- Email: [email protected]
Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients
Journal of Critical Care xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Journal of Critical Care journal homepage: www.jccjournal.org.
Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients☆ Alexander Koch a, Ralf Weiskirchen b, Julian Kunze a, Hanna Dückers a, Jan Bruensing a, Lukas Buendgens a, Michael Matthes a, Tom Luedde a, Christian Trautwein a, Frank Tacke a,⁎ a b
Department of Medicine III, RWTH-University Hospital Aachen, Pauwelsstrasse 30, Aachen, Germany Institute of Clinical Chemistry and Pathobiochemistry, RWTH-University Hospital Aachen, Pauwelsstrasse 30, Aachen, Germany
a r t i c l e Keywords: ADMA ICU Prognosis Sepsis Organ failure
i n f o
a b s t r a c t Objective: Serum concentrations of asymmetric dimethylarginine (ADMA), an endogenous inhibitor of nitric oxide synthase, may contribute to endothelial dysfunction and organ failure in sepsis. We aimed at investigating ADMA levels as a potential diagnostic or prognostic biomarker in critically ill patients. Methods: Two hundred fifty-five patients (164 with sepsis, 91 without sepsis) were studied prospectively upon admission to the medical intensive care unit (ICU) and on day 7, in comparison to 78 healthy controls. ADMA serum concentrations were correlated with clinical data and extensive laboratory parameters. Patients’ survival was followed up for up to 3 years. Results: ADMA serum levels were significantly elevated in critically ill patients at admission compared to controls. ADMA levels did not differ between patients with or without sepsis, but were closely related to hepatic and renal dysfunction, metabolism and clinical scores of disease severity. ADMA levels further increased during the first week of ICU treatment. ADMA serum levels at admission were an independent prognostic biomarker in critically ill patients not only for short-term mortality at the ICU, but also for unfavorable long-term survival. Conclusion: Serum ADMA concentrations are significantly elevated in critically ill patients, associated with organ failure and related to short- and long-term mortality risk. © 2013 Elsevier Inc. All rights reserved.
1. Introduction Systemic inflammation and sepsis, associated with multiple organ dysfunctions, are key characteristics of patients with critical illness requiring treatment at the intensive care unit (ICU). Nonspecific clinical and physiologic criteria of systemic inflammatory response syndrome and sepsis are used to identify patients at immediate need for ICU therapy. A complex network of biological mediators is dysregulated in critically ill patients, and several circulating parameters have been evaluated as potential biomarkers with diagnostic or even prognostic value [1]. There is increasing experimental evidence that the arginine-nitric oxide (NO) pathway is crucially involved in inflammation, infection and organ injury [2]. NO is a potent vasodilatator, which is produced from L-arginine by the enzyme nitric oxide synthase [3]. Asymmetric dimethylarginine (ADMA), on the other hand, is a naturally occurring non-selective inhibitor of NO synthase [4]. Circulating ADMA is mainly
☆ Competing interests: None of the authors declares competing interests. ⁎ Corresponding author. Department of Medicine III, RWTH-University Hospital Aachen, Pauwelsstraße 30, 52074 Aachen, Germany. Tel.: +49 241 80 35848; fax: +49 241 80 82455. E-mail address: [email protected] (F. Tacke).
derived from protein catabolism, because dimethylarginines are released as proteins are hydrolyzed, thus representing an obligatory product of protein turnover [5]. Functionally, ADMA impairs (beneficial) NO-dependent endothelial functions such as vascular dilatation or anti-inflammatory processes [4]. In line, circulating ADMA levels have been linked to endothelial dysfunction in cardiovascular diseases and risk factors for atherosclerosis [6]. It had been speculated that ADMA is also involved in the pathogenesis of microvascular dysfunction in sepsis, as increased ADMA could possibly decrease NO bioavailability at the endothelium [7]. Few small studies have investigated circulating ADMA as a novel biomarker in critical care medicine and found elevated ADMA serum concentrations in patients with acute infections and sepsis [7–11]. Although the exact mechanisms of ADMA regulation in these patients remained obscure, ADMA was suggested as a predictor of ICU mortality [7,8,12]. However, these initial studies were limited by their focus on patients with sepsis, by their relatively small cohorts and by the short observation period with respect to mortality as an endpoint. The aim of this study was to investigate ADMA serum concentrations in a large cohort of 255 consecutively enrolled critically ill patients, with or without sepsis, identify associations between ADMA and organ dysfunction, metabolism, and disease severity as well as to assess the prognostic
0883-9441/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcrc.2013.05.016
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
2
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
value of ADMA to predict ICU and long-term mortality in critically ill patients. 2. Materials and methods 2.1. Study design and patient characteristics Written informed consent was obtained from the patient, his or her spouse or the appointed legal guardian. Patients who were expected to have a short-term (b72 h) intensive care treatment due to post-interventional observation or acute intoxication were not included in this study [13]. All patient data, clinical information and blood samples were collected prospectively. The clinical course of patients was observed in a follow-up period by directly contacting the patients, the patients’ relatives or their primary care physician. Patients who met the criteria proposed by the American College of Chest Physicians & the Society of Critical Care Medicine Consensus Conference Committee for severe sepsis and septic shock were categorized as sepsis patients, the others as nonsepsis patients [14]. The study protocol was approved by the local ethics committee and conducted in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki (ethics committee of the University Hospital Aachen, RWTH-University, Aachen, Germany, reference number EK 150/06). 2.2. ADMA measurements Blood samples were collected upon admission to the ICU as well as in the morning of day 7 after admission. Importantly, ADMA levels at admission were obtained prior to therapeutic inventions at the ICU, which could potentially influence glucose metabolism, such as parenteral nutrition or insulin administration. After centrifugation at 4°C for 10 minutes, serum and plasma aliquots of 1 mL were frozen immediately at − 80°C. ADMA serum concentrations were analyzed using a commercial enzyme immunoassay (Immundiagnostik, Bensheim, Germany). The scientist performing experimental measurements was fully blinded to any clinical or other laboratory data of the patients or controls. 2.3. Statistical analysis Data are given as median and range due to the skewed distribution of most of the parameters. Differences between two groups were assessed by Mann-Whitney U test and multiple comparisons between more than two groups have been conducted by Kruskal-Wallis analysis of variance and Mann-Whitney U test for post hoc analysis. Box plot graphics illustrate comparisons between subgroups and they display a statistical summary of the median, quartiles, range, and extreme values. The whiskers extend from the minimum to the maximum value excluding outside and far out values which are displayed as separate points. An outside value (indicated by an open circle) was defined as a value that is smaller than the lower quartile minus 1.5 times interquartile range, or larger than the upper quartile plus 1.5 times the interquartile range. A far out value was defined as a value that is smaller than the lower quartile minus three times interquartile range, or larger than the upper quartile plus three times the interquartile range [15]. All values, including “outliers”, have been included for statistical analyses. Correlations between variables have been analyzed using the Spearman correlation tests, where values of P b .05 were considered statistically significant. All single parameters that correlated significantly with ADMA levels at admission were included in a multivariate linear regression analysis with ADMA as the dependent variable to identify independent (meaningful) predictors of elevated ADMA. The prognostic value of the variables was tested by univariate and multivariate analysis in the Cox regression model. Kaplan Meier curves were plotted to display the impact on survival
[16]. Receiver operating characteristic (ROC) curve analysis and the derived area under the curve (AUC) statistic provide a global and standardized appreciation of the accuracy of a marker or a composite score for predicting an event. ROC curves were generated by plotting sensitivity against 1-specificity [17]. All statistical analyses were performed with SPSS (SPSS, Chicago, IL).
