Journal of Clinical Neuroscience xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Clinical Neuroscience

journal homepage: www.elsevier .com/ locate/ jocn

Clinical Study

Increased cerebrospinal fluid concentrations of asymmetricdimethylarginine correlate with adverse clinical outcome insubarachnoid hemorrhage patients

http://dx.doi.org/10.1016/j.jocn.2013.11.0380967-5868/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel./fax: +86 25 8086 3310.E-mail address: hang_neurosurgery@163.com (C.-H. Hang).

1 The first two authors contributed equally to this work.

Please cite this article in press as: Li H et al. Increased cerebrospinal fluid concentrations of asymmetric dimethylarginine correlate with adverseoutcome in subarachnoid hemorrhage patients. J Clin Neurosci (2014), http://dx.doi.org/10.1016/j.jocn.2013.11.038

Hua Li a,1, Wei Wu a,1, Ming Liu b, Xin Zhang a, Qing-Rong Zhang a, Li Ni c, Chun-Hua Hang a,⇑a Department of Neurosurgery, Jinling Hospital, School of Medicine, Nanjing University, 305 East Zhongshan Road, Nanjing 210002, Jiangsu Province, Chinab Department of Neurosurgery, School of Medicine, Southern Medical University (Guangzhou), Jinling Hospital, Nanjing, Jiangsu Province, Chinac Department of Pharmacology, Jinling Hospital, Nanjing 210002, Jiangsu Province, China

a r t i c l e i n f o

Article history:Received 14 August 2013Accepted 13 November 2013Available online xxxx

Keywords:Asymmetric dimethylarginineHPLCOutcomeSubarachnoid hemorrhageVasospasm

a b s t r a c t

Elevated cerebrospinal fluid (CSF) concentrations of asymmetric dimethylarginine (ADMA), an endoge-nous inhibitor of nitric oxide synthase, have been found in patients with subarachnoid hemorrhage(SAH). In addition, CSF levels of ADMA are associated with the severity of vasospasm. However, the rela-tion between CSF ADMA levels and the clinical outcome of SAH patients is still unclear. We hypothesizedthat elevated ADMA levels in CSF might be related to the clinical outcome of SAH patients. CSF ADMAlevels were measured in 20 SAH patients at days 3–5, days 7–9 and days 12–14 after SAH onset usinghigh-performance liquid chromatography. Cerebral vasospasm was assessed by transcranial Dopplerultra sonography. Clinical outcome at 2 year follow-up was evaluated using the Karnofsky PerformanceStatus scale (KPS). CSF ADMA concentrations in all SAH patients were significantly increased at days3–5 (p = 0.002) after SAH, peaked on days 7–9 (p < 0.001) and remained elevated until days 12–14(p < 0.001). In subgroup analysis, significant increases of CSF ADMA levels were found in patients bothwith and without vasospasm. The KPS scores significantly correlated with CSF levels of ADMA at days7–9 (correlation coefficient = �0.55, p = 0.012; 95% confidence interval �0.80 to �0.14). Binary logisticregression analysis indicated that higher ADMA level at days 7–9 predicted a poor clinical outcome at2 year follow-up after SAH (odds ratio = 1.722, p = 0.039, 95% confidence interval 1.029 to 2.882). ADMAmay be directly involved in the pathological process and future adverse prognosis of SAH.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Subarachnoid hemorrhage (SAH), especially aneurysmal SAH, isa life-threatening disease of the central nervous system. The inci-dence of SAH is about 22.5 cases per 100,000 people according tothe World Health Organization [1]. Between 4 to 9 days afterSAH, severe cerebral vasospasm develops in 30–70% of patients,resulting in delayed ischemic neurological deficits (DIND) in 25%of patients with vasospasm. Half of these patients suffer severeneurological dysfunction or death due to DIND [2]. However,despite substantial efforts, treatment strategies to reduce the inci-dence of vasospasm and to improve the prognosis of SAH patientsare limited.

