Neurocrit Care DOI 10.1007/s12028-014-0089-2
ORIGINAL ARTICLE
Blood Metabolomic Predictors of 1-Year Outcome in Subarachnoid Hemorrhage Rickard L. Sjo¨berg • Tommy Bergenheim • Lina Mo¨re´n • Henrik Antti Cecilia Lindgren • Silvana Naredi • Peter Lindvall
•
Ó Springer Science+Business Media New York 2015
Abstract Background Delayed neurological deficit (DND) is the most important cause of morbidity and mortality in patients with subarachnoid hemorrhage (SAH) whose aneurysms have been secured. However, the methods currently used to predict the development of DND, such as trans-cranial Doppler or levels biochemical markers in blood and cerebrospinal fluid are not very accurate. Method Venous blood was drawn from 50 patients with SAH, admitted to the neurosurgical department Umea˚ University Hospital, at day 1–3 and day 7 after the bleed. The clinical status of the patients was followed up approximately 1 year after this episode and classified according to the Glasgow Outcome Score (GOS). Results Results showed considerable differences in blood metabolomic patterns between day 1–3 and 7 after the hemorrhage. Fifty-six out of 98 metabolites could be identified from our in-house library and 17 of these
metabolites changed significantly from day 1–3 to 7 after the bleed. One of these, myo-inositol, was predictive of clinical outcome even after correction for multiple testing. An estimation of the diagnostic accuracy of high levels of this substance in predicting good outcome (GOS 4–5) yielded a sensitivity of .763 and a specificity of .5 at the optimal cut off point. Conclusions SAH is an event with a profound effect on blood metabolomics profiles. Myo-inositol might be an interesting compound for future study to focus on in the search for metabolic markers in venous blood of delayed neurological deterioration in SAH patients. Keywords Delayed neurological deficit Myo-inositol Metabolomics Subarachnoid hemorrhage Vasospasm Venous blood
Introduction
R. L. Sjo¨berg T. Bergenheim P. Lindvall Department of Pharmacology and Clinical Neuroscience, Umea˚ University, Umea˚, Sweden R. L. Sjo¨berg (&) Department of Neurosurgery, Umea˚ University Hospital, 901 85 Umea˚, Sweden e-mail: [email protected] L. Mo¨re´n H. Antti Department of Chemistry, Computational Life Science Cluster, Umea˚ University, Umea˚, Sweden C. Lindgren S. Naredi Department of Anaesthesiology and Intensive Care Medicine, Institute of Surgical and Pereoperative Sciences, Umea˚ University, Umea˚, Sweden
Some 3–12 days after an aneurysmal subarachnoid hemorrhage (SAH), many patients will develop so called delayed neurological deficits (DND). During the latter part of the twentieth century this phenomenon was widely understood as caused by vasospasm in the large cerebral arteries [18, 25, 31, 35]. As a remnant of this era, the perhaps most frequently used method to detect threats to a favorable long term neurological outcome in the individual patient is still through monitoring of blood flow velocity in the middle cerebral arteries (MCA) by means of transcranial Doppler ultrasound (TCD) [20, 36]. The literature, however, suggests that despite its relative utility in accurately diagnosing angiographic vasospasm, the ability of TCD to predict DND when defined as neurological outcome is far less impressive [5, 7, 19, 36].
123
Neurocrit Care
Recently several other metabolic and pathophysiological processes have emerged as implicated in the development of DND. Alterations in the balance between the vasodilator nitric oxide (NO) and the vasoconstrictor endothelin 1 and its importance to micro-vascular spasm is one example [30]. Free radical production and oxidation of lipid membranes [28] effects of spreading depolarization [6] and induction of apoptotic cascades [4] are others. However, when it comes to early markers for the development of cerebral ischemia that might become useful in clinical practice only a few attempts to investigate levels of specific substances such as asymmetric dimethylL-arginine (ADMA), endothelin 1 and neuropeptide Y (NPY) in venous blood and CSF, respectively, has recently been made [4, 15, 16, 22, 32]. The results have been modest. Even though several of these substances have been linked to vasospasm their relation to DND are at best small or at worst absent. In a search for possible metabolic markers of pathology two different strategies may be applied. The one most frequently used so far, is the theory-driven one. When this strategy is used, single, or a small set of metabolites, that from a theoretical point of view could be expected to be of importance in this particular context, are studied. The second possibility would be to apply a non-hypothesisdriven approach by ‘‘globally’’ investigating the metabolome by mass spectroscopic methods. Mainly two techniques for this are available today: nuclear magnetic resonance spectroscopy (1H-NMR) and more potent, mass spectrometry coupled to some type of chromatographic pre-separation step usually by means of gas chromatography (GC/MS) or liquid chromatography (LC/MS). Mass spectrometry based techniques are superior in terms of sensitivity and both can detect and quantify hundreds of metabolites in biological samples [26, 33]. In the present study, we used GC/MS to study metabolite levels and changes in blood metabolites at day 1–3 and 7 in 50 patients treated at our institution after an aneurysmal subarachnoid hemorrhage. The aim was to explore possible relations between metabolite levels and parameters describing the clinical status of the patients at admission and at a 1-year follow-up.