3. Results 3.1. ADMA serum levels are significantly elevated in critically ill patients, but not related to sepsis In order to investigate ADMA in critical illness, we measured ADMA serum concentrations in a large cohort of medical ICU patients at admission (before therapeutic intervention) and on day 7 (Table 1). We enrolled 255 patients (149 male, 106 female with a median age of 63 years; range 18-90 years) who were admitted to the General Internal Medicine ICU at the RWTH-University Hospital Aachen, Germany (Table 1). As a control population we analyzed 78 healthy blood donors (50 male, 28 female; median age 30, range 18-67 years) with normal values for blood counts, C-reactive protein and liver enzymes. ADMA serum levels were significantly higher in ICU patients (n = 255, median 0.48 μmol/L, range 0.14-2.0) as compared with healthy controls (n = 78, median 0.36 μmol/L, range 0.23-0.57, P b .001; Fig.1 A). No association between ADMA levels and sex or age were observed in controls (data not shown). Among the 255 critically ill patients enrolled in this study, 164 patients conformed to the criteria of bacterial sepsis (Table 1). Pneumonia was identified in the majority of sepsis patients as origin of infection (not shown). Non-sepsis patients were admitted to the ICU mainly due to cardiopulmonary diseases (myocardial infarction, pulmonary embolism, and acute decompensated heart failure), decompensated liver cirrhosis or other critical conditions and did not differ in age or sex from sepsis patients. Sepsis patients were more often in need of mechanical ventilation in longer terms as compared to the non-sepsis patients’ cohort (Table 2). In sepsis patients significantly higher levels of routinely used biomarkers of inflammation (ie, C-reactive protein, procalcitonin, white blood cell count) were found (data not shown). Both groups did not differ in Acute Physiology and Chronic Health Evaluation (APACHE) II-, Sequential Organ Failure Assessment (SOFA) and simplified acute physiology score-2 score, vasopressor demand, or laboratory parameters indicating liver or renal dysfunction (Tables 1-2, and data not shown). Among all critical care patients about 25% died at the ICU; during the follow-up period of up to 3 years, a total of 47% of the initial cohort had died (Table 2). By direct comparison between septic and non-septic patients, ADMA serum levels did not differ between patients with or without sepsis (Table 1, Fig. 1B). In 44 patients, paired blood samples were available for ADMA measurements at ICU admission and at day 7 of ICU treatment. Remarkably, individual ADMA levels increased during the first week Table 1 Baseline patient characteristics and ADMA serum measurements Parameter
All patients
Sepsis
Non-sepsis
Number Sex (male/female) Age median (range) [years] APACHE-II score median (range) SOFA score median (range) pre-existing diabetes n(%) ADMA day 1 median (range)[μmol/L] ADMA day 7 median (range) [μmol/L]
255 149/106 63 (18-90) 17 (2-43)
164 96/68 64 (20-90) 19 (3-43)
91 53/38 61 (18-85) 15 (2-33)
9.5 (0-19) 11 (3-19) 7 (0-16) 75 (29.4%) 46 (28%) 29 (31.9%) 0.48 (0.14-2.00) 0.47 (0.14-2.00) 0.49 (0.16-1.65) 0.71 (0.43-1.94) 0.70 (0.43-1.94) 0.78 (0.51-1.62)
For quantitative variables, median and range (in parenthesis) are given.
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
A 2.0
B
2.0
n.s.
P < .001 1.5
ADMA µmol/L
ADMA µmol/L
1.5
1.0
1.0
0.5
0.5
0
0
controls
patients
n = 78
n = 255
C 2.0
no n = 91
D
2.0
P < .001
sepsis
yes n = 164
ADMA day 1 ADMA day 7
1.5
1.5
ADMA µmol/L
ADMA µmol/L
3
1.0
P < .001 1.0
0.5
0.5
0
0
ADMA day 1 n = 44
P < .001
no n = 11
ADMA day 7 n = 44
sepsis
yes n = 33
Fig. 1. Serum ADMA concentrations in critically ill patients. (A) At admission to the Medical ICU, serum ADMA levels were significantly (P b .001, U test) elevated in critically ill patients (n = 255) as compared to healthy controls (n = 78). (B) ADMA serum levels at ICU admission did not differ between patients with or without sepsis. (C-D) In 44 patients, ADMA levels were measured at admission (day 1) and at 1 week (day 7) of ICU therapy. ADMA levels increased significantly (P b .001, paired Wilcoxon-test) during treatment at the ICU (C), independent of sepsis (D).
of ICU therapy (Table 1, Fig. 1C; P b .001, paired Wilcoxon-Test). This was observed for patients with sepsis as well as for non-sepsis patients (Fig. 1D).
3.2. ADMA serum concentrations at admission to the ICU are closely correlated to organ function, metabolism and clinical scores To determine the factors possibly promoting elevated serum ADMA levels in critically ill patients, correlation analyses with laboratory parameters were performed. For these analyses, serum ADMA levels at admission were applied, in order to exclude possible confounding effects due to patients that died or were discharged from the ICU during the first week. At admission to the ICU, serum ADMA concentrations were closely correlated to biomarkers displaying hepatic and renal dysfunction. In detail, ADMA was found to correlate significantly with markers reflecting the hepatic biosynthetic capacity (eg, international normalized ratio [INR], r = 0.272, P b .001; albumin, r =−0.285, P b .001) and also to parameters indicating cholestasis (eg, bilirubin, r = 0.361, P b .001; gamma-GT, r = 0.357, P b .001) as well as with markers indicating renal dysfunction (eg, creatinine, r = 0.205, P = .001; cystatin C, r = 0.334, P b .001). In line, patients with preexisting renal failure, defined as a glomerular filtration rate for cystatin C b 50 mL/min, or with preexisting hepatic dysfunction,
defined as prothrombin time b50%, showed significantly elevated ADMA levels (Fig. 2A and B, P b .001, U test). Besides its relation to cardiovascular risk, ADMA levels had been linked to metabolic syndrome and type 2 diabetes [4]. In our cohort of critically ill patients, ADMA levels were significantly decreased in ICU patients with a pre-existing type 2 diabetes upon admission, but did not differ between patients with obesity, defined as a body mass index N30 kg/m 2 (detailed data not shown). Interestingly, ADMA was also correlated to established clinical scores indicating disease severity, such as the APACHE II (r = 0.154, P = .032) or SOFA score (r = 0.224, P = .021), as well as to soluble urokinase plasminogen activator receptor (suPAR, r = 0.434, P b .001), a prognostic biomarker in ICU patients [18]. Table 2 Outcome measures of the study cohort Parameter
All patients
Sepsis
Non-sepsis
Number ICU days median (range) Death during ICU n (%) Death during follow-up, n (%) Mechanical ventilation, n (%) Ventilation time median (range) [h]
255 8 (1-137) 63 (24.7%) 120 (47.1%) 186 (72.9%) 122 (0-3628)
164 10 (1-137) 47 (28.7%) 88 (53.7%) 120 (73.2%) 183 (0-2966)
91 6 (1-45) 16 (17.6%) 32 (35.2%) 61 (67%) 46 (0-3628)
For quantitative variables, median and range (in parenthesis) are given.
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
4
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
A
B 2.0
2.0
P < .001
P < .001 1.5
ADMA µmol/L
ADMA µmol/L
1.5
1.0
1.0
0.5
0.5
0
0 GFR >50 ml/min
GFR <50 ml/min
GFR cystatin C day 1
PT >50%
PT <50%
prothrombin time day 1
Fig. 2. Serum ADMA concentrations in critically ill patients with organ failure and comorbidities. Serum ADMA levels were measured in n = 255 critically ill patients at admission to the ICU. Patients with renal failure (cystatin C-based glomerular filtration rate [GFR] b50 mL/min, A) or hepatic failure (prothrombin time b50%, B) had significantly elevated ADMA levels.
When all parameters that were correlated with ADMA serum levels by univariate analysis were included in a multivariate regression analysis, only albumin (P = .040) and suPAR (P b .001), but not renal function markers or metabolic parameters, remained independent predictors of ADMA concentrations (Table 3).
3.3. ADMA serum levels are an independent prognostic biomarker for ICU mortality in critically ill patients Based on the close correlation between ADMA levels at admission and disease severity scores, we hypothesized that circulating ADMA might be capable of identifying patients at high risk for mortality. Indeed, patients that died during the course of ICU treatment (about one quarter of the total cohort) had significantly higher serum ADMA levels at admission compared with the ICU survivors (median 0.62 vs. 0.44 μmol/L, P b .001; Fig. 3A). We thus performed Cox regression analyses and Kaplan-Meier curves to assess the impact of the initial ADMA serum concentrations on ICU-mortality among critically ill patients. Low ADMA levels upon admission to the ICU were a strong prognostic predictor for ICU-survival (P b .001, Cox regression analyses). In multivariate Cox regression analyses, including markers of inflammation/infection (CRP, WBC), hematologic (hematocrit, platelet count), circulatory (lactate), hepatic (bilirubin, INR) and renal (creatinine, urea) deterioration at admission, ADMA remained an independent significant prognostic parameter (hazard ratios and p-values are presented in Table 4). Kaplan-Meier curves displayed that patients with ADMA levels of the upper quartile (N0.63 μmol/L) had highest mortality (Fig. 3 B). We found the best cut-off value to discriminate survivors from non-ICU-survivors for serum ADMA of 0.65 μmol/L (Fig. 3C). Table 3 Multivariate regression analysis of parameters predicting ADMA levels Parameter
Standardized coefficient beta
t value
P
suPAR Albumin Creatinine Bilirubin INR APACHE-II
0.479 −0.202 0.055 0.011 −0.021 −0.121
4.220 −2.079 0.598 0.110 −0.222 −1.233
b.001 0.040 n.s. n.s. n.s. n.s.
suPAR, soluble urokinase plasminogen activator receptor.