Asymmetric dimethylarginine (ADMA) is an endogenous com-petitive inhibitor of nitric oxide synthase (NOS). Elevated concen-trations of ADMA decrease the availability of nitric oxide (NO)and thus induces vasoconstriction [3,4]. During the past decade,studies have demonstrated that ADMA is not only a marker ofendothelial dysfunction, but also an independent risk factor forcardiovascular and cerebrovascular diseases [5]. ADMA has alsobeen shown to predict outcomes of cardiovascular diseases andischemic stroke [6–10].

In SAH studies, cerebrospinal fluid (CSF) concentrations ofADMA have been found to be significantly increased after SAH,and elevated levels of ADMA are correlated with the course anddegree of vasospasm as well as with the decrease of NO afterSAH [11–13]. However, the relationship between CSF ADMA con-centrations and the clinical outcome of SAH patients is stillunknown. The present study was carried out to investigatewhether levels of CSF ADMA were associated with clinical outcomein patients with SAH.

clinical

2 H. Li et al. / Journal of Clinical Neuroscience xxx (2014) xxx–xxx

2. Material and methods

2.1. Patients

Between December 2008 and March 2009, 25 consecutivepatients diagnosed with SAH were prospectively recruited uponadmission to the Neurosurgery Unit of the Jinling Hospital, Nanjing,China. The diagnosis of SAH was supported in every patient by axialCT scan. All patients underwent detailed physical examination androutine blood chemistry analysis. Those who had been previouslydiagnosed with hypertension, hyperlipidemia, diabetes mellitus,arterial atherosclerosis, chronic renal failure or peripheral vasculardisease were excluded.

The control group consisted of 18 patients without a history ofcerebrovascular disease or contraindication for lumbar puncture,who underwent spinal anesthesia for elective urogenital tractintervention or lower limb fracture surgery. The study wasapproved by the local Ethics Committee, and all patients gave writ-ten informed consent.

2.2. Patient management

All patients received intravenous nimodipine at a dose of 2 mg/hour from admission until at least day 14 after SAH onset. The typeof surgical procedure (clipping or coiling) was decided according toboth size and location of the aneurysm by the treating neurosur-geon and neuroradiologist. After clipping or coiling, those patientswho had delayed ischemic neurological deficit or cerebral vaso-spasm were managed with ‘‘triple H’’ therapy (hypertension witha mean arterial pressure goal greater than 100 mmHg, hypervole-mia and hemodilution with a goal hematocrit of 30). Clinical onsetof vasospasm is defined below. Each vasospasm episode was trea-ted with cerebral intra-arterial administration of nimodipine. Thistherapy was repeated if necessary. Balloon angioplasty was used asa second-line therapy when nimodipine was judged insufficient.

2.3. Assessment of vasospasm

Transcranial Doppler ultra sonography (TCD) was performeddaily to assess the presence of vasospasm. In accordance with a2004 technology assessment by the American Academy of Neurol-ogy [14], the following three independent criteria were defined asmiddle cerebral artery vasospasm: (1) velocity greater than200 cm/s, (2) a rapid rise (greater than 50 cm/s) between serialTCD measurements, or (3) a Lindegaard Index greater than 6. TheLindegaard Index refers to the ratio of velocities between the middlecerebral artery and the ipsilateral extracranial internal carotidartery.

2.4. Sample collection and preparation

In the study group, lumbar puncture was performed on days 3–5,7–9 and 12–14 after SAH. In the control group, a single CSF samplewas collected during spinal anesthesia before surgery. In total,approximately 15 ml of CSF was collected in polypropylene vialswhich were immediately centrifuged at 3000 rotations/minute for15 minutes at 4 �C. Supernatants were stored at �80 �C untilanalysis.

2.5. Determination of CSF ADMA concentration

CSF concentrations of ADMA were determined simultaneouslyby high-performance liquid chromatography (HPLC) as previouslydescribed [15]. Briefly, HPLC was performed on a Shimadzu LC-10AD system equipped with a Shimadzu RF-10Axl fluorescence

Please cite this article in press as: Li H et al. Increased cerebrospinal fluid concoutcome in subarachnoid hemorrhage patients. J Clin Neurosci (2014), http://d

detector (both Shimadzu, Kyoto, Japan) for excitation at 338 nmand emission at 450 nm with a Hypersil ODS (C18) Column(4.6 mm � 150 mm) (Thermo Fisher Scientific, Waltham, MA,USA). All reagents and chemicals for the analysis of ADMA werefrom Sigma-Aldrich (St Louis, MO, USA).