Methods Patients The study included patients C18 years of age with an aneurysmal SAH treated at Umea˚ University Hospital between March 2008 and August 2009 arriving at the hospital within 48 h of initial symptoms of the bleeding. Patients were consecutively included when resources in
123
terms of research nurses were available. All patients were scored according to Hunt and Hess at arrival [12] and the primary CT investigations were evaluated according to the Fisher scale [10]. All patients that needed intensive care and controlled ventilation were treated according to a local protocol that includes normoventilation, normovolaemia and keeping glucose, sodium, hemoglobin, and albumin within normal limits. Nimodipine (NimotopÒ, Bayer, Leverkusen, Germany) 0.2 mg/ml, was administered 1.0–15 ml/h intravenously during the first 10 days after bleeding as prophylaxis for vasospasm. Repeated venous blood samples were collected, the first at day 1, 2, or (in 1 patient) on day 3 and the second at day 7 after initial symptoms. Venous serum samples were collected in 10 ml plain glass blood-tubes (BD VacutainerÒ) spun down, fractionated and frozen in 2 ml polypropylene vials (Sarstedt AG & Co) at -80 °C within 45 min. At follow-up, approximately 1 year after the SAH, patients were scored according to Glasgow Outcome Scale (GOS), extending from; 1 (dead) to 5 (good recovery) [14]. At this time informed consent was obtained from either patients or relatives. The study was approved by the ethics committee at Umea˚ University. Sample Preparation Serum samples were allowed to thaw in room temperature for 30 min. To 100 ll aliquots of serum, 900 ll extraction solution (methanol (90 %) and water (10 %)) spiked with 11 IS (7 ng/ll) was added followed by rigorous agitation at 30 Hz for 2 min in a bead mill (MM 400, Retsch GmbH, Haan, Germany) and storage on ice for 120 min, prior to centrifugation at 14,000 rpm (10 min at 4 °C (Centrifuge 5417R, Eppendorf, Hamburg, Germany). Two hundred microliter of supernatants were transferred to GC vials and evaporated to dryness using a speedvac (miVac, Quattro concentrator, Barnstead Genevac, Ipswich, UK). Evaporated samples were stored in -80 °C waiting for derivatization. Derivatization was carried out in two steps. First, methoxyamination, by adding 30 ll methoxyamine in pyridine (15 lg/ll), 10 min of shaking and 60 min of heating at 70 °C with a 16 h reaction time (in room temperature). Second, trimethylsilylation, by adding 30 ll MSFTA (Nmethyl-N-trimethylsilyl-trifluoroacetamide) + 1 % TMCS (Trimethylchlorosilane) for 1 h (in room temperature). Finally, 30 ll heptane including methyl stearate (15 ng/ll) was added as injection standard. GC/MS One microliter of derivatized sample was injected splitless by an Agilent 7683 Series autosampler (Agilent, Atlanta,
Neurocrit Care Fig. 1 Cross-validated score plot based on the final OPLS model discriminating the early sample time points (black dots) from the later sample time points (grey diamonds). The xaxis consists of the patients included. (R2X:0.415, R2Y:0.837, Q2:0.754)
GA) into an Agilent 6980 GC equipped with a 10 m 9 0.18 mm i.d. fused-silica capillary column chemically bonded with 0.18 lm DB5-MS stationary phase (J&W Scientific, Folsom, CA). The injector temperature was set to 270 °C. Helium was used as carrier gas at a constant flow rate of 1 ml/min through the column. The purge time was set to 60 s at a purge flow rate of 20 ml/min and an equilibration time of 1 min for every analysis. Initially, the column temperature was kept to 70 °C for 2 min and then increased to 320 °C at 30 °C/min, where it was kept for 2 min. The column effluent was introduced into the ion source of a Pegasus III TOFMS (Leco Corp., St Joseph, MI). The transfer line temperature was set to 250 °C and the ion source temperature to 200 °C. Ions were generated by a 70 eV electron beam at a current of 2.0 mA. Masses were acquired from m/z 50 to 800 at a rate of 30 spectra/s, and the acceleration voltage was turned on after a solvent delay of 165 s. Hierarchical Multivariate Curve Resolution NetCDF files of acquired data were exported to MATLAB 7.11.0 (R2010b) (Mathworks, Natick, MA) where baseline correction, alignment, time-window settings were carried out prior to resolving of pure chromatographic and spectral profiles using hierarchical multivariate curve resolution (HMCR).
(MKS-Umetrics, Umea˚, Sweden) comparing spectroscopic patterns between day 1–3 and day 7. Variables unaffected by sample time, variable with low model weight values (|w*| < .05), were excluded to obtain the final OPLS-DA model. Model validation was performed using crossvalidation and p values for cross-validated model were calculated by ANOVA. Second, two tailed t tests for independent samples were performed comparing levels of all metabolic compounds identified between day 1–3 and day 7. Level of significance was set to p < 0.05 and Bonferroni correction applied as described in the results section. Third, levels of substances that changed significantly (after Bonferroni correction) were related to the following variables by means of Spearman’s rho: I. Glasgow Outcome Score at 1 year. II. Hunt and Hess score at admission. III. Fisher grade at admission. Level of significance for these tests was set at p < 0.05 and Bonferroni correction applied as described in the results section. Finally, the prognostic value of metabolites that were significantly related to GOS at 1-year follow-up was described using a Reciever Operating Charachteristic (ROC) curve calculated with the help of the computer software SPSS (IBM). In all statistical analysis the level of substances that could not be detected was deemed to be zero.
Statistical Analysis
Results
First, Orthogonal Partial Least Squares discriminant analysis (OPLS-DA) was performed on unit variance (UV) scaled data using the computer software SIMCA-P 13
During the study period 95 patients were assessed for inclusion in the study. Twenty-one patients were not included since they arrived during summer holidays,
123
Neurocrit Care Table 1 Identified substances that changed significantly between day 1–3 and day 7 in venous blood samples from 50 patients with aneurysmal subarachnoid hemorrhage Substance
b-D-Methylglucopyranoside
t
3.45
Sign at .05 (p < 8.9e-4)
Sign at .01 (p < 1.78e-4)
Increase or decrease
p (2-tailed independent t test with equal variances assumed)
*
NS
:
8.30e-4
Valine
3.66
*
NS
:
4.10e-4
Threonine
4.04
*
**
:
1.07e-4
3,4-Dihydroxybutanoic acid
4.30
*
**
:
3.98e-5
Benzenamine
4.64
*
**
:
1.08e-5
Glutamine
4.62
*
**
:
1.17e-5
Leucine
4.69
*
**
:
8.99e-6
Creatinine
4.90
*
**
:
3.67e-6
Methionine
5.46
*
**
:
3.68e-7
Trypthophan
6.04
*
**
:
2.76e-8
Phenylalanine
6.33
*
**
:
7.36e-9
Uric acid
-4.69
*
**
;
9.03e-6
Lauric acid
-4.71
*
**
;
8.32e-6
Myo-inositol Stearic acid
-4.74 -5.60
* *
** **
; ;
7.20e-6 2.06e-7
Octadecenoic acid
-6.01
*
**
;
3.15e-8
Linoleic acid
-6.23
*
**
;
1.16e-8
* Signifies values significant at the .05 level after correction for multiple testing ** Signifies values significant at the .01 level after correction for multiple testing
Christmas, or New Year when research nurses were not available. Six patients were excluded due to late arrival, >48 h after initial symptoms, 6 patients had a nonaneurysmal hemorrhage as determined either trough digital subtraction angiography, or computed tomography (CT) angiography. Five patients were excluded since they had previous aneurysmal bleedings and five patients were excluded for other reasons. Fifty patients remained and they had a median age of 59 years (range 26–82). There were 35 females and 15 males. The median Hunt and Hess score was 3 (range 1–5) and the median Fisher grade was 4 (range 2–4) the median GCS grade at admission was 11.5 (range 3–15), and median for GOS at 1 year was 5 (range 1–5). Thirty-three patients were treated for their aneurysm by endovascular intervention, 15 were treated by open surgery and two patients were not treated. The median number of days patients were subject to controlled ventilation was 4.5 (range 0–10) and the median amount of days patients were on vasopressor, or inotropic support was 3 (range 0–10). Further details on this patient cohort have previously been described by Lindgren et al. [21]. Metabolomic Analyses Ninety-eight metabolites were resolved from the GC/MS data. Metabolite changes between day 1–3 and day 7 were first explored by comparing the metabolic profiles from the
123