By ROC curve analyses, we compared the prognostic value of ADMA to other clinically used biomarkers. In comparison to biomarkers reflecting hepatic or renal organ dysfunction (INR, creatinine) or inflammation (CRP), ADMA levels displayed higher prognostic power (Fig. 3D). The prognostic value of serum ADMA levels at admission (AUC 0.636, 95% CI 0.513-0.760) was comparable to clinical disease severity scores such as APACHE-II (AUC 0.662, 95% CI 0.533-0.791) or SOFA (AUC 0.667, 95% CI 0.549-0.785). Nevertheless, the use of the individual increase in ADMA between admission to the ICU and day 7 of ICU treatment, as available for 44 patients with matched measurements, did not further increase the prognostic value of ADMA (AUC 0.583). 3.4. High ADMA levels at ICU admission indicate unfavorable long-term prognosis During the follow-up observation period of approximately three years, the overall case fatality rate increased to 47.1% of the study cohort (Table 1). ADMA serum concentrations at admission to the ICU were significantly higher in patients with unfavorable outcome (median 0.54 vs. 0.43 μmol/L, P b .001, Fig. 4A). By Cox regression analysis, initial serum ADMA levels significantly predicted long-term prognosis (P b .001). The prognostic value remained significant also by multivariate analysis (Table 5). Kaplan-Meier curves proved that ADMA levels of the highest quartile (N0.63 μmol/L) were strongly associated with fatal outcome (Fig. 4B). The log-rank test value of the Kaplan-Meier curves for ADMA quartiles was substantially higher for the overall (32.88) than for the ICU mortality (16.29). Again, ADMA levels of 0.65 μmol/L discriminated the long-term prognosis of critically ill patients (Fig. 4C). In accordance with the findings for ICU mortality, serum ADMA levels at ICU admission had higher prognostic value for long-term outcome as compared with single biomarkers (Fig. 4D). Interestingly, the prognostic accuracy of serum ADMA levels could be even improved, if ADMA levels were adjusted to total protein concentrations of the patients. Based on the close association between liver dysfunction and ADMA levels and the highly extended mortality risk of critically ill patients with liver failure, we reasoned that normalization of ADMA to protein levels might allow to better delineate the specific prognostic value of ADMA. For ICU mortality, the area-under-the-ROC-curve (AUC)
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
A
B
ICU survival
5
ICU survival 1.0
2.0
ADMA lower 25% ADMA 25-75% ADMA upper 25%
P < .001 cumulative survival
0.8
ADMA µmol/L
1.5
1.0
0.5
0.6
0.4
0.2
log rank 32.88
survivors
deaths
D
40
60
80
100
120
140
1.0
prediction of ICU survival
ADMA < 0.65 µmol/L ADMA > 0.65 µmol/L 0.8
sensitivity
0.8
0.6
0.4
0.2
0
20
time (days)
ICU survival 1.0
cumulative survival
0
n = 63
n = 192
C
P < .001
0
0
0.6
0.4
ADMA INR creatinine CRP
0.2
log rank 8.52 P = .0035 0 0
20
40
60
80
100
120
140
0
time (days)
0.2
0.4
0.6
0.8
1.0
1 - specificity
Fig. 3. Prediction of ICU mortality by ADMA serum concentrations. (A) Patients that died during the course of ICU treatment had significantly higher serum ADMA levels on admittance to ICU (P b .001 than survivors. (B-C) Kaplan-Meier survival curves of ICU patients are displayed, showing that patients with ADMA levels of upper quartile (on admission N0.63 μmol/L; black, B) had an increased short-term mortality at the ICU as compared to patients with ADMA serum concentrations of lower quartile (on admission b0.36 μmol/L; light grey) or middle 50% (grey). Best discrimination between ICU survivors and non-survivors was achieved with an ADMA cut-off value of 0.65 μmol/L (C). P values are given in the figure. (D) ROC curve analyses comparing the prognostic value of ADMA at admission for ICU survival (dark grey, AUC 0.686, 95% CI 0.606-0.765) with INR (light grey, AUC 0.626, 95% CI 0.547-0.706) and creatinine (black, AUC 0.602, 95% CI 0.522-0.683) as makers of hepatic and renal function as well as CRP (dotted line, AUC 0.521, 95% CI 0.442-0.601) as a marker of inflammation and infection.
increased from 0.713 for ADMA itself to 0.765 for ADMA/proteinratio. For overall mortality, the AUC increased from 0.659 for ADMA to 0.709 for ADMA/protein-ratio, respectively. Again, the use of the individual increase in ADMA between admission to the ICU and day
7 of ICU treatment, as available for 44 patients with matched measurements, did not further increase the prognostic value of ADMA (AUC 0.636). 4. Discussion
Table 4 Uni- and multivariate Cox regression analyses for ADMA levels at admission to predict ICU mortality Parameter
Unadjusted HR (95% CI)
P
Adjusted HR (95% CI)
P
ADMA Creatinine Urea Bilirubin INR White blood cell count C-reactive protein Platelets Lactate
3.803 (2.189-6.607) 1.031 (0.951-1.118) 1.001 (1.000-1.003) 1.051 (1.014-1.089) 1.309 (1,037-1.652) 1.015 (1.003-1.028)
b0.001 NS NS 0.007 0.023 0.039
3.141 (1.721-5.730) -
b0.001 NS NS NS NS NS
0.999 (0.996-1.002) 0.999 (0.997-1.001) 1.127 (1.062-1.197)
NS NS b0.001
1.095 (1.030-1.163)
NS NS 0.003
Significant results are highlighted in bold.
Based on the increasing experimental evidence on the importance of the arginine-nitric oxide pathway for inflammation, infection and organ injury, ADMA as the most relevant inhibitor of nitric oxide synthase has been hypothesized as a novel biomarker in critical care medicine [3]. Initial studies have linked ADMA levels in critically ill and/or septic patients with the severity of organ failure, degree of inflammation and the presence of shock in severe sepsis [7,8,12]. In line with these findings from smaller studies, we observed significantly upregulated circulating ADMA levels in 255 prospectively enrolled critically ill patients compared to healthy controls. Unexpectedly, however, ADMA levels did not differ between patients with or without sepsis. An association between ADMA levels and microvascular dysfunction has been recently documented in selected patients with sepsis [7]. As accumulation of ADMA and consequently
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
6
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
A
B
overall survival 2.0
overall survival 1.0
ADMA lower 25% ADMA 25-75% ADMA upper 25%
P < .001 0.8
cumulative survival
ADMA µmol/L
1.5
1.0
0.5
0.6
0.4
0.2
log rank 32.88 0
C
survivors
deaths
n = 131
n = 120
200
400
600
800
1000
time (days)
D 1.0
prediction of overall survival
ADMA < 0.65 µmol/L ADMA > 0.65 µgmol/L 0.8
0.8
sensitivity
cumulative survival
0
overall survival 1.0
0.6
0.4
0.2
0
P < .001
0
0.6
0.4
ADMA INR creatinine CRP
0.2
log rank 13.96 P = .0002 0 0
200
400
600
800
1000
0
0.2
0.4
0.6
0.8
1.0
1 - specificity
time (days)
Fig. 4. Prediction of overall mortality by ADMA serum concentrations. (A) Patients that died during the total observation period had significantly higher serum ADMA levels on admittance to ICU (P b .001 than survivors. (B-C) Kaplan-Meier survival curves of ICU patients are displayed, showing that patients with ADMA levels of upper quartile (on admission N0.63 μmol/L; black, B) had an increased short-term mortality at the ICU as compared to patients with ADMA serum concentrations of lower quartile (on admission b0.36 μmol/L; light grey) or middle 50% (grey). Best discrimination between ICU survivors and non-survivors was achieved with an ADMA cut-off value of 0.65 μmol/L (C). P values are given in the figure. (D) ROC curve analyses comparing the prognostic value of ADMA at admission for ICU survival (dark grey, AUC 0.636, 95% CI 0.566-0.707) with INR (light grey, AUC 0.568, 95% CI 0.496-0.640) and creatinine (black, AUC 0.568, 95% CI 0.495-0.641) as makers of hepatic and renal function as well as CRP (dotted line, AUC 0.572, 95% CI 0.500-0.644) as a marker of inflammation and infection.