2.6. Observational follow-up

Patients were clinically re-examined at 3 to 6 month intervalsafter hospital discharge. If the regular re-evaluations did not occur,personal telephone contact to the patients or their relatives wasestablished. Clinical conditions were recorded semiannually untilApril 2011 using the Karnofsky Performance Status scale (KPS)[16]. During the follow-up period, one patient died during the sec-ond month after discharge due to cardiorespiratory failure, but allothers were followed up for a period of 2 years. Outcome was adju-dicated by two independent observers blinded to the patients’other clinical and laboratory data.

Patients with the ability to complete activities of daily living(KPS P 60) were classified as having a relatively good outcome.Those who could not take care of themselves or those who neededto be admitted to care (KPS < 60) were classified as having a pooroutcome.

2.7. Statistical analysis

Data are presented as mean value ± standard deviation. Chi-squared test was used for categorical variables and unpaired Stu-dent’s t-test for continuous variables, to assess differences betweenthe groups. Pearson’s correlation coefficient (CC) was used to eval-uate a possible correlation in continuous variables. The influence ofADMA levels on vasospasm and clinical outcome was assessedusing binary logistic regression. Results are presented as oddsratios with the corresponding 95% confidence intervals (CI). Statis-tical significance was defined as p < 0.05. All analyses were per-formed with the Statistical Package for the Social Sciencessoftware, version 15 (SPSS, Chicago, IL, USA).

3. Results

3.1. Clinical characteristics and ADMA concentrations

Five patients were excluded from the study because clinical orfollow-up data were incomplete. The remaining 20 patients (sixmen and 14 women, age 52.35 ± 7.21 years) were included intothe analysis. Of these patients, 15, 20 and 17 CSF samples wereobtained on days 3–5, days 7–9 and days 12–14, respectively.Seven patients (35%) developed vasospasm, while the rest showedno evidence of vasospasm. Clinical characteristics and follow-updata are shown in Table 1.

A low level of ADMA was identified in the control group(2.06 ± 0.66 ug/l) while the level of ADMA increased significantlyat days 3–5 after SAH (4.48 ± 2.41 ug/l, p = 0.002), peaked on days7–9 (6.71 ± 3.25 ug/l, p < 0.001) and remained elevated until days12–14 (5.61 ± 2.19 ug/l, p < 0.001) (Fig. 1A).

3.2. CSF ADMA levels in relation to cerebral vasospasm

The CSF concentration of ADMA significantly increased at days3–5 (5.89 ± 3.32 ug/l, p = 0.061), days 7–9 (9.64 ± 2.62 ug/l,p < 0.001) and days 12–14 (6.89 ± 1.35 ug/l, p < 0.001) in patientswho developed vasospasm. Patients without vasospasm alsoshowed a significant increase in CSF ADMA level at days 3–5(5.89 ± 3.32 ug/l, p = 0.007), days 7–9 (5.13 ± 2.37 ug/l, p = 0.001)and days 12–14 (4.92 ± 2.29 ug/l, p = 0.002). CSF ADMA level at

entrations of asymmetric dimethylarginine correlate with adverse clinicalx.doi.org/10.1016/j.jocn.2013.11.038

Table 1Clinical characteristics and outcome at 2 years in patients with subarachnoid hemorrhage

Patient Age, years Sex Fisher grade Hunt-Hess grade Reason for SAH or aneurysm site Treatment Presence of vasospasm KPS

1 40 M 2 2 ACoA Coil N 602 41 M 3 3 ACoA, L-PCoA Clip Y 1003 55 M 2 1 R-ACA Clip N 1004 54 F 3 4 ACoA Coil N 505 44 M 2 1 DAVF Coil N 1006 65 F 4 3 ACoA Coil Y 407 47 F 1 2 DSA(�) – N 908 51 M 1 1 DSA(�) – N 1009 54 F 4 3 ACoA Coil Y 50