different sample time points with OPLS-DA. This analysis which is illustrated in Fig. 1 show an almost complete separation between these two points in time. The final model shown is based on 59 metabolites that survived the (|w*| < .05) criteria (F = 48.6 DF = 6 p = 2,24049e-27). Fifty-six of the 98 metabolites could be identified from our in-house library. Statistical analyses revealed that 17 of these metabolites did change significantly from day 1–3 to 7 after Bonferroni correction for multiple testing which changed the level of significance from p < .05 to < 8.9e-4. Levels of these substances and t tests are displayed in Table 1. Relation of Significantly Changed Substances to Clinical Variables Levels of 3,4-Dihydroxybutanoic acid day 1–3 was significantly related to HH score at admission (Spearman’s rho = .411 p = .003). None of the other 17 identified substances described in Table 1 were at day 1–3 significantly related to any of the clinical variables after Bonferroni correction for multiple testing (which changed the level of significance from p < .05 to 160 m/s had a sensitivity in predicting DND of only 52 % and a sensitivity of 48 %. In a larger and more recent study Carrera et al. [5] analyzed 1877 TCD examinations in 441 SAH patients within 14 days of onset. They described the sensitivity and specificity for different levels of MCA flow velocities as measured by TCD to predict Delayed Cerebral Ischemia. With this method a ROC AUC value of 0.5 would mean that the diagnostic utility would be similar to that of tossing a coin. Carrera et al. found that TCD examinations barely surpassed this level with a ROC AUC value of 0.59. Several other studies have yielded similar modest results [19, 36]. This illustrates that even though DND correlates with angiographic vasospasm in major arteries it may actually not be caused by it [25, 31]. Similar demonstrations of dissociation between angiographic vasospasm and clinical outcome have been demonstrated in pharmacological studies. Here the substance clazosentan that has a demonstrated ability to prevent vasospasm is apparently unable to prevent the development of DND whereas nimodipine, a substance with a disputed effect on vasospasm does prevent DND [8, 11, 24, 27, 29]. However, it should be noted that Budohoski et al. recently reported a AUC of .81 for the ability of a more sophisticated TCD measurement, combined with simultaneous measurements of systemic blood pressure, to predict early DND [2]. The ability of levels of myo-inositol to predict neurological outcome in SAH patients described in this study is modest with a ROC AUC of .763. Still, it should be noted that this result, by far exceeds many of those shown by TCD, as discussed above and that they are also better than any other results of any other predictive substance such as asymmetric dimethyl-L-arginine (ADMA), endothelin 1 and neuropeptide Y (NPY) [15, 16, 22, 32]. This result clearly suggests that myo-inositol may be an interesting compound for further study in this context. However, replication in prospective studies is of course necessary before any recommendation for monitoring of serum myoinositol levels in SAH patients in clinical practice can be made. For a study with the purpose of identifying substances with potential predictive value for outcome in SAH patients, our approach has some important advantages. First, we focused on metabolite levels from venous blood, which is an entity that is regularly sampled during routine care of these patients in most hospitals. Second, the
123
unbiased approach of mass spectrometry may be important in overcoming some of the risks of false positive results common in theory-driven studies. At the same time an obvious disadvantage of this approach is that interpretations of the biomedical mechanisms underlying any positive results may be difficult. Another weakness of the non-hypothesis-driven approach used in the present study is of course that the exact mechanism underlying our findings remains to be elucidated. However, three different aspects of the biological function of myo-inositol deserve mention. First, myo-inositol is a precursor to inositol 1,4,5-triphosphate [13]. The last step in the formation of inositol 1,4,5-triphosphate is induced by endothelin-1 stimulation of the endothelin receptor type A. This stimulation will in turn lead to increases of intracellular calcium in cerebral smooth muscle cells resulting in vasoconstriction [25]. Second, it has recently been experimentally demonstrated that exposure to CSF from SAH patients will lead to a inositol 1,4,5-triphosphate dependent rise in calcium levels in human astrocytes that will eventually increase mitochondrial permeability leading to necrosis of the cells [17]. It may thus be speculated that high serum levels of myo-inositol may be a marker for a disruption in the formation of inositol 1,4,5-thriphosphate that will eventually work to balance these destructive mechanisms. Third, myo-inositol is a potent osmolyte that protects cells exposed to hyperosmotic stress [3, 34] and is also despite its vasodilating and osmotic properties known to decrease blood pressure in ways similar to Mannitol [1]. Although speculative one possible explanation for beneficial effects of endogenous myo-inositol in the blood stream may thus be that it might counteract the effects of cytotoxic and vasogenic edema both by effects on blood pressure leading to a decrease in the pressure gradient across the blood brain barrier and by binding fluid to the blood stream. Finally, a third finding of our study was that high levels of 3,4-dihydroxybutanoic acid were significantly positively related to Hunt and Hess scores at both day 1–3 and 7. This compound is a phenolic acid which has previously been shown to have anti-oxidative and anti-inflammatory properties [23]. Though the possible clinical significance of this finding is less clear, it might be speculated that levels of this compound might be a marker for a cardiovascular profile conferring a higher risk of severe SAH.
Conclusion The present study, in which metabolites in venous blood of 50 SAH patients was studied with a high resolution
Neurocrit Care
spectroscopy approach revealed clear patterns of metabolic changes between day 1–3 and 7, possibly indicating the entering of the patients into a catabolic state. Furthermore, the individual metabolite myo-inositol was shown to correlate significantly with 1-year outcome after SAH and levels of this metabolite on day 7 after SAH showed a reasonably good ability to predict a good clinical outcome (ability to return to work or school) in this patient cohort. Though these results should be replicated and explored further before clinical applications, they do suggest that myo-inositol might be an interesting compound for future study to focus on in the search for metabolic markers in venous blood of delayed neurological deterioration in SAH patients. Acknowledgment This research was supported by Grants from the County Council of Va¨sterbotten, Sweden, and by Umea˚ University. Conflict of interest declare.
The authors have no conflicts of interest to
References 1. Bernstein LM, Ivy AC. Inositol and mannitol hexanitrates in hypertension management. Circulation. 1955;12:353–60. 2. Budohoski KP, Czosnyka M, Smielewski P, Kasprowicz M, Helmy A, Bulters D, Pickard JD, Kirkpatrick PJ. Impairment of cerebral autoregulation predicts delayed cerebral ischemia after subarachnoid hemorrhage: a prospective observational study. Stroke. 2012;43(12):3230–7. 3. Burg MB. Molecular basis for osmoregulation of organic osmolytes in renal medullary cells. J Exp Zoo. 1994;268:171–5. 4. Cahill WJ, Calvert JH, Zhang JH. Mechanisms of early brain injury after subarachnoid hemorrhage. J Cereb Blood Flow Metab. 2006;26:1341–55. 5. Carrera E, Schmidt JM, Oddo M, Fernandez L, Claassen J, Seder D, Lee K, Badjatia M, Connoly ES, Mayer S. Transcranial doppler for predicting delayed cerebral ischemia after subarachnoid hemorrhage. Neurosurgery. 2009;65:316–24. 6. Dreier J, Ebert N, Priller J, Megow D, Lindauer U, Klee R, Reuter U, Imai Y, Einha¨upl KM, Victorov I, Dimagl U. Products of hemolysis in the subarachnoid space inducing spreading ischemia in the cortex and focal necrosis in rats: a model for delayed ischemic neurological deficits after subarachnoid hemorrhage? J Neurosurg. 2000;93:658–66. 7. Ekelund A, Sa¨veland H, Romner B, Brandt L. Is transcranial doppler sonography useful in detecting late cerebral ischaemia after aneurysmal subarachnoid hemorrhage. Br J Neurosurg. 1996;10:19–25. 8. Etminan N, Vergouwen MD, Ilodigwe D, Macdonald RL. Effect of pharmaceutical treatment on vasospasm, delayed cerebral ischemia, and clinical outcome in patients with aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis. J Cereb Blood Flow Metab. 2011;31(6):1443–51. 9. Felig P, Owen OE, Wahren J, Cahill GF Jr. Amino acid metabolism during prolonged starvation. J Clin Invest. 1969;48(3):584–94. 10. Fisher C, Kistler J, Davis J. Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery. 1980;6(1):1–9.