reduced NO signaling is known to promote endothelial dysfunction and increased systemic vascular resistance [19], our findings indicate that ADMA up-regulation may represent a uniform response in critical illness, independent of the presence of infection. On the other hand, a reduced renal and/or hepatic clearance might contribute to elevated ADMA levels in critically ill patients, because we Table 5 Uni- and multivariate Cox regression analyses for ADMA levels at admission to predict overall mortality Parameter
Unadjusted HR (95% CI)
P
Adjusted HR (95% CI)
P
ADMA Creatinine Urea Bilirubin INR White blood cell count C-reactive protein Platelets Lactate
3.300 1.014 1.001 1.046 1.010 1.009
(2.082-5.232) (0.967-1.062) (1.000-1.002) (1.023-1.070) (0.996-1.024) (1.001-1.017)
b0.001 NS 0.008 b0.001 0.171 0.034
3.115 (1.862-5.210) 1.002 (1.001-1.003) -
b0.001 0.014 0.001 NS NS NS
1.001 (1.000-1.003) 0.999 (0.998-1.000) 1.075 (1.039-1.111)
NS NS b0.001
1.118 (1.047-1.194)
NS NS 0.001
observed that ADMA serum concentrations correlated closely with hepatic and renal organ function. However, there is no clear experimental evidence at present for hepatic or renal clearance of ADMA. Studies from patients with chronic renal insufficiency and animals models of acute liver failure did not show altered ADMA clearance as a main mechanism of elevated ADMA serum levels [20,21]. From patients with chronic cardiovascular diseases, ADMA levels have been linked to metabolic syndrome and surrogate parameters of insulin resistance [4,22]. In non-diabetic patients, plasma glucose levels were positively correlated with circulating ADMA, and diabetic patients showed highly elevated ADMA levels [22]. This association indicated that ADMA might activate signaling pathways involved in development of the metabolic syndrome [4]. In our study, ADMA levels were not increased in patients with pre-existing type 2 diabetes. This discrepancy between prior studies on chronically diseased patients and our findings in ICU patients might indicate that the physiological regulation of ADMA is massively altered in the acute phase of critical illness, which is characterized by increased protein turnover due to catabolism and the induction of acute phase proteins [12]. Three studies including 47, 52, and 67 critically ill patients, respectively, even indicated that ADMA levels might serve as a
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
potential biomarker predicting ICU mortality [7,8,12]. However, these small studies focused on septic patients and solely assessed (shortterm) ICU mortality. On the basis of 255 consecutively included critically ill patients, we now identified ADMA as a biomarker related to short-term and long-term mortality risk. In comparison with routinely used single biomarkers reflecting organ dysfunction such as creatinine or INR, ADMA levels showed superior prognostic power, reaching a similar accuracy as composite clinical disease severity scores (eg, APACHE-II, SOFA). Importantly, multivariate Cox regression analyses confirmed circulating ADMA as an independent prognostic indicator, specifically independent of hepatic or renal failure, inflammation and even of lactate as a clinically used biomarker for disturbed tissue perfusion [23]. These analyses support that ADMA could be a marker for a combined algorithm for risk in ICU. However, the cut-off values for mortality prediction that have been identified in our study need to be validated in independent patient cohorts. The association of high ADMA levels with increased risk of case-related fatality, both short- and long-term, might be caused by the inhibition of endothelial nitric oxide synthesis and of vascular reactivity, as experimentally demonstrated for animals and humans [24,25]. Interestingly, the prognostic accuracy of serum ADMA levels could be even improved, if ADMA levels were adjusted to total protein concentrations of the patients. This supports that an ADMA increase in conditions of protein catabolism is a distinct indicator of disease severity. A recent study has demonstrated that in patients with severe septic shock, protein catabolism resulted in massively increased plasma arginine levels [8]. Possibly, the L-arginine-to-ADMA ratio might then be reduced, which would imply reduced endothelial nitric oxide availability and impaired microvascular reactivity in these patients [8], potentially linking ADMA levels and mortality risk. The association of ADMA levels with mortality risk raises the question if this reflects a potentially modifiable process in critical illness. In fact, ADMA and its related pathways have been suggested as possible new therapeutic targets in several disease conditions [26]. With respect to sepsis, it is important to consider that ADMA is physiologically eliminated through active metabolism by dimethylarginine dimethylaminohydrolase (DDAH) [3]. Experimental data from DDAH-deficient mice and administration of a chemical DDAH inhibitor had indicated that pharmacological inhibition of DDAH could be harnessed therapeutically to increase the vascular tension during the hyperdynamic state of sepsis [19]. However, our study revealed strongly increased ADMA levels already at the time-point of admission to the ICU before therapeutic interventions in septic and non-septic patients and clearly demonstrated high ADMA levels as an independent prognostic indicator for adverse outcome. Thus, blocking DDAH and consequently further increasing circulating ADMA could also be potentially harmful in critical illness. Therefore, further experimental and clinical studies are clearly warranted to understand the underlying regulatory mechanisms and pathogenic consequences of elevated ADMA serum concentrations in the setting of intensive care medicine. Acknowledgments We sincerely thank Philip Kim for excellent technical assistance. This work was supported by the German Research Foundation (DFG Ta434/ 2-1 & SFB/TRR57) and the Interdisciplinary Centre for Clinical Research (IZKF) within the faculty of Medicine at the RWTH Aachen University.
7
References [1] Pierrakos C, Vincent JL. Sepsis biomarkers: a review. Crit Care 2010;14:R15. [2] Landry DW, Oliver JA. The pathogenesis of vasodilatory shock. N Engl J Med 2001;345:588–95. [3] Boger RH. Live and let die: asymmetric dimethylarginine and septic shock. Crit Care 2006;10:169. [4] Bestermann Jr WH. The ADMA-metformin hypothesis: linking the cardiovascular consequences of the metabolic syndrome and type 2 diabetes. Cardiorenal Med 2011;1:211–9. [5] Leiper J, Vallance P. Biological significance of endogenous methylarginines that inhibit nitric oxide synthases. Cardiovasc Res 1999;43:542–8. [6] Boger RH. Asymmetric dimethylarginine (ADMA): a novel risk marker in cardiovascular medicine and beyond. Ann Med 2006;38:126–36. [7] Davis JS, Darcy CJ, Yeo TW, et al. Asymmetric dimethylarginine, endothelial nitric oxide bioavailability and mortality in sepsis. PLoS One 2011;6:e17260. [8] O'Dwyer MJ, Dempsey F, Crowley V, et al. Septic shock is correlated with asymmetrical dimethyl arginine levels, which may be influenced by a polymorphism in the dimethylarginine dimethylaminohydrolase II gene: a prospective observational study. Crit Care 2006;10:R139. [9] Iapichino G, Umbrello M, Albicini M, et al. Time course of endogenous nitric oxide inhibitors in severe sepsis in humans. Minerva Anestesiol 2010;76: 325–33. [10] Nakamura T, Sato E, Fujiwara N, et al. Circulating levels of advanced glycation end products (AGE) and interleukin-6 (IL-6) are independent determinants of serum asymmetric dimethylarginine (ADMA) levels in patients with septic shock. Pharmacol Res 2009;60:515–8. [11] Zoccali C, Maas R, Cutrupi S, et al. Asymmetric dimethyl-arginine (ADMA) response to inflammation in acute infections. Nephrol Dial Transplant 2007;22: 801–6. [12] Nijveldt RJ, Teerlink T, Van Der Hoven B, et al. Asymmetrical dimethylarginine (ADMA) in critically ill patients: high plasma ADMA concentration is an independent risk factor of ICU mortality. Clin Nutr 2003;22:23–30. [13] Koch A, Gressner OA, Sanson E, Tacke F, Trautwein C. Serum resistin levels in critically ill patients are associated with inflammation, organ dysfunction and metabolism and may predict survival of non-septic patients. Crit Care 2009;13: R95. [14] Koch A, Voigt S, Kruschinski C, et al. Circulating soluble urokinase plasminogen activator receptor is stably elevated during the first week of treatment in the intensive care unit and predicts mortality in critically ill patients. Crit Care 2011;15:R63. [15] Koch A, Sanson E, Voigt S, et al. Serum adiponectin upon admission to the intensive care unit may predict mortality in critically ill patients. J Crit Care 2011;26:166–74. [16] Koch A, Voigt S, Sanson E, et al. Prognostic value of circulating aminoterminal pro-C-type natriuretic peptide in critically ill patients. Crit Care 2011;15:R45. [17] Koch A, Sanson E, Helm A, et al. Regulation and prognostic relevance of serum ghrelin concentrations in critical illness and sepsis. Crit Care 2010;14:R94. [18] Backes Y, van der Sluijs KF, Mackie DP, et al. Usefulness of suPAR as a biological marker in patients with systemic inflammation or infection: a systematic review. Intensive Care Med 2012;38:1418–28. [19] Leiper J, Nandi M, Torondel B, et al. Disruption of methylarginine metabolism impairs vascular homeostasis. Nat Med 2007;13:198–203. [20] Ronden RA, Houben AJ, Teerlink T, et al. Reduced renal plasma clearance does not explain increased plasma asymmetric dimethylarginine in hypertensive subjects with mild to moderate renal insufficiency. Am J Physiol Renal Physiol 2012;303: F149–56. [21] Sharma V, Ten Have GA, Ytrebo L, et al. Nitric oxide and L-arginine metabolism in a devascularized porcine model of acute liver failure. Am J Physiol Gastrointest Liver Physiol 2012;303:G435–41. [22] Yeni-Komshian H, Carantoni M, Abbasi F, Reaven GM. Relationship between several surrogate estimates of insulin resistance and quantification of insulinmediated glucose disposal in 490 healthy nondiabetic volunteers. Diabetes Care 2000;23:171–5. [23] Nichol A, Bailey M, Egi M, et al. Dynamic lactate indices as predictors of outcome in critically ill patients. Crit Care 2011;15:R242. [24] Ogawa T, Kimoto M, Sasaoka K. Purification and properties of a new enzyme, NG, NG-dimethylarginine dimethylaminohydrolase, from rat kidney. J Biol Chem 1989;264:10205–9. [25] Boger RH, Bode-Boger SM, Szuba A, et al. Asymmetric dimethylarginine (ADMA): a novel risk factor for endothelial dysfunction: its role in hypercholesterolemia. Circulation 1998;98:1842–7. [26] Leiper J, Nandi M. The therapeutic potential of targeting endogenous inhibitors of nitric oxide synthesis. Nat Rev Drug Discov 2011;10:277–91.
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
Contents lists available at SciVerse ScienceDirect
Journal of Critical Care journal homepage: www.jccjournal.org.
Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients☆ Alexander Koch a, Ralf Weiskirchen b, Julian Kunze a, Hanna Dückers a, Jan Bruensing a, Lukas Buendgens a, Michael Matthes a, Tom Luedde a, Christian Trautwein a, Frank Tacke a,⁎ a b
Department of Medicine III, RWTH-University Hospital Aachen, Pauwelsstrasse 30, Aachen, Germany Institute of Clinical Chemistry and Pathobiochemistry, RWTH-University Hospital Aachen, Pauwelsstrasse 30, Aachen, Germany
a r t i c l e Keywords: ADMA ICU Prognosis Sepsis Organ failure
i n f o
a b s t r a c t Objective: Serum concentrations of asymmetric dimethylarginine (ADMA), an endogenous inhibitor of nitric oxide synthase, may contribute to endothelial dysfunction and organ failure in sepsis. We aimed at investigating ADMA levels as a potential diagnostic or prognostic biomarker in critically ill patients. Methods: Two hundred fifty-five patients (164 with sepsis, 91 without sepsis) were studied prospectively upon admission to the medical intensive care unit (ICU) and on day 7, in comparison to 78 healthy controls. ADMA serum concentrations were correlated with clinical data and extensive laboratory parameters. Patients’ survival was followed up for up to 3 years. Results: ADMA serum levels were significantly elevated in critically ill patients at admission compared to controls. ADMA levels did not differ between patients with or without sepsis, but were closely related to hepatic and renal dysfunction, metabolism and clinical scores of disease severity. ADMA levels further increased during the first week of ICU treatment. ADMA serum levels at admission were an independent prognostic biomarker in critically ill patients not only for short-term mortality at the ICU, but also for unfavorable long-term survival. Conclusion: Serum ADMA concentrations are significantly elevated in critically ill patients, associated with organ failure and related to short- and long-term mortality risk. © 2013 Elsevier Inc. All rights reserved.
1. Introduction Systemic inflammation and sepsis, associated with multiple organ dysfunctions, are key characteristics of patients with critical illness requiring treatment at the intensive care unit (ICU). Nonspecific clinical and physiologic criteria of systemic inflammatory response syndrome and sepsis are used to identify patients at immediate need for ICU therapy. A complex network of biological mediators is dysregulated in critically ill patients, and several circulating parameters have been evaluated as potential biomarkers with diagnostic or even prognostic value [1]. There is increasing experimental evidence that the arginine-nitric oxide (NO) pathway is crucially involved in inflammation, infection and organ injury [2]. NO is a potent vasodilatator, which is produced from L-arginine by the enzyme nitric oxide synthase [3]. Asymmetric dimethylarginine (ADMA), on the other hand, is a naturally occurring non-selective inhibitor of NO synthase [4]. Circulating ADMA is mainly
☆ Competing interests: None of the authors declares competing interests. ⁎ Corresponding author. Department of Medicine III, RWTH-University Hospital Aachen, Pauwelsstraße 30, 52074 Aachen, Germany. Tel.: +49 241 80 35848; fax: +49 241 80 82455. E-mail address: [email protected] (F. Tacke).
derived from protein catabolism, because dimethylarginines are released as proteins are hydrolyzed, thus representing an obligatory product of protein turnover [5]. Functionally, ADMA impairs (beneficial) NO-dependent endothelial functions such as vascular dilatation or anti-inflammatory processes [4]. In line, circulating ADMA levels have been linked to endothelial dysfunction in cardiovascular diseases and risk factors for atherosclerosis [6]. It had been speculated that ADMA is also involved in the pathogenesis of microvascular dysfunction in sepsis, as increased ADMA could possibly decrease NO bioavailability at the endothelium [7]. Few small studies have investigated circulating ADMA as a novel biomarker in critical care medicine and found elevated ADMA serum concentrations in patients with acute infections and sepsis [7–11]. Although the exact mechanisms of ADMA regulation in these patients remained obscure, ADMA was suggested as a predictor of ICU mortality [7,8,12]. However, these initial studies were limited by their focus on patients with sepsis, by their relatively small cohorts and by the short observation period with respect to mortality as an endpoint. The aim of this study was to investigate ADMA serum concentrations in a large cohort of 255 consecutively enrolled critically ill patients, with or without sepsis, identify associations between ADMA and organ dysfunction, metabolism, and disease severity as well as to assess the prognostic
0883-9441/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcrc.2013.05.016
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
2
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
value of ADMA to predict ICU and long-term mortality in critically ill patients. 2. Materials and methods 2.1. Study design and patient characteristics Written informed consent was obtained from the patient, his or her spouse or the appointed legal guardian. Patients who were expected to have a short-term (b72 h) intensive care treatment due to post-interventional observation or acute intoxication were not included in this study [13]. All patient data, clinical information and blood samples were collected prospectively. The clinical course of patients was observed in a follow-up period by directly contacting the patients, the patients’ relatives or their primary care physician. Patients who met the criteria proposed by the American College of Chest Physicians & the Society of Critical Care Medicine Consensus Conference Committee for severe sepsis and septic shock were categorized as sepsis patients, the others as nonsepsis patients [14]. The study protocol was approved by the local ethics committee and conducted in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki (ethics committee of the University Hospital Aachen, RWTH-University, Aachen, Germany, reference number EK 150/06). 2.2. ADMA measurements Blood samples were collected upon admission to the ICU as well as in the morning of day 7 after admission. Importantly, ADMA levels at admission were obtained prior to therapeutic inventions at the ICU, which could potentially influence glucose metabolism, such as parenteral nutrition or insulin administration. After centrifugation at 4°C for 10 minutes, serum and plasma aliquots of 1 mL were frozen immediately at − 80°C. ADMA serum concentrations were analyzed using a commercial enzyme immunoassay (Immundiagnostik, Bensheim, Germany). The scientist performing experimental measurements was fully blinded to any clinical or other laboratory data of the patients or controls. 2.3. Statistical analysis Data are given as median and range due to the skewed distribution of most of the parameters. Differences between two groups were assessed by Mann-Whitney U test and multiple comparisons between more than two groups have been conducted by Kruskal-Wallis analysis of variance and Mann-Whitney U test for post hoc analysis. Box plot graphics illustrate comparisons between subgroups and they display a statistical summary of the median, quartiles, range, and extreme values. The whiskers extend from the minimum to the maximum value excluding outside and far out values which are displayed as separate points. An outside value (indicated by an open circle) was defined as a value that is smaller than the lower quartile minus 1.5 times interquartile range, or larger than the upper quartile plus 1.5 times the interquartile range. A far out value was defined as a value that is smaller than the lower quartile minus three times interquartile range, or larger than the upper quartile plus three times the interquartile range [15]. All values, including “outliers”, have been included for statistical analyses. Correlations between variables have been analyzed using the Spearman correlation tests, where values of P b .05 were considered statistically significant. All single parameters that correlated significantly with ADMA levels at admission were included in a multivariate linear regression analysis with ADMA as the dependent variable to identify independent (meaningful) predictors of elevated ADMA. The prognostic value of the variables was tested by univariate and multivariate analysis in the Cox regression model. Kaplan Meier curves were plotted to display the impact on survival
[16]. Receiver operating characteristic (ROC) curve analysis and the derived area under the curve (AUC) statistic provide a global and standardized appreciation of the accuracy of a marker or a composite score for predicting an event. ROC curves were generated by plotting sensitivity against 1-specificity [17]. All statistical analyses were performed with SPSS (SPSS, Chicago, IL).
3. Results 3.1. ADMA serum levels are significantly elevated in critically ill patients, but not related to sepsis In order to investigate ADMA in critical illness, we measured ADMA serum concentrations in a large cohort of medical ICU patients at admission (before therapeutic intervention) and on day 7 (Table 1). We enrolled 255 patients (149 male, 106 female with a median age of 63 years; range 18-90 years) who were admitted to the General Internal Medicine ICU at the RWTH-University Hospital Aachen, Germany (Table 1). As a control population we analyzed 78 healthy blood donors (50 male, 28 female; median age 30, range 18-67 years) with normal values for blood counts, C-reactive protein and liver enzymes. ADMA serum levels were significantly higher in ICU patients (n = 255, median 0.48 μmol/L, range 0.14-2.0) as compared with healthy controls (n = 78, median 0.36 μmol/L, range 0.23-0.57, P b .001; Fig.1 A). No association between ADMA levels and sex or age were observed in controls (data not shown). Among the 255 critically ill patients enrolled in this study, 164 patients conformed to the criteria of bacterial sepsis (Table 1). Pneumonia was identified in the majority of sepsis patients as origin of infection (not shown). Non-sepsis patients were admitted to the ICU mainly due to cardiopulmonary diseases (myocardial infarction, pulmonary embolism, and acute decompensated heart failure), decompensated liver cirrhosis or other critical conditions and did not differ in age or sex from sepsis patients. Sepsis patients were more often in need of mechanical ventilation in longer terms as compared to the non-sepsis patients’ cohort (Table 2). In sepsis patients significantly higher levels of routinely used biomarkers of inflammation (ie, C-reactive protein, procalcitonin, white blood cell count) were found (data not shown). Both groups did not differ in Acute Physiology and Chronic Health Evaluation (APACHE) II-, Sequential Organ Failure Assessment (SOFA) and simplified acute physiology score-2 score, vasopressor demand, or laboratory parameters indicating liver or renal dysfunction (Tables 1-2, and data not shown). Among all critical care patients about 25% died at the ICU; during the follow-up period of up to 3 years, a total of 47% of the initial cohort had died (Table 2). By direct comparison between septic and non-septic patients, ADMA serum levels did not differ between patients with or without sepsis (Table 1, Fig. 1B). In 44 patients, paired blood samples were available for ADMA measurements at ICU admission and at day 7 of ICU treatment. Remarkably, individual ADMA levels increased during the first week Table 1 Baseline patient characteristics and ADMA serum measurements Parameter
All patients
Sepsis
Non-sepsis
Number Sex (male/female) Age median (range) [years] APACHE-II score median (range) SOFA score median (range) pre-existing diabetes n(%) ADMA day 1 median (range)[μmol/L] ADMA day 7 median (range) [μmol/L]
255 149/106 63 (18-90) 17 (2-43)
164 96/68 64 (20-90) 19 (3-43)
91 53/38 61 (18-85) 15 (2-33)
9.5 (0-19) 11 (3-19) 7 (0-16) 75 (29.4%) 46 (28%) 29 (31.9%) 0.48 (0.14-2.00) 0.47 (0.14-2.00) 0.49 (0.16-1.65) 0.71 (0.43-1.94) 0.70 (0.43-1.94) 0.78 (0.51-1.62)
For quantitative variables, median and range (in parenthesis) are given.