10 44 F 2 1 ACoA Clip N 9011 58 M 2 2 ACoA Clip Y 3012 56 F 2 2 ACoA Coil N 9013 55 F 1 2 L-PCoA Clip N 8014 52 F 2 2 L-CA Clip N 9015 46 F 2 1 L-PCoA, L-CA Coil N 9016 56 F 3 2 R-PCoA Coil N 9017 51 F 2 3 L-PCoA Coil N 9018 65 F 3 4 R-PCoA – Y 019 62 F 3 4 L-PCoA, R-MCA Coil Y 8020 51 F 4 2 L-PICA Coil Y 90

ACA = anterior cerebral artery, ACoA = anterior communicating artery, CA = carotid artery, DAVF = dural arteriovenous fistula, DSA = digital subtraction angiography,F = female, KPS = Karnofsky Performance Status, L = left, M = male, MCA = middle cerebral artery, N = no, PCoA = posterior communicating artery, PICA = posterior inferiorcerebellar artery, R = right, SAH = subarachnoid hemorrhage, Y = yes.

Fig. 1. (A) Mean ± standard deviation of cerebrospinal fluid asymmetric dimethylarginine (ADMA) concentrations in subarachnoid hemorrhage (SAH) patients and a controlgroup. (B) Elevated levels of ADMA are found in patients both with and without cerebral vasospasm following SAH. CSF = cerebrospinal fluid. ⁄p < 0.05.

H. Li et al. / Journal of Clinical Neuroscience xxx (2014) xxx–xxx 3

days 7–9 in patients with vasospasm was significantly higher thanthat in non-vasospasm patients (p = 0.001) (Fig. 1B). In binarylogistic regression analysis, higher ADMA level at days 7–9 pre-dicted cerebral vasospasm (p = 0.022) (Table 2).

3.3. CSF ADMA levels in relation to SAH outcome

KPS at 2 years showed a significant correlation with CSF levelsof ADMA at days 3–5 (CC = �0.57, p = 0.028; 95% CI �0.84 to�0.08) and days 7–9 (CC = �0.55, p = 0.012; 95% CI �0.80 to�0.14), but not at days 12–14 (CC = �0.43, p = 0.086; 95% CI�0.75 to �0.07) (Fig. 2A–C). SAH patients with a good outcomeat 2 years (KPS P 60) had much lower ADMA levels at days 3–5(p = 0.010) and days 7–9 (p = 0.008) than those who with a pooroutcome (KPS < 60) (Fig. 2D). Higher ADMA level at days 7–9 was

Table 2Binary logistic regression analysis of ADMA, cerebral vasospasm and 2 year outcome after

Cerebrovasospasm

OR 95% CI

ADMA at days 3–5 1.154 0.856–2.777ADMA at days 7–9 2.051 1.111–3.788

ADMA at days 12–14 1.678 0.917–3.072

ADMA = asymmetric dimethylarginine, CI = confidence interval, OR = odds ratio.

Please cite this article in press as: Li H et al. Increased cerebrospinal fluid concoutcome in subarachnoid hemorrhage patients. J Clin Neurosci (2014), http://d

a predictive marker of poor outcome in binary logistic regressionanalysis (p = 0.039) (Table 2).

4. Discussion

In this prospective study, we aimed to investigate the changesof CSF ADMA levels in patients with SAH and to determine whetherthese changes were related to the long-term outcomes of SAHpatients. Therefore, patients with other existing systemic diseases,including hypertension, hyperlipidemia, diabetes mellitus, arterialatherosclerosis, chronic renal failure and peripheral vascular dis-ease, that may be associated with CSF ADMA levels [9,17–19] wereexcluded. Our data demonstrated that CSF ADMA levels were sig-nificantly elevated after SAH and the increased concentrations atdays 7–9 after SAH correlated with adverse clinical outcomes. To

subarachnoid hemorrhage

Poor outcome

p value OR 95% CI p value

0.149 2.253 0.873–5.815 0.0930.022 1.722 1.029–2.882 0.0390.093 1.828 0.862–3.877 0.116

entrations of asymmetric dimethylarginine correlate with adverse clinicalx.doi.org/10.1016/j.jocn.2013.11.038