11. Hou J, Zhang JH. Does prevention of vasospasm in subarachnoid hemorrhage improve clinical outcome? No. Stroke. 2013;44: S34–6. 12. Hunt WE, Hess RM. Surgical risk as related to the time of intervention in the repair of intracranial aneurysms. J Neurosurg. 1968;28:14–20. 13. Inhorn RC, Bansal VS, Majerus PW. Pathway for inositol 1,3,4trisphosphate and 1,4-bisphosphate metabolism. Proc Natl Acad Sci. 1987;84:2170–4. 14. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet. 1975;305:480–4. 15. Jung CS, Lange B, Zimmerman M, Zimmerman S. The CSF concentration of ADMA, but not of ET-1, is correlated with the occurrence and severity of cerebral vasospasm after subarachnoid hemorrhage. Neurosci Lett. 2012;22(524):20416. 16. Jung CS, Lange B, Zimmerman M, Seifert V. CSF and serum biomarkers focusing on cerebral vasospasm and ischemia after subarachnoid hemorrhage. Stroke Res Treat. 2013. doi:10.1155/ 2013/560305. 17. Kasseckert SA, Shahzad T, Miqdad M, Stein M, Abdallah Y, Scharbrodt W, Oertel M. The mechanisms of energy crisis in human astrocytes after subarachnoid hemorrhage. Neurosurgery. 2013;72:468–74. 18. Kassell N, Torner J, Haley E Jr. The international cooperative study on the timing of aneurysm surgery. Part 1: overall management results. J Neurosurg. 1990;73:18–36. 19. Laumer R, Steinmeier R, Go¨nner F, Vogtmann T, Priem R, Fahlbush R. Cerebral hemodynamics in subarachnoid hemorrhage evaluated by transcranial doppler sonography. Part 1: reliability of flow velocities in clinical management. Neurosurgery. 2003;33:1–9. 20. Lindegaard KF, Nornes H, Bakke SJ, Sorteberg W, Nakstad P. Cerebral vasospasm diagnosis by means of angiography and blood velocity measurements. Acta Neurochir. 1989;100:12–24. 21. Lindgren C, Dahlqvist P, Lindvall P, Nilsson L, Koskinen L-O, Naredi S. Cortisol levels are influenced by sedation in the acute phase after subarachnoid haemorrhage. Acta Anaesthesiol Scand. 2013;57:452–60. 22. Lindgren C, Hultin M, Koskinen LO, Lindvall P, Borota L, Naredi S. ADMA levels and arginine/ADMA ratios reflect severity of disease and extent of inflammation after subarachnoid hemorrhage. Neurocrit Care. 2014;21(1):91–101. doi:10.1007/ s12028-013-9945-8. 23. Liu C-L, Wang J-M, Chu C-Y, Cheng M-T, Tseng T-H. In vivo protective effect of protocatechuic acid on tert-butyl hydroperoxide-induced rat hepatotoxicity. Food Chem Toxicol. 2002;40: 635–41. 24. MacDonald RL. Does prevention of vasospasm in subarachnoid hemorrhage improve clinical outcome? Yes. Stroke. 2013;44:S31–3. 25. MacDonald RL, Pluta RM, Zhang JH. Cerebral vasospasm after subarachnoid hemorrhage: the emerging revolution. Nat Clin Pract Neurol. 2009;3(5):256–63. 26. Nicholson J, Connelly JC, Lindon JC, Holmes E. Metabonomics: a platform for studying drug and gene function. Nat Rev Drug Metab. 2002;1:153–61. ¨ hman J, Servo A, Heiskanen O. Long-term effects of nimodipine 27. O on cerebral infarcts and outcome after aneurysmal subarachnoid hemorrhage and surgery. J Neurosurg. 1991;74:8–13. 28. Ostrowski RP, Colohan AR, Zhang JH. Molecular mechanisms of early brain injury after subarachnoid hemorrhage. Neurol Res. 2006;28(4):399–414. 29. Pickard JD, Murray GD, Illingworth R, Shaw MDM, Teasdale GM, Foy PM, Humphrey PRD, Lang DA, Nelson R, Richards P, Sinar J, Bailey S, Skene A. Effect of oral nimodipine on cerebral infarction and outcome after subarachnoid haemorrhage: British aneurysm nimodipine trial. BMJ. 1989;298:636–42.
123
Neurocrit Care 30. Pluta R. Delayed cerebral vasospasm and nitric oxide: review, new hypothesis, and proposed treatment. Pharmacol Ther. 2005;105:23–56. 31. Pluta RM, Hansen-Schwartz J, Dreier J, Vajkoczy P, MacDonald L, Nishizawa S, Kasuva H, Wellman G, Keller E, Zauner A, Dorsch N, Clark J, Ono S, Kiris T, LeRoux P, Zhang JH. Cerebral vasospasm following subarachnoid hemorrhage: time for a new world of thought. Neurol Res. 2009;31:151–8. 32. Schebesch KM, Brawanski A, Bele S, Scho¨del P, Herbst A, Bru¨ndl E, Kagerbauer SM, Martin J, Lohmeier A, Stoerr EM, Proescholdt M. Neuropeptide Y—an early biomarker for cerebral vasospasm after aneurysmal subarachnoid hemorrhage. Neurol Res. 2013;35(10):1038–43. 33. Serkova NJ, Brown MS. Quantitative analysis in magnetic resonance spectroscopy: from metabolic profiling to in vivo biomarkers. Bioanalysis. 2012;4(3):321–41.
123
34. Strange K, Morrison R, Heilig CW, Dipietro S, Gullans SR. Upregulation of inositol transport mediates inositol accumulation in hyperosmolar brain cells. Am J Physiol. 1991;260:C784–90. 35. Weir B, Grace M, Hansen J, Rothberg C. Time course of vasospasm in man. J Neurosurg. 1978;48:173–8. 36. Westermeier T, Pham M, Stetter C, Willner N, Solymosi L, Ernestus RI, Vince GH, Kunze E. Value of transcranial Doppler, perfusion-CT and neurological evaluation to forecast secondary ischemia after aneurysmal SAH. Neurocrit Care. 2013. doi:10. 1007/s12028-013-9896-0.