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
A 2.0
B
2.0
n.s.
P < .001 1.5
ADMA µmol/L
ADMA µmol/L
1.5
1.0
1.0
0.5
0.5
0
0
controls
patients
n = 78
n = 255
C 2.0
no n = 91
D
2.0
P < .001
sepsis
yes n = 164
ADMA day 1 ADMA day 7
1.5
1.5
ADMA µmol/L
ADMA µmol/L
3
1.0
P < .001 1.0
0.5
0.5
0
0
ADMA day 1 n = 44
P < .001
no n = 11
ADMA day 7 n = 44
sepsis
yes n = 33
Fig. 1. Serum ADMA concentrations in critically ill patients. (A) At admission to the Medical ICU, serum ADMA levels were significantly (P b .001, U test) elevated in critically ill patients (n = 255) as compared to healthy controls (n = 78). (B) ADMA serum levels at ICU admission did not differ between patients with or without sepsis. (C-D) In 44 patients, ADMA levels were measured at admission (day 1) and at 1 week (day 7) of ICU therapy. ADMA levels increased significantly (P b .001, paired Wilcoxon-test) during treatment at the ICU (C), independent of sepsis (D).
of ICU therapy (Table 1, Fig. 1C; P b .001, paired Wilcoxon-Test). This was observed for patients with sepsis as well as for non-sepsis patients (Fig. 1D).
3.2. ADMA serum concentrations at admission to the ICU are closely correlated to organ function, metabolism and clinical scores To determine the factors possibly promoting elevated serum ADMA levels in critically ill patients, correlation analyses with laboratory parameters were performed. For these analyses, serum ADMA levels at admission were applied, in order to exclude possible confounding effects due to patients that died or were discharged from the ICU during the first week. At admission to the ICU, serum ADMA concentrations were closely correlated to biomarkers displaying hepatic and renal dysfunction. In detail, ADMA was found to correlate significantly with markers reflecting the hepatic biosynthetic capacity (eg, international normalized ratio [INR], r = 0.272, P b .001; albumin, r =−0.285, P b .001) and also to parameters indicating cholestasis (eg, bilirubin, r = 0.361, P b .001; gamma-GT, r = 0.357, P b .001) as well as with markers indicating renal dysfunction (eg, creatinine, r = 0.205, P = .001; cystatin C, r = 0.334, P b .001). In line, patients with preexisting renal failure, defined as a glomerular filtration rate for cystatin C b 50 mL/min, or with preexisting hepatic dysfunction,
defined as prothrombin time b50%, showed significantly elevated ADMA levels (Fig. 2A and B, P b .001, U test). Besides its relation to cardiovascular risk, ADMA levels had been linked to metabolic syndrome and type 2 diabetes [4]. In our cohort of critically ill patients, ADMA levels were significantly decreased in ICU patients with a pre-existing type 2 diabetes upon admission, but did not differ between patients with obesity, defined as a body mass index N30 kg/m 2 (detailed data not shown). Interestingly, ADMA was also correlated to established clinical scores indicating disease severity, such as the APACHE II (r = 0.154, P = .032) or SOFA score (r = 0.224, P = .021), as well as to soluble urokinase plasminogen activator receptor (suPAR, r = 0.434, P b .001), a prognostic biomarker in ICU patients [18]. Table 2 Outcome measures of the study cohort Parameter
All patients
Sepsis
Non-sepsis
Number ICU days median (range) Death during ICU n (%) Death during follow-up, n (%) Mechanical ventilation, n (%) Ventilation time median (range) [h]
255 8 (1-137) 63 (24.7%) 120 (47.1%) 186 (72.9%) 122 (0-3628)
164 10 (1-137) 47 (28.7%) 88 (53.7%) 120 (73.2%) 183 (0-2966)
91 6 (1-45) 16 (17.6%) 32 (35.2%) 61 (67%) 46 (0-3628)
For quantitative variables, median and range (in parenthesis) are given.
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
4
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
A
B 2.0
2.0
P < .001
P < .001 1.5
ADMA µmol/L
ADMA µmol/L
1.5
1.0
1.0
0.5
0.5
0
0 GFR >50 ml/min
GFR <50 ml/min
GFR cystatin C day 1
PT >50%
PT <50%
prothrombin time day 1
Fig. 2. Serum ADMA concentrations in critically ill patients with organ failure and comorbidities. Serum ADMA levels were measured in n = 255 critically ill patients at admission to the ICU. Patients with renal failure (cystatin C-based glomerular filtration rate [GFR] b50 mL/min, A) or hepatic failure (prothrombin time b50%, B) had significantly elevated ADMA levels.
When all parameters that were correlated with ADMA serum levels by univariate analysis were included in a multivariate regression analysis, only albumin (P = .040) and suPAR (P b .001), but not renal function markers or metabolic parameters, remained independent predictors of ADMA concentrations (Table 3).
3.3. ADMA serum levels are an independent prognostic biomarker for ICU mortality in critically ill patients Based on the close correlation between ADMA levels at admission and disease severity scores, we hypothesized that circulating ADMA might be capable of identifying patients at high risk for mortality. Indeed, patients that died during the course of ICU treatment (about one quarter of the total cohort) had significantly higher serum ADMA levels at admission compared with the ICU survivors (median 0.62 vs. 0.44 μmol/L, P b .001; Fig. 3A). We thus performed Cox regression analyses and Kaplan-Meier curves to assess the impact of the initial ADMA serum concentrations on ICU-mortality among critically ill patients. Low ADMA levels upon admission to the ICU were a strong prognostic predictor for ICU-survival (P b .001, Cox regression analyses). In multivariate Cox regression analyses, including markers of inflammation/infection (CRP, WBC), hematologic (hematocrit, platelet count), circulatory (lactate), hepatic (bilirubin, INR) and renal (creatinine, urea) deterioration at admission, ADMA remained an independent significant prognostic parameter (hazard ratios and p-values are presented in Table 4). Kaplan-Meier curves displayed that patients with ADMA levels of the upper quartile (N0.63 μmol/L) had highest mortality (Fig. 3 B). We found the best cut-off value to discriminate survivors from non-ICU-survivors for serum ADMA of 0.65 μmol/L (Fig. 3C). Table 3 Multivariate regression analysis of parameters predicting ADMA levels Parameter
Standardized coefficient beta
t value
P
suPAR Albumin Creatinine Bilirubin INR APACHE-II
0.479 −0.202 0.055 0.011 −0.021 −0.121
4.220 −2.079 0.598 0.110 −0.222 −1.233
b.001 0.040 n.s. n.s. n.s. n.s.
suPAR, soluble urokinase plasminogen activator receptor.