Fig. 2. Karnofsky Performance Status (KPS) score at 2 years after subarachnoid hemorrhage (SAH) shows significant correlation with cerebrospinal fluid levels of asymmetricdimethylarginine (ADMA) at (A) days 3–5 and (B) days 7–9, (C) but not at days 12–14. (D) SAH patients with a good outcome (KPS P 60) had significantly lower levels ofADMA at days 3–5 and days 7–9 than those with a poor outcome (KPS < 60). CSF = cerebrospinal fluid. ⁄p < 0.05.

4 H. Li et al. / Journal of Clinical Neuroscience xxx (2014) xxx–xxx

the best of our knowledge, this is the first report of such correlationin SAH studies.

The biochemical pathways relevant to generation and elimina-tion of ADMA, NO production and SAH are complex. After SAH, thefunction of NOS is destroyed by hemoglobin, which is graduallyreleased from erythrocytes in the subarachnoid space, creatingan increased level of ADMA [20]. Decreased ADMA eliminationdue to a reduction in the ADMA-hydrolyzing enzyme immunoreac-tivity in the arteries in spasm is another reason for the increase ofADMA levels [11,20]. Data from clinical and experimental studiesdemonstrate that elevation of ADMA levels in the CSF of patientswith SAH strongly correlates with the degree and time course ofcerebral vasospasm [11–13,21]. Our present findings are compati-ble with these data and add to the evidence that increased levels ofADMA in CSF may play an important role in initiation andmaintenance of vasospasm following SAH. It should be noted thatwe found significant increases of CSF ADMA levels in patients bothwith and without vasospasm, which differs from a previous reportthat significant changes of CSF ADMA concentrations were onlyfound in patients with vasospasm [12]. This difference might dueto the relatively low patient numbers in our study groups.

In ischemic stroke studies, increased CSF ADMA concentrationswere found to be associated with stroke severity [17]. Additionally,elevated plasma ADMA levels independently predicted adverseoutcomes of acute ischemic stroke at 90 days after stroke onset[22]. Recently, a prospective observational study found that alow plasma arginine:ADMA ratio predicted poorer survival forSAH patients [23]. Our present data, showing that the KPS scoreat 2 year follow-up significantly negatively correlated with CSF lev-els of ADMA at days 7–9 (p = 0.012), and that higher CSF ADMA lev-els at days 7–9 predicted adverse outcomes of SAH patients(p = 0.039), suggest a strong relationship between elevation ofCSF ADMA at days 7–9 and clinical outcome. This conclusion wasconfirmed by the finding that SAH patients with good outcomes(KPS P 60) had significantly lower levels of ADMA at days 7–9

Please cite this article in press as: Li H et al. Increased cerebrospinal fluid concoutcome in subarachnoid hemorrhage patients. J Clin Neurosci (2014), http://d

than those with poor outcomes (KPS < 60) (p = 0.008). The mecha-nism behind this is not clear, but it can be hypothesized that earlyincreased ADMA levels impair the activity of endothelial NOS andthereby inhibit the production of NO, ultimately leading to cerebralvasospasm, which is a well-recognized contributor to poor out-come [24,25]. As CSF ADMA concentrations were significantly ele-vated in patients both with and without vasospasm, wehypothesize that increased CSF ADMA levels lead to adverse clini-cal outcomes of SAH patients not only by initiating vasospasm butalso by other mechanisms. This hypothesis is supported by in vitroand in vivo studies showing that ADMA increases vascular tone incerebral blood vessels [22,26]. Moreover, infusion of ADMA in tohealthy volunteers significantly reduces cerebral perfusion [27].Furthermore, ADMA has been reported to increase endothelialsuperoxide radical production and nuclear factor kappa B activa-tion, resulting in enhanced endothelial adhesiveness for mono-cytes, and might induce apoptosis of endothelial cells [28,29].Therefore, elevation of ADMA may be involved in other pathologi-cal process that contribute to an adverse prognosis of SAH.