ORIGINAL ARTICLE
Blood Metabolomic Predictors of 1-Year Outcome in Subarachnoid Hemorrhage Rickard L. Sjo¨berg • Tommy Bergenheim • Lina Mo¨re´n • Henrik Antti Cecilia Lindgren • Silvana Naredi • Peter Lindvall
•
Ó Springer Science+Business Media New York 2015
Abstract Background Delayed neurological deficit (DND) is the most important cause of morbidity and mortality in patients with subarachnoid hemorrhage (SAH) whose aneurysms have been secured. However, the methods currently used to predict the development of DND, such as trans-cranial Doppler or levels biochemical markers in blood and cerebrospinal fluid are not very accurate. Method Venous blood was drawn from 50 patients with SAH, admitted to the neurosurgical department Umea˚ University Hospital, at day 1–3 and day 7 after the bleed. The clinical status of the patients was followed up approximately 1 year after this episode and classified according to the Glasgow Outcome Score (GOS). Results Results showed considerable differences in blood metabolomic patterns between day 1–3 and 7 after the hemorrhage. Fifty-six out of 98 metabolites could be identified from our in-house library and 17 of these
metabolites changed significantly from day 1–3 to 7 after the bleed. One of these, myo-inositol, was predictive of clinical outcome even after correction for multiple testing. An estimation of the diagnostic accuracy of high levels of this substance in predicting good outcome (GOS 4–5) yielded a sensitivity of .763 and a specificity of .5 at the optimal cut off point. Conclusions SAH is an event with a profound effect on blood metabolomics profiles. Myo-inositol might be an interesting compound for future study to focus on in the search for metabolic markers in venous blood of delayed neurological deterioration in SAH patients. Keywords Delayed neurological deficit Myo-inositol Metabolomics Subarachnoid hemorrhage Vasospasm Venous blood
Introduction
R. L. Sjo¨berg T. Bergenheim P. Lindvall Department of Pharmacology and Clinical Neuroscience, Umea˚ University, Umea˚, Sweden R. L. Sjo¨berg (&) Department of Neurosurgery, Umea˚ University Hospital, 901 85 Umea˚, Sweden e-mail: [email protected] L. Mo¨re´n H. Antti Department of Chemistry, Computational Life Science Cluster, Umea˚ University, Umea˚, Sweden C. Lindgren S. Naredi Department of Anaesthesiology and Intensive Care Medicine, Institute of Surgical and Pereoperative Sciences, Umea˚ University, Umea˚, Sweden
Some 3–12 days after an aneurysmal subarachnoid hemorrhage (SAH), many patients will develop so called delayed neurological deficits (DND). During the latter part of the twentieth century this phenomenon was widely understood as caused by vasospasm in the large cerebral arteries [18, 25, 31, 35]. As a remnant of this era, the perhaps most frequently used method to detect threats to a favorable long term neurological outcome in the individual patient is still through monitoring of blood flow velocity in the middle cerebral arteries (MCA) by means of transcranial Doppler ultrasound (TCD) [20, 36]. The literature, however, suggests that despite its relative utility in accurately diagnosing angiographic vasospasm, the ability of TCD to predict DND when defined as neurological outcome is far less impressive [5, 7, 19, 36].
123
Neurocrit Care
Recently several other metabolic and pathophysiological processes have emerged as implicated in the development of DND. Alterations in the balance between the vasodilator nitric oxide (NO) and the vasoconstrictor endothelin 1 and its importance to micro-vascular spasm is one example [30]. Free radical production and oxidation of lipid membranes [28] effects of spreading depolarization [6] and induction of apoptotic cascades [4] are others. However, when it comes to early markers for the development of cerebral ischemia that might become useful in clinical practice only a few attempts to investigate levels of specific substances such as asymmetric dimethylL-arginine (ADMA), endothelin 1 and neuropeptide Y (NPY) in venous blood and CSF, respectively, has recently been made [4, 15, 16, 22, 32]. The results have been modest. Even though several of these substances have been linked to vasospasm their relation to DND are at best small or at worst absent. In a search for possible metabolic markers of pathology two different strategies may be applied. The one most frequently used so far, is the theory-driven one. When this strategy is used, single, or a small set of metabolites, that from a theoretical point of view could be expected to be of importance in this particular context, are studied. The second possibility would be to apply a non-hypothesisdriven approach by ‘‘globally’’ investigating the metabolome by mass spectroscopic methods. Mainly two techniques for this are available today: nuclear magnetic resonance spectroscopy (1H-NMR) and more potent, mass spectrometry coupled to some type of chromatographic pre-separation step usually by means of gas chromatography (GC/MS) or liquid chromatography (LC/MS). Mass spectrometry based techniques are superior in terms of sensitivity and both can detect and quantify hundreds of metabolites in biological samples [26, 33]. In the present study, we used GC/MS to study metabolite levels and changes in blood metabolites at day 1–3 and 7 in 50 patients treated at our institution after an aneurysmal subarachnoid hemorrhage. The aim was to explore possible relations between metabolite levels and parameters describing the clinical status of the patients at admission and at a 1-year follow-up.