By ROC curve analyses, we compared the prognostic value of ADMA to other clinically used biomarkers. In comparison to biomarkers reflecting hepatic or renal organ dysfunction (INR, creatinine) or inflammation (CRP), ADMA levels displayed higher prognostic power (Fig. 3D). The prognostic value of serum ADMA levels at admission (AUC 0.636, 95% CI 0.513-0.760) was comparable to clinical disease severity scores such as APACHE-II (AUC 0.662, 95% CI 0.533-0.791) or SOFA (AUC 0.667, 95% CI 0.549-0.785). Nevertheless, the use of the individual increase in ADMA between admission to the ICU and day 7 of ICU treatment, as available for 44 patients with matched measurements, did not further increase the prognostic value of ADMA (AUC 0.583). 3.4. High ADMA levels at ICU admission indicate unfavorable long-term prognosis During the follow-up observation period of approximately three years, the overall case fatality rate increased to 47.1% of the study cohort (Table 1). ADMA serum concentrations at admission to the ICU were significantly higher in patients with unfavorable outcome (median 0.54 vs. 0.43 μmol/L, P b .001, Fig. 4A). By Cox regression analysis, initial serum ADMA levels significantly predicted long-term prognosis (P b .001). The prognostic value remained significant also by multivariate analysis (Table 5). Kaplan-Meier curves proved that ADMA levels of the highest quartile (N0.63 μmol/L) were strongly associated with fatal outcome (Fig. 4B). The log-rank test value of the Kaplan-Meier curves for ADMA quartiles was substantially higher for the overall (32.88) than for the ICU mortality (16.29). Again, ADMA levels of 0.65 μmol/L discriminated the long-term prognosis of critically ill patients (Fig. 4C). In accordance with the findings for ICU mortality, serum ADMA levels at ICU admission had higher prognostic value for long-term outcome as compared with single biomarkers (Fig. 4D). Interestingly, the prognostic accuracy of serum ADMA levels could be even improved, if ADMA levels were adjusted to total protein concentrations of the patients. Based on the close association between liver dysfunction and ADMA levels and the highly extended mortality risk of critically ill patients with liver failure, we reasoned that normalization of ADMA to protein levels might allow to better delineate the specific prognostic value of ADMA. For ICU mortality, the area-under-the-ROC-curve (AUC)
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
A
B
ICU survival
5
ICU survival 1.0
2.0
ADMA lower 25% ADMA 25-75% ADMA upper 25%
P < .001 cumulative survival
0.8
ADMA µmol/L
1.5
1.0
0.5
0.6
0.4
0.2
log rank 32.88
survivors
deaths
D
40
60
80
100
120
140
1.0
prediction of ICU survival
ADMA < 0.65 µmol/L ADMA > 0.65 µmol/L 0.8
sensitivity
0.8
0.6
0.4
0.2
0
20
time (days)
ICU survival 1.0
cumulative survival
0
n = 63
n = 192
C
P < .001
0
0
0.6
0.4
ADMA INR creatinine CRP
0.2
log rank 8.52 P = .0035 0 0
20
40
60
80
100
120
140
0
time (days)
0.2
0.4
0.6
0.8
1.0
1 - specificity
Fig. 3. Prediction of ICU mortality by ADMA serum concentrations. (A) Patients that died during the course of ICU treatment had significantly higher serum ADMA levels on admittance to ICU (P b .001 than survivors. (B-C) Kaplan-Meier survival curves of ICU patients are displayed, showing that patients with ADMA levels of upper quartile (on admission N0.63 μmol/L; black, B) had an increased short-term mortality at the ICU as compared to patients with ADMA serum concentrations of lower quartile (on admission b0.36 μmol/L; light grey) or middle 50% (grey). Best discrimination between ICU survivors and non-survivors was achieved with an ADMA cut-off value of 0.65 μmol/L (C). P values are given in the figure. (D) ROC curve analyses comparing the prognostic value of ADMA at admission for ICU survival (dark grey, AUC 0.686, 95% CI 0.606-0.765) with INR (light grey, AUC 0.626, 95% CI 0.547-0.706) and creatinine (black, AUC 0.602, 95% CI 0.522-0.683) as makers of hepatic and renal function as well as CRP (dotted line, AUC 0.521, 95% CI 0.442-0.601) as a marker of inflammation and infection.
increased from 0.713 for ADMA itself to 0.765 for ADMA/proteinratio. For overall mortality, the AUC increased from 0.659 for ADMA to 0.709 for ADMA/protein-ratio, respectively. Again, the use of the individual increase in ADMA between admission to the ICU and day
7 of ICU treatment, as available for 44 patients with matched measurements, did not further increase the prognostic value of ADMA (AUC 0.636). 4. Discussion
Table 4 Uni- and multivariate Cox regression analyses for ADMA levels at admission to predict ICU mortality Parameter
Unadjusted HR (95% CI)
P
Adjusted HR (95% CI)
P
ADMA Creatinine Urea Bilirubin INR White blood cell count C-reactive protein Platelets Lactate
3.803 (2.189-6.607) 1.031 (0.951-1.118) 1.001 (1.000-1.003) 1.051 (1.014-1.089) 1.309 (1,037-1.652) 1.015 (1.003-1.028)
b0.001 NS NS 0.007 0.023 0.039
3.141 (1.721-5.730) -
b0.001 NS NS NS NS NS
0.999 (0.996-1.002) 0.999 (0.997-1.001) 1.127 (1.062-1.197)
NS NS b0.001
1.095 (1.030-1.163)
NS NS 0.003
Significant results are highlighted in bold.
Based on the increasing experimental evidence on the importance of the arginine-nitric oxide pathway for inflammation, infection and organ injury, ADMA as the most relevant inhibitor of nitric oxide synthase has been hypothesized as a novel biomarker in critical care medicine [3]. Initial studies have linked ADMA levels in critically ill and/or septic patients with the severity of organ failure, degree of inflammation and the presence of shock in severe sepsis [7,8,12]. In line with these findings from smaller studies, we observed significantly upregulated circulating ADMA levels in 255 prospectively enrolled critically ill patients compared to healthy controls. Unexpectedly, however, ADMA levels did not differ between patients with or without sepsis. An association between ADMA levels and microvascular dysfunction has been recently documented in selected patients with sepsis [7]. As accumulation of ADMA and consequently
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
6
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
A
B
overall survival 2.0
overall survival 1.0
ADMA lower 25% ADMA 25-75% ADMA upper 25%
P < .001 0.8
cumulative survival
ADMA µmol/L
1.5
1.0
0.5
0.6
0.4
0.2
log rank 32.88 0
C
survivors
deaths
n = 131
n = 120
200
400
600
800
1000
time (days)
D 1.0
prediction of overall survival
ADMA < 0.65 µmol/L ADMA > 0.65 µgmol/L 0.8
0.8
sensitivity
cumulative survival
0
overall survival 1.0
0.6
0.4
0.2
0
P < .001
0
0.6
0.4
ADMA INR creatinine CRP
0.2
log rank 13.96 P = .0002 0 0
200
400
600
800
1000
0
0.2
0.4
0.6
0.8
1.0
1 - specificity
time (days)
Fig. 4. Prediction of overall mortality by ADMA serum concentrations. (A) Patients that died during the total observation period had significantly higher serum ADMA levels on admittance to ICU (P b .001 than survivors. (B-C) Kaplan-Meier survival curves of ICU patients are displayed, showing that patients with ADMA levels of upper quartile (on admission N0.63 μmol/L; black, B) had an increased short-term mortality at the ICU as compared to patients with ADMA serum concentrations of lower quartile (on admission b0.36 μmol/L; light grey) or middle 50% (grey). Best discrimination between ICU survivors and non-survivors was achieved with an ADMA cut-off value of 0.65 μmol/L (C). P values are given in the figure. (D) ROC curve analyses comparing the prognostic value of ADMA at admission for ICU survival (dark grey, AUC 0.636, 95% CI 0.566-0.707) with INR (light grey, AUC 0.568, 95% CI 0.496-0.640) and creatinine (black, AUC 0.568, 95% CI 0.495-0.641) as makers of hepatic and renal function as well as CRP (dotted line, AUC 0.572, 95% CI 0.500-0.644) as a marker of inflammation and infection.