Despite the value of these findings, our study has some limita-tions. First, the sample size is relatively small compared with otherlarge-scale studies. Further studies with a larger study populationare warranted. Second, this study was a single-center observa-tional study, and most of the patients (11/20, 55%) were treatedwith endovascular intervention, which might have a potentialimpact on the long-term outcome. However, the results of thisstudy, as well as the findings by Jung et al. [13], should encouragefurther investigation in to ADMA in patients with SAH.

5. Conclusions

Our study suggests that increased CSF ADMA level at days 7–9after SAH is correlated with lower KPS score and predicts anadverse clinical outcome at 2 years after SAH. Based on these find-

entrations of asymmetric dimethylarginine correlate with adverse clinicalx.doi.org/10.1016/j.jocn.2013.11.038

H. Li et al. / Journal of Clinical Neuroscience xxx (2014) xxx–xxx 5

ings, we believe that the prognostic value and clinicopathologicalrole of ADMA deserves further investigation with larger patientpopulations.

Acknowledgements

We are grateful to Professor Wei-yan Li and the staff of theDepartment of Anesthesiology for sample collection.

References

[1] Schuette AJ, Barrow DL. Epidemiology and long-term mortality insubarachnoid hemorrhage. World Neurosurg 2013;80:264–5.

[2] Pluta RM, Oldfield EH. Analysis of nitric oxide (NO) in cerebral vasospasm afteraneurysmal bleeding. Rev Recent Clin Trials 2007;2:59–67.

[3] Rodling-Wahlström M, Olivecrona M, Koskinen LO, et al. Subarachnoidhaemorrhage induces an inflammatory response followed by a delayedpersisting increase in asymmetric dimethylarginine. Scand J Clin Lab Invest2012;72:484–9.

[4] Vallance P, Leone A, Calver A, et al. Endogenous dimethylarginine as aninhibitor of nitric oxide synthesis. J Cardiovasc Pharmacol 1992;20:S60–2.

[5] Wanby P, Teerlink T, Brudin L, et al. Asymmetric dimethylarginine (ADMA) as arisk marker for stroke and TIA in a Swedish population. Atherosclerosis2006;185:271–7.

[6] Mittermayer F, Krzyzanowska K, Exner M, et al. Asymmetric dimethylargininepredicts major adverse cardiovascular events in patients with advancedperipheral artery disease. Arterioscler Thromb Vasc Biol 2006;26:2536–40.

[7] Leong T, Zylberstein D, Graham I, et al. Asymmetric dimethylarginineindependently predicts fatal and nonfatal myocardial infarction and stroke inwomen: 24-year follow-up of the population study of women in Gothenburg.Arterioscler Thromb Vasc Biol 2008;28:961–7.

[8] Chen S, Li N, Deb-Chatterji M, et al. Asymmetric dimethyarginine as markerand mediator in ischemic stroke. Int J Mol Sci 2012;13:15983–6004.

[9] Hsu CP, Lin SJ, Chung MY, et al. Asymmetric dimethylarginine predicts clinicaloutcomes in ischemic chronic heart failure. Atherosclerosis 2012;225:504–10.

[10] Mamatha SN, Nagaraja D, Philip M, et al. Asymmetric dimethylarginine as arisk marker for early-onset ischemic stroke in Indian population. Clin ChimActa 2011;412:139–42.

[11] Jung CS, Iuliano BA, Harvey-White J, et al. Association between cerebrospinalfluid levels of asymmetric dimethyl-L-arginine, an endogenous inhibitor ofendothelial nitric oxide synthase, and cerebral vasospasm in a primate modelof subarachnoid hemorrhage. J Neurosurg 2004;101:836–42.

[12] Jung CS, Oldfield EH, Harvey-White J, et al. Association of an endogenousinhibitor of nitric oxide synthase with cerebral vasospasm in patients withaneurysmal subarachnoid hemorrhage. J Neurosurg 2007;107:945–50.

[13] Jung CS, Lange B, Zimmermann M, et al. The CSF concentration of ADMA, butnot of ET-1, is correlated with the occurrence and severity of cerebralvasospasm after subarachnoid hemorrhage. Neurosci Lett 2012;524:20–4.