Methods Patients The study included patients C18 years of age with an aneurysmal SAH treated at Umea˚ University Hospital between March 2008 and August 2009 arriving at the hospital within 48 h of initial symptoms of the bleeding. Patients were consecutively included when resources in
123
terms of research nurses were available. All patients were scored according to Hunt and Hess at arrival [12] and the primary CT investigations were evaluated according to the Fisher scale [10]. All patients that needed intensive care and controlled ventilation were treated according to a local protocol that includes normoventilation, normovolaemia and keeping glucose, sodium, hemoglobin, and albumin within normal limits. Nimodipine (NimotopÒ, Bayer, Leverkusen, Germany) 0.2 mg/ml, was administered 1.0–15 ml/h intravenously during the first 10 days after bleeding as prophylaxis for vasospasm. Repeated venous blood samples were collected, the first at day 1, 2, or (in 1 patient) on day 3 and the second at day 7 after initial symptoms. Venous serum samples were collected in 10 ml plain glass blood-tubes (BD VacutainerÒ) spun down, fractionated and frozen in 2 ml polypropylene vials (Sarstedt AG & Co) at -80 °C within 45 min. At follow-up, approximately 1 year after the SAH, patients were scored according to Glasgow Outcome Scale (GOS), extending from; 1 (dead) to 5 (good recovery) [14]. At this time informed consent was obtained from either patients or relatives. The study was approved by the ethics committee at Umea˚ University. Sample Preparation Serum samples were allowed to thaw in room temperature for 30 min. To 100 ll aliquots of serum, 900 ll extraction solution (methanol (90 %) and water (10 %)) spiked with 11 IS (7 ng/ll) was added followed by rigorous agitation at 30 Hz for 2 min in a bead mill (MM 400, Retsch GmbH, Haan, Germany) and storage on ice for 120 min, prior to centrifugation at 14,000 rpm (10 min at 4 °C (Centrifuge 5417R, Eppendorf, Hamburg, Germany). Two hundred microliter of supernatants were transferred to GC vials and evaporated to dryness using a speedvac (miVac, Quattro concentrator, Barnstead Genevac, Ipswich, UK). Evaporated samples were stored in -80 °C waiting for derivatization. Derivatization was carried out in two steps. First, methoxyamination, by adding 30 ll methoxyamine in pyridine (15 lg/ll), 10 min of shaking and 60 min of heating at 70 °C with a 16 h reaction time (in room temperature). Second, trimethylsilylation, by adding 30 ll MSFTA (Nmethyl-N-trimethylsilyl-trifluoroacetamide) + 1 % TMCS (Trimethylchlorosilane) for 1 h (in room temperature). Finally, 30 ll heptane including methyl stearate (15 ng/ll) was added as injection standard. GC/MS One microliter of derivatized sample was injected splitless by an Agilent 7683 Series autosampler (Agilent, Atlanta,
Neurocrit Care Fig. 1 Cross-validated score plot based on the final OPLS model discriminating the early sample time points (black dots) from the later sample time points (grey diamonds). The xaxis consists of the patients included. (R2X:0.415, R2Y:0.837, Q2:0.754)
GA) into an Agilent 6980 GC equipped with a 10 m 9 0.18 mm i.d. fused-silica capillary column chemically bonded with 0.18 lm DB5-MS stationary phase (J&W Scientific, Folsom, CA). The injector temperature was set to 270 °C. Helium was used as carrier gas at a constant flow rate of 1 ml/min through the column. The purge time was set to 60 s at a purge flow rate of 20 ml/min and an equilibration time of 1 min for every analysis. Initially, the column temperature was kept to 70 °C for 2 min and then increased to 320 °C at 30 °C/min, where it was kept for 2 min. The column effluent was introduced into the ion source of a Pegasus III TOFMS (Leco Corp., St Joseph, MI). The transfer line temperature was set to 250 °C and the ion source temperature to 200 °C. Ions were generated by a 70 eV electron beam at a current of 2.0 mA. Masses were acquired from m/z 50 to 800 at a rate of 30 spectra/s, and the acceleration voltage was turned on after a solvent delay of 165 s. Hierarchical Multivariate Curve Resolution NetCDF files of acquired data were exported to MATLAB 7.11.0 (R2010b) (Mathworks, Natick, MA) where baseline correction, alignment, time-window settings were carried out prior to resolving of pure chromatographic and spectral profiles using hierarchical multivariate curve resolution (HMCR).
(MKS-Umetrics, Umea˚, Sweden) comparing spectroscopic patterns between day 1–3 and day 7. Variables unaffected by sample time, variable with low model weight values (|w*| < .05), were excluded to obtain the final OPLS-DA model. Model validation was performed using crossvalidation and p values for cross-validated model were calculated by ANOVA. Second, two tailed t tests for independent samples were performed comparing levels of all metabolic compounds identified between day 1–3 and day 7. Level of significance was set to p < 0.05 and Bonferroni correction applied as described in the results section. Third, levels of substances that changed significantly (after Bonferroni correction) were related to the following variables by means of Spearman’s rho: I. Glasgow Outcome Score at 1 year. II. Hunt and Hess score at admission. III. Fisher grade at admission. Level of significance for these tests was set at p < 0.05 and Bonferroni correction applied as described in the results section. Finally, the prognostic value of metabolites that were significantly related to GOS at 1-year follow-up was described using a Reciever Operating Charachteristic (ROC) curve calculated with the help of the computer software SPSS (IBM). In all statistical analysis the level of substances that could not be detected was deemed to be zero.
Statistical Analysis
Results
First, Orthogonal Partial Least Squares discriminant analysis (OPLS-DA) was performed on unit variance (UV) scaled data using the computer software SIMCA-P 13
During the study period 95 patients were assessed for inclusion in the study. Twenty-one patients were not included since they arrived during summer holidays,
123
Neurocrit Care Table 1 Identified substances that changed significantly between day 1–3 and day 7 in venous blood samples from 50 patients with aneurysmal subarachnoid hemorrhage Substance
b-D-Methylglucopyranoside
t
3.45
Sign at .05 (p < 8.9e-4)
Sign at .01 (p < 1.78e-4)
Increase or decrease
p (2-tailed independent t test with equal variances assumed)
*
NS
:
8.30e-4
Valine
3.66
*
NS
:
4.10e-4
Threonine
4.04
*
**
:
1.07e-4
3,4-Dihydroxybutanoic acid
4.30
*
**
:
3.98e-5
Benzenamine
4.64
*
**
:
1.08e-5
Glutamine
4.62
*
**
:
1.17e-5
Leucine
4.69
*
**
:
8.99e-6
Creatinine
4.90
*
**
:
3.67e-6
Methionine
5.46
*
**
:
3.68e-7
Trypthophan
6.04
*
**
:
2.76e-8
Phenylalanine
6.33
*
**
:
7.36e-9
Uric acid
-4.69
*
**
;
9.03e-6
Lauric acid
-4.71
*
**
;
8.32e-6
Myo-inositol Stearic acid
-4.74 -5.60
* *
** **
; ;
7.20e-6 2.06e-7
Octadecenoic acid
-6.01
*
**
;
3.15e-8
Linoleic acid
-6.23
*
**
;
1.16e-8
* Signifies values significant at the .05 level after correction for multiple testing ** Signifies values significant at the .01 level after correction for multiple testing
Christmas, or New Year when research nurses were not available. Six patients were excluded due to late arrival, >48 h after initial symptoms, 6 patients had a nonaneurysmal hemorrhage as determined either trough digital subtraction angiography, or computed tomography (CT) angiography. Five patients were excluded since they had previous aneurysmal bleedings and five patients were excluded for other reasons. Fifty patients remained and they had a median age of 59 years (range 26–82). There were 35 females and 15 males. The median Hunt and Hess score was 3 (range 1–5) and the median Fisher grade was 4 (range 2–4) the median GCS grade at admission was 11.5 (range 3–15), and median for GOS at 1 year was 5 (range 1–5). Thirty-three patients were treated for their aneurysm by endovascular intervention, 15 were treated by open surgery and two patients were not treated. The median number of days patients were subject to controlled ventilation was 4.5 (range 0–10) and the median amount of days patients were on vasopressor, or inotropic support was 3 (range 0–10). Further details on this patient cohort have previously been described by Lindgren et al. [21]. Metabolomic Analyses Ninety-eight metabolites were resolved from the GC/MS data. Metabolite changes between day 1–3 and day 7 were first explored by comparing the metabolic profiles from the
123
