, plasma GDF-15 is elevated in patients with CAD, and levels are even higher in patients with STEMI. GDF-15 is independently associated with NT-proBNP and cardiac troponin T (cTnT) levels at presentation in STEMI patients[8], [26]. Persisting elevated GDF-15 levels may be related to myocardial injury, revealed by the increase in circulating levels of cTnI. However, to our knowledge, this is the first study to report that elevated GDF-15 levels are positively associated with markers of cardiac injury after CABG with CPB.
GDF-15 and other risk factors during CABG
We found a significant association between GDF-15 and hemoglobin levels during the post-operative period. GDF-15 is known to be implicated in dyserythropoietic syndromes such as thalassemia or sickle cell disease[37]. Theurl et al. demonstrated that GDF15 was significantly increased in chronic anemia in subjects with or without true iron deficiency[38]_ENREF_69. GDF-15 is probably involved in mechanisms that regulate hemoglobin levels; however, further studies are needed to elucidate these pathways.
Finally, our results showed that higher pre-, per- and post-operative GDF-15 levels were significantly associated with higher EuroSCORE values. Our results are in agreement with a recent study showing that pre-operative plasma GDF-15 levels improved the predictive capacity of the additive EuroSCORE. Collectively, these results suggest that the addition of GDF-15 to conventional risk stratification tools could provide important prognostic information[12].
Conclusions
In conclusion, our prospective study demonstrated for the first time the kinetic increase in plasma GDF-15 levels during cardiac surgery in patients undergoing CABG associated with CPB. We confirm here the important role of GDF-15 as a marker of renal and cardiac injury in patients with CAD who underwent CABG associated with CPB. Given the value of GDF-15 as a marker of heart disease and cardiovascular risk and the prevalence of major adverse cardiovascular events associated with the post-CPB period, it seems important to measure GDF-15 levels in patients undergoing CABG with CPB so as to improve diagnosis and follow-up.
Acknowledgments
The authors thank Françoise Bechet, Norenn Baverel and Séverine Claes for technical assistance. The authors thank Mr. Philip Bastable for English revision of the manuscript.
Author Contributions
Conceived and designed the experiments: CV LR CG. Performed the experiments: AK KS. Analyzed the data: MZ CV CG. Contributed reagents/materials/analysis tools: SG. Wrote the paper: CV AK CG.
References
- 1. Wildhirt SM, Schulze C, Schulz C, Egi K, Brenner P, et al. (2001) Reduction of systemic and cardiac adhesion molecule expression after off-pump versus conventional coronary artery bypass grafting. Shock 16 Suppl 155–59.
- 2. Goudeau JJ, Clermont G, Guillery O, Lemaire-Ewing S, Musat A, et al. (2007) In high-risk patients, combination of antiinflammatory procedures during cardiopulmonary bypass can reduce incidences of inflammation and oxidative stress. J Cardiovasc Pharmacol 49: 39–45.
- 3. Clermont G, Vergely C, Jazayeri S, Lahet JJ, Goudeau JJ, et al. (2002) Systemic free radical activation is a major event involved in myocardial oxidative stress related to cardiopulmonary bypass. Anesthesiology 96: 80–87.
- 4. Shi Y, Massague J (2003) Mechanisms of TGF-beta signaling from cell membrane to the nucleus. Cell 113: 685–700.
- 5. Ago T, Sadoshima J (2006) GDF15, a cardioprotective TGF-beta superfamily protein. Circ Res 98: 294–297.
- 6. Kempf T, Eden M, Strelau J, Naguib M, Willenbockel C, et al. (2006) The transforming growth factor-beta superfamily member growth-differentiation factor-15 protects the heart from ischemia/reperfusion injury. Circ Res 98: 351–360.
- 7. Wollert KC, Kempf T, Peter T, Olofsson S, James S, et al. (2007) Prognostic value of growth-differentiation factor-15 in patients with non-ST-elevation acute coronary syndrome. Circulation 115: 962–971.
- 8. Khan SQ, Ng K, Dhillon O, Kelly D, Quinn P, et al. (2009) Growth differentiation factor-15 as a prognostic marker in patients with acute myocardial infarction. Eur Heart J 30: 1057–1065.
- 9. Kempf T, von Haehling S, Peter T, Allhoff T, Cicoira M, et al. (2007) Prognostic utility of growth differentiation factor-15 in patients with chronic heart failure. J Am Coll Cardiol 50: 1054–1060.
- 10. Hochholzer W, Morrow DA, Giugliano RP (2010) Novel biomarkers in cardiovascular disease: update 2010. Am Heart J 160: 583–594.
- 11. Widera C, Giannitsis E, Kempf T, Korf-Klingebiel M, Fiedler B, et al. (2012) Identification of follistatin-like 1 by expression cloning as an activator of the growth differentiation factor 15 gene and a prognostic biomarker in acute coronary syndrome. Clin Chem 58: 1233–1241.
- 12. Heringlake M, Charitos EI, Gatz N, Kabler JH, Beilharz A, et al. (2013) Growth differentiation factor 15: a novel risk marker adjunct to the EuroSCORE for risk stratification in cardiac surgery patients. J Am Coll Cardiol 61: 672–681.
- 13. Breit SN, Carrero JJ, Tsai VW, Yagoutifam N, Luo W, et al. (2012) Macrophage inhibitory cytokine-1 (MIC-1/GDF15) and mortality in end-stage renal disease. Nephrol Dial Transplant 27: 70–75.
- 14. Ho JE, Hwang SJ, Wollert KC, Larson MG, Cheng S, et al. (2013) Biomarkers of cardiovascular stress and incident chronic kidney disease. Clin Chem 59: 1613–1620.