reduced NO signaling is known to promote endothelial dysfunction and increased systemic vascular resistance [19], our findings indicate that ADMA up-regulation may represent a uniform response in critical illness, independent of the presence of infection. On the other hand, a reduced renal and/or hepatic clearance might contribute to elevated ADMA levels in critically ill patients, because we Table 5 Uni- and multivariate Cox regression analyses for ADMA levels at admission to predict overall mortality Parameter
Unadjusted HR (95% CI)
P
Adjusted HR (95% CI)
P
ADMA Creatinine Urea Bilirubin INR White blood cell count C-reactive protein Platelets Lactate
3.300 1.014 1.001 1.046 1.010 1.009
(2.082-5.232) (0.967-1.062) (1.000-1.002) (1.023-1.070) (0.996-1.024) (1.001-1.017)
b0.001 NS 0.008 b0.001 0.171 0.034
3.115 (1.862-5.210) 1.002 (1.001-1.003) -
b0.001 0.014 0.001 NS NS NS
1.001 (1.000-1.003) 0.999 (0.998-1.000) 1.075 (1.039-1.111)
NS NS b0.001
1.118 (1.047-1.194)
NS NS 0.001
observed that ADMA serum concentrations correlated closely with hepatic and renal organ function. However, there is no clear experimental evidence at present for hepatic or renal clearance of ADMA. Studies from patients with chronic renal insufficiency and animals models of acute liver failure did not show altered ADMA clearance as a main mechanism of elevated ADMA serum levels [20,21]. From patients with chronic cardiovascular diseases, ADMA levels have been linked to metabolic syndrome and surrogate parameters of insulin resistance [4,22]. In non-diabetic patients, plasma glucose levels were positively correlated with circulating ADMA, and diabetic patients showed highly elevated ADMA levels [22]. This association indicated that ADMA might activate signaling pathways involved in development of the metabolic syndrome [4]. In our study, ADMA levels were not increased in patients with pre-existing type 2 diabetes. This discrepancy between prior studies on chronically diseased patients and our findings in ICU patients might indicate that the physiological regulation of ADMA is massively altered in the acute phase of critical illness, which is characterized by increased protein turnover due to catabolism and the induction of acute phase proteins [12]. Three studies including 47, 52, and 67 critically ill patients, respectively, even indicated that ADMA levels might serve as a
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016
A. Koch et al. / Journal of Critical Care xxx (2013) xxx–xxx
potential biomarker predicting ICU mortality [7,8,12]. However, these small studies focused on septic patients and solely assessed (shortterm) ICU mortality. On the basis of 255 consecutively included critically ill patients, we now identified ADMA as a biomarker related to short-term and long-term mortality risk. In comparison with routinely used single biomarkers reflecting organ dysfunction such as creatinine or INR, ADMA levels showed superior prognostic power, reaching a similar accuracy as composite clinical disease severity scores (eg, APACHE-II, SOFA). Importantly, multivariate Cox regression analyses confirmed circulating ADMA as an independent prognostic indicator, specifically independent of hepatic or renal failure, inflammation and even of lactate as a clinically used biomarker for disturbed tissue perfusion [23]. These analyses support that ADMA could be a marker for a combined algorithm for risk in ICU. However, the cut-off values for mortality prediction that have been identified in our study need to be validated in independent patient cohorts. The association of high ADMA levels with increased risk of case-related fatality, both short- and long-term, might be caused by the inhibition of endothelial nitric oxide synthesis and of vascular reactivity, as experimentally demonstrated for animals and humans [24,25]. Interestingly, the prognostic accuracy of serum ADMA levels could be even improved, if ADMA levels were adjusted to total protein concentrations of the patients. This supports that an ADMA increase in conditions of protein catabolism is a distinct indicator of disease severity. A recent study has demonstrated that in patients with severe septic shock, protein catabolism resulted in massively increased plasma arginine levels [8]. Possibly, the L-arginine-to-ADMA ratio might then be reduced, which would imply reduced endothelial nitric oxide availability and impaired microvascular reactivity in these patients [8], potentially linking ADMA levels and mortality risk. The association of ADMA levels with mortality risk raises the question if this reflects a potentially modifiable process in critical illness. In fact, ADMA and its related pathways have been suggested as possible new therapeutic targets in several disease conditions [26]. With respect to sepsis, it is important to consider that ADMA is physiologically eliminated through active metabolism by dimethylarginine dimethylaminohydrolase (DDAH) [3]. Experimental data from DDAH-deficient mice and administration of a chemical DDAH inhibitor had indicated that pharmacological inhibition of DDAH could be harnessed therapeutically to increase the vascular tension during the hyperdynamic state of sepsis [19]. However, our study revealed strongly increased ADMA levels already at the time-point of admission to the ICU before therapeutic interventions in septic and non-septic patients and clearly demonstrated high ADMA levels as an independent prognostic indicator for adverse outcome. Thus, blocking DDAH and consequently further increasing circulating ADMA could also be potentially harmful in critical illness. Therefore, further experimental and clinical studies are clearly warranted to understand the underlying regulatory mechanisms and pathogenic consequences of elevated ADMA serum concentrations in the setting of intensive care medicine. Acknowledgments We sincerely thank Philip Kim for excellent technical assistance. This work was supported by the German Research Foundation (DFG Ta434/ 2-1 & SFB/TRR57) and the Interdisciplinary Centre for Clinical Research (IZKF) within the faculty of Medicine at the RWTH Aachen University.
7
References [1] Pierrakos C, Vincent JL. Sepsis biomarkers: a review. Crit Care 2010;14:R15. [2] Landry DW, Oliver JA. The pathogenesis of vasodilatory shock. N Engl J Med 2001;345:588–95. [3] Boger RH. Live and let die: asymmetric dimethylarginine and septic shock. Crit Care 2006;10:169. [4] Bestermann Jr WH. The ADMA-metformin hypothesis: linking the cardiovascular consequences of the metabolic syndrome and type 2 diabetes. Cardiorenal Med 2011;1:211–9. [5] Leiper J, Vallance P. Biological significance of endogenous methylarginines that inhibit nitric oxide synthases. Cardiovasc Res 1999;43:542–8. [6] Boger RH. Asymmetric dimethylarginine (ADMA): a novel risk marker in cardiovascular medicine and beyond. Ann Med 2006;38:126–36. [7] Davis JS, Darcy CJ, Yeo TW, et al. Asymmetric dimethylarginine, endothelial nitric oxide bioavailability and mortality in sepsis. PLoS One 2011;6:e17260. [8] O'Dwyer MJ, Dempsey F, Crowley V, et al. Septic shock is correlated with asymmetrical dimethyl arginine levels, which may be influenced by a polymorphism in the dimethylarginine dimethylaminohydrolase II gene: a prospective observational study. Crit Care 2006;10:R139. [9] Iapichino G, Umbrello M, Albicini M, et al. Time course of endogenous nitric oxide inhibitors in severe sepsis in humans. Minerva Anestesiol 2010;76: 325–33. [10] Nakamura T, Sato E, Fujiwara N, et al. Circulating levels of advanced glycation end products (AGE) and interleukin-6 (IL-6) are independent determinants of serum asymmetric dimethylarginine (ADMA) levels in patients with septic shock. Pharmacol Res 2009;60:515–8. [11] Zoccali C, Maas R, Cutrupi S, et al. Asymmetric dimethyl-arginine (ADMA) response to inflammation in acute infections. Nephrol Dial Transplant 2007;22: 801–6. [12] Nijveldt RJ, Teerlink T, Van Der Hoven B, et al. Asymmetrical dimethylarginine (ADMA) in critically ill patients: high plasma ADMA concentration is an independent risk factor of ICU mortality. Clin Nutr 2003;22:23–30. [13] Koch A, Gressner OA, Sanson E, Tacke F, Trautwein C. Serum resistin levels in critically ill patients are associated with inflammation, organ dysfunction and metabolism and may predict survival of non-septic patients. Crit Care 2009;13: R95. [14] Koch A, Voigt S, Kruschinski C, et al. Circulating soluble urokinase plasminogen activator receptor is stably elevated during the first week of treatment in the intensive care unit and predicts mortality in critically ill patients. Crit Care 2011;15:R63. [15] Koch A, Sanson E, Voigt S, et al. Serum adiponectin upon admission to the intensive care unit may predict mortality in critically ill patients. J Crit Care 2011;26:166–74. [16] Koch A, Voigt S, Sanson E, et al. Prognostic value of circulating aminoterminal pro-C-type natriuretic peptide in critically ill patients. Crit Care 2011;15:R45. [17] Koch A, Sanson E, Helm A, et al. Regulation and prognostic relevance of serum ghrelin concentrations in critical illness and sepsis. Crit Care 2010;14:R94. [18] Backes Y, van der Sluijs KF, Mackie DP, et al. Usefulness of suPAR as a biological marker in patients with systemic inflammation or infection: a systematic review. Intensive Care Med 2012;38:1418–28. [19] Leiper J, Nandi M, Torondel B, et al. Disruption of methylarginine metabolism impairs vascular homeostasis. Nat Med 2007;13:198–203. [20] Ronden RA, Houben AJ, Teerlink T, et al. Reduced renal plasma clearance does not explain increased plasma asymmetric dimethylarginine in hypertensive subjects with mild to moderate renal insufficiency. Am J Physiol Renal Physiol 2012;303: F149–56. [21] Sharma V, Ten Have GA, Ytrebo L, et al. Nitric oxide and L-arginine metabolism in a devascularized porcine model of acute liver failure. Am J Physiol Gastrointest Liver Physiol 2012;303:G435–41. [22] Yeni-Komshian H, Carantoni M, Abbasi F, Reaven GM. Relationship between several surrogate estimates of insulin resistance and quantification of insulinmediated glucose disposal in 490 healthy nondiabetic volunteers. Diabetes Care 2000;23:171–5. [23] Nichol A, Bailey M, Egi M, et al. Dynamic lactate indices as predictors of outcome in critically ill patients. Crit Care 2011;15:R242. [24] Ogawa T, Kimoto M, Sasaoka K. Purification and properties of a new enzyme, NG, NG-dimethylarginine dimethylaminohydrolase, from rat kidney. J Biol Chem 1989;264:10205–9. [25] Boger RH, Bode-Boger SM, Szuba A, et al. Asymmetric dimethylarginine (ADMA): a novel risk factor for endothelial dysfunction: its role in hypercholesterolemia. Circulation 1998;98:1842–7. [26] Leiper J, Nandi M. The therapeutic potential of targeting endogenous inhibitors of nitric oxide synthesis. Nat Rev Drug Discov 2011;10:277–91.
Please cite this article as: Koch A, et al, Elevated asymmetric dimethylarginine levels predict short- and long-term mortality risk in critically ill patients, J Crit Care (2013), http://dx.doi.org/10.1016/j.jcrc.2013.05.016