different sample time points with OPLS-DA. This analysis which is illustrated in Fig. 1 show an almost complete separation between these two points in time. The final model shown is based on 59 metabolites that survived the (|w*| < .05) criteria (F = 48.6 DF = 6 p = 2,24049e-27). Fifty-six of the 98 metabolites could be identified from our in-house library. Statistical analyses revealed that 17 of these metabolites did change significantly from day 1–3 to 7 after Bonferroni correction for multiple testing which changed the level of significance from p < .05 to < 8.9e-4. Levels of these substances and t tests are displayed in Table 1. Relation of Significantly Changed Substances to Clinical Variables Levels of 3,4-Dihydroxybutanoic acid day 1–3 was significantly related to HH score at admission (Spearman’s rho = .411 p = .003). None of the other 17 identified substances described in Table 1 were at day 1–3 significantly related to any of the clinical variables after Bonferroni correction for multiple testing (which changed the level of significance from p < .05 to 160 m/s had a sensitivity in predicting DND of only 52 % and a sensitivity of 48 %. In a larger and more recent study Carrera et al. [5] analyzed 1877 TCD examinations in 441 SAH patients within 14 days of onset. They described the sensitivity and specificity for different levels of MCA flow velocities as measured by TCD to predict Delayed Cerebral Ischemia. With this method a ROC AUC value of 0.5 would mean that the diagnostic utility would be similar to that of tossing a coin. Carrera et al. found that TCD examinations barely surpassed this level with a ROC AUC value of 0.59. Several other studies have yielded similar modest results [19, 36]. This illustrates that even though DND correlates with angiographic vasospasm in major arteries it may actually not be caused by it [25, 31]. Similar demonstrations of dissociation between angiographic vasospasm and clinical outcome have been demonstrated in pharmacological studies. Here the substance clazosentan that has a demonstrated ability to prevent vasospasm is apparently unable to prevent the development of DND whereas nimodipine, a substance with a disputed effect on vasospasm does prevent DND [8, 11, 24, 27, 29]. However, it should be noted that Budohoski et al. recently reported a AUC of .81 for the ability of a more sophisticated TCD measurement, combined with simultaneous measurements of systemic blood pressure, to predict early DND [2]. The ability of levels of myo-inositol to predict neurological outcome in SAH patients described in this study is modest with a ROC AUC of .763. Still, it should be noted that this result, by far exceeds many of those shown by TCD, as discussed above and that they are also better than any other results of any other predictive substance such as asymmetric dimethyl-L-arginine (ADMA), endothelin 1 and neuropeptide Y (NPY) [15, 16, 22, 32]. This result clearly suggests that myo-inositol may be an interesting compound for further study in this context. However, replication in prospective studies is of course necessary before any recommendation for monitoring of serum myoinositol levels in SAH patients in clinical practice can be made. For a study with the purpose of identifying substances with potential predictive value for outcome in SAH patients, our approach has some important advantages. First, we focused on metabolite levels from venous blood, which is an entity that is regularly sampled during routine care of these patients in most hospitals. Second, the
123
unbiased approach of mass spectrometry may be important in overcoming some of the risks of false positive results common in theory-driven studies. At the same time an obvious disadvantage of this approach is that interpretations of the biomedical mechanisms underlying any positive results may be difficult. Another weakness of the non-hypothesis-driven approach used in the present study is of course that the exact mechanism underlying our findings remains to be elucidated. However, three different aspects of the biological function of myo-inositol deserve mention. First, myo-inositol is a precursor to inositol 1,4,5-triphosphate [13]. The last step in the formation of inositol 1,4,5-triphosphate is induced by endothelin-1 stimulation of the endothelin receptor type A. This stimulation will in turn lead to increases of intracellular calcium in cerebral smooth muscle cells resulting in vasoconstriction [25]. Second, it has recently been experimentally demonstrated that exposure to CSF from SAH patients will lead to a inositol 1,4,5-triphosphate dependent rise in calcium levels in human astrocytes that will eventually increase mitochondrial permeability leading to necrosis of the cells [17]. It may thus be speculated that high serum levels of myo-inositol may be a marker for a disruption in the formation of inositol 1,4,5-thriphosphate that will eventually work to balance these destructive mechanisms. Third, myo-inositol is a potent osmolyte that protects cells exposed to hyperosmotic stress [3, 34] and is also despite its vasodilating and osmotic properties known to decrease blood pressure in ways similar to Mannitol [1]. Although speculative one possible explanation for beneficial effects of endogenous myo-inositol in the blood stream may thus be that it might counteract the effects of cytotoxic and vasogenic edema both by effects on blood pressure leading to a decrease in the pressure gradient across the blood brain barrier and by binding fluid to the blood stream. Finally, a third finding of our study was that high levels of 3,4-dihydroxybutanoic acid were significantly positively related to Hunt and Hess scores at both day 1–3 and 7. This compound is a phenolic acid which has previously been shown to have anti-oxidative and anti-inflammatory properties [23]. Though the possible clinical significance of this finding is less clear, it might be speculated that levels of this compound might be a marker for a cardiovascular profile conferring a higher risk of severe SAH.
Conclusion The present study, in which metabolites in venous blood of 50 SAH patients was studied with a high resolution
Neurocrit Care
spectroscopy approach revealed clear patterns of metabolic changes between day 1–3 and 7, possibly indicating the entering of the patients into a catabolic state. Furthermore, the individual metabolite myo-inositol was shown to correlate significantly with 1-year outcome after SAH and levels of this metabolite on day 7 after SAH showed a reasonably good ability to predict a good clinical outcome (ability to return to work or school) in this patient cohort. Though these results should be replicated and explored further before clinical applications, they do suggest that myo-inositol might be an interesting compound for future study to focus on in the search for metabolic markers in venous blood of delayed neurological deterioration in SAH patients. Acknowledgment This research was supported by Grants from the County Council of Va¨sterbotten, Sweden, and by Umea˚ University. Conflict of interest declare.
The authors have no conflicts of interest to
References 1. Bernstein LM, Ivy AC. Inositol and mannitol hexanitrates in hypertension management. Circulation. 1955;12:353–60. 2. Budohoski KP, Czosnyka M, Smielewski P, Kasprowicz M, Helmy A, Bulters D, Pickard JD, Kirkpatrick PJ. Impairment of cerebral autoregulation predicts delayed cerebral ischemia after subarachnoid hemorrhage: a prospective observational study. Stroke. 2012;43(12):3230–7. 3. Burg MB. Molecular basis for osmoregulation of organic osmolytes in renal medullary cells. J Exp Zoo. 1994;268:171–5. 4. Cahill WJ, Calvert JH, Zhang JH. Mechanisms of early brain injury after subarachnoid hemorrhage. J Cereb Blood Flow Metab. 2006;26:1341–55. 5. Carrera E, Schmidt JM, Oddo M, Fernandez L, Claassen J, Seder D, Lee K, Badjatia M, Connoly ES, Mayer S. Transcranial doppler for predicting delayed cerebral ischemia after subarachnoid hemorrhage. Neurosurgery. 2009;65:316–24. 6. Dreier J, Ebert N, Priller J, Megow D, Lindauer U, Klee R, Reuter U, Imai Y, Einha¨upl KM, Victorov I, Dimagl U. Products of hemolysis in the subarachnoid space inducing spreading ischemia in the cortex and focal necrosis in rats: a model for delayed ischemic neurological deficits after subarachnoid hemorrhage? J Neurosurg. 2000;93:658–66. 7. Ekelund A, Sa¨veland H, Romner B, Brandt L. Is transcranial doppler sonography useful in detecting late cerebral ischaemia after aneurysmal subarachnoid hemorrhage. Br J Neurosurg. 1996;10:19–25. 8. Etminan N, Vergouwen MD, Ilodigwe D, Macdonald RL. Effect of pharmaceutical treatment on vasospasm, delayed cerebral ischemia, and clinical outcome in patients with aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis. J Cereb Blood Flow Metab. 2011;31(6):1443–51. 9. Felig P, Owen OE, Wahren J, Cahill GF Jr. Amino acid metabolism during prolonged starvation. J Clin Invest. 1969;48(3):584–94. 10. Fisher C, Kistler J, Davis J. Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery. 1980;6(1):1–9.