Please cite this article in press as: Li H et al. Increased cerebrospinal fluid concoutcome in subarachnoid hemorrhage patients. J Clin Neurosci (2014), http://d

[14] Sloan MA, Alexandrov AV, Tegeler CH, et al. Assessment: transcranial Dopplerultrasonography: report of the Therapeutics and Technology AssessmentSubcommittee of the American Academy of Neurology. Neurology2004;62:1468–81.

[15] Teerlink T, Nijveldt RJ, de Jong S, et al. Determination of arginine, asymmetricdimethylarginine, and symmetric dimethylarginine in human plasma andother biological samples by high-performance liquid chromatography. AnalBiochem 2002;303:131–7.

[16] Karnofsky DA, Burchenal JH. The clinical evaluation of chemotherapeuticagents in cancer. In: MacLeod CM, editor. Evaluation of chemotherapeuticagents. New York: Columbia Univ Press; 1949. p. 196.

[17] Brouns R, Marescau B, Possemiers I, et al. Dimethylarginine levels incerebrospinal fluid of hyperacute ischemic stroke patients are associatedwith stroke severity. Neurochem Res 2009;34:1642–9.

[18] Yoo JH, Lee SC. Elevated levels of plasma homocyst(e)ine and asymmetricdimethylarginine in elderly patients with stroke. Atherosclerosis2001;158:425–30.

[19] Krzyzanowska K, Mittermayer F, Wolzt M, et al. Asymmetric dimethylargininepredicts cardiovascular events in patients with type 2 diabetes. Diabetes Care2007;30:1834–9.

[20] Pluta RM. Dysfunction of nitric oxide synthases as a cause and therapeutictarget in delayed cerebral vasospasm after SAH. Acta Neurochir Suppl2008;104:139–47.

[21] Martens-Lobenhoffer J, Sulyok E, Czeiter E, et al. Determination ofcerebrospinal fluid concentrations of arginine and dimethylarginines inpatients with subarachnoid haemorrhage. J Neurosci Methods2007;164:155–60.

[22] Worthmann H, Chen S, Martens-Lobenhoffer J, et al. High plasmadimethylarginine levels are associated with adverse clinical outcome afterstroke. J Atheroscler Thromb 2011;18:753–61.

[23] Staalsø JM, Bergström A, Edsen T, et al. Low plasma arginine:asymmetricdimethyl arginine ratios predict mortality after intracranial aneurysm rupture.Stroke 2013;44:1273–81.

[24] Macdonald RL, Higashida RT, Keller E, et al. Preventing vasospasm improvesoutcome after aneurysmal subarachnoid hemorrhage: rationale and design ofCONSCIOUS-2 and CONSCIOUS-3 trials. Neurocrit Care 2010;13:416–24.

[25] Bederson JB, Connolly Jr ES, Batjer HH, et al. Guidelines for the management ofaneurysmal subarachnoid hemorrhage: a statement for healthcareprofessionals from a special writing group of the Stroke Council, AmericanHeart Association. Stroke 2009;40:994–1025.

[26] Faraci FM, Brian Jr JE, Heistad DD. Response of cerebral blood vessels to anendogenous inhibitor of nitric oxide synthase. Am J Physiol1995;269:H1522–7.

[27] Kielstein JT, Donnerstag F, Gasper S, et al. ADMA increases arterial stiffness anddecreases cerebral blood flow in humans. Stroke 2006;37:2024–9.

[28] Böger RH, Bode-Böger SM, Tsao PS, et al. An endogenous inhibitor of nitricoxide synthase regulates endothelial adhesiveness for monocytes. J Am CollCardiol 2000;36:2287–95.

[29] Jiang DJ, Jia SJ, Dai Z, et al. Asymmetric dimethylarginine induces apoptosis viap38 MAPK/caspase-3-dependent signaling pathway in endothelial cells. J MolCell Cardiol 2006;40:529–39.

entrations of asymmetric dimethylarginine correlate with adverse clinicalx.doi.org/10.1016/j.jocn.2013.11.038