- 15. Kempf T, Horn-Wichmann R, Brabant G, Peter T, Allhoff T, et al. (2007) 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 53: 284–291.
- 16. Doise JM, Aho LS, Quenot JP, Guilland JC, Zeller M, et al. (2008) Plasma antioxidant status in septic critically ill patients: a decrease over time. Fundam Clin Pharmacol 22: 203–209.
- 17. Lorgis L, Zeller M, Dentan G, Sicard P, Richard C, et al. (2010) The free oxygen radicals test (FORT) to assess circulating oxidative stress in patients with acute myocardial infarction. Atherosclerosis 213: 616–621.
- 18. Mishra J, Ma Q, Prada A, Mitsnefes M, Zahedi K, et al. (2003) Identification of neutrophil gelatinase-associated lipocalin as a novel early urinary biomarker for ischemic renal injury. J Am Soc Nephrol 14: 2534–2543.
- 19. Wagener G, Jan M, Kim M, Mori K, Barasch JM, et al. (2006) Association between increases in urinary neutrophil gelatinase-associated lipocalin and acute renal dysfunction after adult cardiac surgery. Anesthesiology 105: 485–491.
- 20. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, et al. (1999) A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 130: 461–470.
- 21. Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, et al. (2007) Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 11: R31.
- 22. Ho JE, Mahajan A, Chen MH, Larson MG, McCabe EL, et al. (2012) Clinical and genetic correlates of growth differentiation factor 15 in the community. Clin Chem 58: 1582–1591.
- 23. Eggers KM, Kempf T, Wallentin L, Wollert KC, Lind L (2013) Change in growth differentiation factor 15 concentrations over time independently predicts mortality in community-dwelling elderly individuals. Clin Chem 59: 1091–1098.
- 24. Kempf T, Guba-Quint A, Torgerson J, Magnone MC, Haefliger C, et al. (2012) Growth differentiation factor 15 predicts future insulin resistance and impaired glucose control in obese nondiabetic individuals: results from the XENDOS trial. Eur J Endocrinol 167: 671–678.
- 25. Rohatgi A, Patel P, Das SR, Ayers CR, Khera A, et al. (2012) Association of growth differentiation factor-15 with coronary atherosclerosis and mortality in a young, multiethnic population: observations from the Dallas Heart Study. Clin Chem 58: 172–182.
- 26. Kempf T, Bjorklund E, Olofsson S, Lindahl B, Allhoff T, et al. (2007) Growth-differentiation factor-15 improves risk stratification in ST-segment elevation myocardial infarction. Eur Heart J 28: 2858–2865.
- 27. Foley PW, Stegemann B, Ng K, Ramachandran S, Proudler A, et al. (2009) Growth differentiation factor-15 predicts mortality and morbidity after cardiac resynchronization therapy. Eur Heart J 30: 2749–2757.
- 28. Zhang M, Lu S, Wu X, Chen Y, Song X, et al. (2012) Multimarker approach for the prediction of cardiovascular events in patients with mild to moderate coronary artery lesions. A 3-year follow-up study. Int Heart J 53: 85–90.
- 29. Wollert KC (2007) Growth-differentiation factor-15 in cardiovascular disease: from bench to bedside, and back. Basic Res Cardiol 102: 412–415.
- 30. Wollert KC, Kempf T, Lagerqvist B, Lindahl B, Olofsson S, et al. (2007) Growth differentiation factor 15 for risk stratification and selection of an invasive treatment strategy in non ST-elevation acute coronary syndrome. Circulation 116: 1540–1548.
- 31. Xu X, Li Z, Gao W (2011) Growth differentiation factor 15 in cardiovascular diseases: from bench to bedside. Biomarkers 16: 466–475.
- 32. Warren OJ, Smith AJ, Alexiou C, Rogers PL, Jawad N, et al. (2009) The inflammatory response to cardiopulmonary bypass: part 1-mechanisms of pathogenesis. J Cardiothorac Vasc Anesth 23: 223–231.
- 33. Oshima Y, Ouchi N, Sato K, Izumiya Y, Pimentel DR, et al. (2008) Follistatin-like 1 is an Akt-regulated cardioprotective factor that is secreted by the heart. Circulation 117: 3099–3108.
- 34. Widera C, Horn-Wichmann R, Kempf T, Bethmann K, Fiedler B, et al. (2009) Circulating concentrations of follistatin-like 1 in healthy individuals and patients with acute coronary syndrome as assessed by an immunoluminometric sandwich assay. Clin Chem 55: 1794–1800.
- 35. van der Veen BS, de Winther MP, Heeringa P (2009) Myeloperoxidase: molecular mechanisms of action and their relevance to human health and disease. Antioxid Redox Signal 11: 2899–2937.
- 36. Simonson MS, Tiktin M, Debanne SM, Rahman M, Berger B, et al. (2012) The renal transcriptome of db/db mice identifies putative urinary biomarker proteins in patients with type 2 diabetes: a pilot study. Am J Physiol Renal Physiol 302: F820–829.
- 37. Tanno T, Bhanu NV, Oneal PA, Goh SH, Staker P, et al. (2007) High levels of GDF15 in thalassemia suppress expression of the iron regulatory protein hepcidin. Nat Med 13: 1096–1101.
- 38. Theurl I, Finkenstedt A, Schroll A, Nairz M, Sonnweber T, et al. (2010) Growth differentiation factor 15 in anaemia of chronic disease, iron deficiency anaemia and mixed type anaemia. Br J Haematol 148: 449–455.