11. Hou J, Zhang JH. Does prevention of vasospasm in subarachnoid hemorrhage improve clinical outcome? No. Stroke. 2013;44: S34–6. 12. Hunt WE, Hess RM. Surgical risk as related to the time of intervention in the repair of intracranial aneurysms. J Neurosurg. 1968;28:14–20. 13. Inhorn RC, Bansal VS, Majerus PW. Pathway for inositol 1,3,4trisphosphate and 1,4-bisphosphate metabolism. Proc Natl Acad Sci. 1987;84:2170–4. 14. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet. 1975;305:480–4. 15. Jung CS, Lange B, Zimmerman M, Zimmerman S. The CSF concentration of ADMA, but not of ET-1, is correlated with the occurrence and severity of cerebral vasospasm after subarachnoid hemorrhage. Neurosci Lett. 2012;22(524):20416. 16. Jung CS, Lange B, Zimmerman M, Seifert V. CSF and serum biomarkers focusing on cerebral vasospasm and ischemia after subarachnoid hemorrhage. Stroke Res Treat. 2013. doi:10.1155/ 2013/560305. 17. Kasseckert SA, Shahzad T, Miqdad M, Stein M, Abdallah Y, Scharbrodt W, Oertel M. The mechanisms of energy crisis in human astrocytes after subarachnoid hemorrhage. Neurosurgery. 2013;72:468–74. 18. Kassell N, Torner J, Haley E Jr. The international cooperative study on the timing of aneurysm surgery. Part 1: overall management results. J Neurosurg. 1990;73:18–36. 19. Laumer R, Steinmeier R, Go¨nner F, Vogtmann T, Priem R, Fahlbush R. Cerebral hemodynamics in subarachnoid hemorrhage evaluated by transcranial doppler sonography. Part 1: reliability of flow velocities in clinical management. Neurosurgery. 2003;33:1–9. 20. Lindegaard KF, Nornes H, Bakke SJ, Sorteberg W, Nakstad P. Cerebral vasospasm diagnosis by means of angiography and blood velocity measurements. Acta Neurochir. 1989;100:12–24. 21. Lindgren C, Dahlqvist P, Lindvall P, Nilsson L, Koskinen L-O, Naredi S. Cortisol levels are influenced by sedation in the acute phase after subarachnoid haemorrhage. Acta Anaesthesiol Scand. 2013;57:452–60. 22. Lindgren C, Hultin M, Koskinen LO, Lindvall P, Borota L, Naredi S. ADMA levels and arginine/ADMA ratios reflect severity of disease and extent of inflammation after subarachnoid hemorrhage. Neurocrit Care. 2014;21(1):91–101. doi:10.1007/ s12028-013-9945-8. 23. Liu C-L, Wang J-M, Chu C-Y, Cheng M-T, Tseng T-H. In vivo protective effect of protocatechuic acid on tert-butyl hydroperoxide-induced rat hepatotoxicity. Food Chem Toxicol. 2002;40: 635–41. 24. MacDonald RL. Does prevention of vasospasm in subarachnoid hemorrhage improve clinical outcome? Yes. Stroke. 2013;44:S31–3. 25. MacDonald RL, Pluta RM, Zhang JH. Cerebral vasospasm after subarachnoid hemorrhage: the emerging revolution. Nat Clin Pract Neurol. 2009;3(5):256–63. 26. Nicholson J, Connelly JC, Lindon JC, Holmes E. Metabonomics: a platform for studying drug and gene function. Nat Rev Drug Metab. 2002;1:153–61. ¨ hman J, Servo A, Heiskanen O. Long-term effects of nimodipine 27. O on cerebral infarcts and outcome after aneurysmal subarachnoid hemorrhage and surgery. J Neurosurg. 1991;74:8–13. 28. Ostrowski RP, Colohan AR, Zhang JH. Molecular mechanisms of early brain injury after subarachnoid hemorrhage. Neurol Res. 2006;28(4):399–414. 29. Pickard JD, Murray GD, Illingworth R, Shaw MDM, Teasdale GM, Foy PM, Humphrey PRD, Lang DA, Nelson R, Richards P, Sinar J, Bailey S, Skene A. Effect of oral nimodipine on cerebral infarction and outcome after subarachnoid haemorrhage: British aneurysm nimodipine trial. BMJ. 1989;298:636–42.
123
Neurocrit Care 30. Pluta R. Delayed cerebral vasospasm and nitric oxide: review, new hypothesis, and proposed treatment. Pharmacol Ther. 2005;105:23–56. 31. Pluta RM, Hansen-Schwartz J, Dreier J, Vajkoczy P, MacDonald L, Nishizawa S, Kasuva H, Wellman G, Keller E, Zauner A, Dorsch N, Clark J, Ono S, Kiris T, LeRoux P, Zhang JH. Cerebral vasospasm following subarachnoid hemorrhage: time for a new world of thought. Neurol Res. 2009;31:151–8. 32. Schebesch KM, Brawanski A, Bele S, Scho¨del P, Herbst A, Bru¨ndl E, Kagerbauer SM, Martin J, Lohmeier A, Stoerr EM, Proescholdt M. Neuropeptide Y—an early biomarker for cerebral vasospasm after aneurysmal subarachnoid hemorrhage. Neurol Res. 2013;35(10):1038–43. 33. Serkova NJ, Brown MS. Quantitative analysis in magnetic resonance spectroscopy: from metabolic profiling to in vivo biomarkers. Bioanalysis. 2012;4(3):321–41.
123
34. Strange K, Morrison R, Heilig CW, Dipietro S, Gullans SR. Upregulation of inositol transport mediates inositol accumulation in hyperosmolar brain cells. Am J Physiol. 1991;260:C784–90. 35. Weir B, Grace M, Hansen J, Rothberg C. Time course of vasospasm in man. J Neurosurg. 1978;48:173–8. 36. Westermeier T, Pham M, Stetter C, Willner N, Solymosi L, Ernestus RI, Vince GH, Kunze E. Value of transcranial Doppler, perfusion-CT and neurological evaluation to forecast secondary ischemia after aneurysmal SAH. Neurocrit Care. 2013. doi:10. 1007/s12028-013-9896-0.
