Molecular Signatures of Vascular Injury Are Associated With Early Growth of Intracerebral Hemorrhage

    the Department of Neurology (Y.S., A.D.), Hospital Doctor Josep Trueta, Girona
    the Department of Neurology (R.L., J.C.), Hospital Clínico Universitario, Santiago de Compostela
    the Department of Neurology (J.T.), Hospital Virgen Blanca, León
    the Department of Neurology (J.M.L.), Hospital Clínico Universitario, Valencia, Spain.

    Abstract

    Background and Purpose— To investigate whether molecular markers of inflammation and endothelial injury are associated with early growth of intracerebral hemorrhage (ICH).

    Methods— In a multicenter prospective study, we determined concentrations of interleukin-6 (IL-6), tumor necrosis factor- (TNF-), matrix metalloproteinase-9 (MMP-9), and cellular fibronectin (c-Fn) in blood samples obtained on admission from 183 patients with primary hemispheric ICH of <12 hours’ duration. Patients had a neurological evaluation and a computed tomography (CT) scan performed at baseline and at 48±6 hours. Early growth of the ICH was defined as a volume increase >33% between the 2 CT examinations for ICH with a baseline volume <20 mL and >10% for ICH 20 mL. Clinical, radiological, and biochemical predictive factors of ICH enlargement were analyzed by logistic regression analysis.

    Results— Fifty-four (29.5%) patients showed a relevant early growth of ICH. High leukocyte count and fibrinogen levels, low platelet count, and intraventricular bleeding were associated with early ICH growth in bivariate analyses. Plasma concentrations of IL-6 (median : 19.6 [13.6; 29.9] versus 15.9 [11.5; 19.8] pg/mL), TNF- (13.5 [8.4; 30.5] versus 8.7 [4.7; 13.5] pg/mL), MMP-9 (153.3 [117.7; 204.7] versus 70.6 [47.8; 103.8] ng/mL), and c-Fn (8.8 [6.2; 12.5] versus 2.8 [1.6; 4.2] μg/mL) were significantly higher in patients with early growth of ICH (all P<0.001). C-Fn levels >6 μg/mL (OR, 92; 95%CI, 22 to 381; P<0.0001) and IL-6>24 pg/mL (OR, 16; 95%CI, 2.3 to 119; P=0.005) were independently associated with ICH enlargement in the logistic regression analysis.

    Conclusions— Molecular signatures of vascular injury and inflammatory markers in the early acute phase of ICH are associated with subsequent enlargement of the hematoma.

    Key Words: blood–brain barrier  computed tomography  hematoma  inflammation  intracerebral hemorrhage  outcome  prognosis

    Introduction

    Intracerebral hemorrhage (ICH) causes a 35% to 50% 30-day mortality. Half of this mortality occurs within the first 2 days as a result of brain herniation, mainly caused by the continued bleeding that provokes an enlargement of the hematoma during the first 24 hours.1 Early hematoma growth (EHG) has been associated with early neurological worsening and poor outcome, but no clinical or radiological predictive factors have been identified and the pathogenesis remains unclear.2,3 A better knowledge of the molecular mechanisms involved in the early growth and perilesional brain injury of ICH may help to find a more appropriate treatment in the acute phase.

    EHG has been related to multifocal bleeding in the periphery of the clot caused by the rupture of arterioles and venules in the perilesional low-flow zone.4 Secondary brain injury has been attributed to ischemic damage and particularly to the toxic effects of thrombin generation by the clot.5 In experimental ICH, thrombin activates the inflammatory cascade and the expression of matrix metalloproteinases (MMPs), causing the breakdown of the blood–brain barrier and edema formation.6–8 In this context, high serum concentrations of cytokines and MMP-9 have been associated with a large volume of peripheral hypodensity in human ICH.9,10 MMPs are able to degrade the basal membrane components, such as cellular fibronectin (c-Fn), a glycoprotein especially important for the adhesion of platelets to fibrin, a function necessary for the blockade of bleeding.11

    Increased levels of MMP-9 and c-Fn have been found in the blood of patients with hemorrhagic transformation after cerebral infarction.12,13 Therefore, we hypothesize that the activation of the inflammatory cascade, MMP-9 overexpression, and c-Fn degradation may occur after ICH in the tissue around the hematoma, and that these molecules might be involved in vessel rupture and hematoma growth. The aim of this study was to investigate whether high concentrations of cytokines, MMP-9, and c-Fn in peripheral blood are associated with the early growth of ICH.

    Subjects and Methods

    This is a secondary analysis of a prospective, multicenter study performed to identify predictors of early neurological deterioration in 266 patients with primary hemispheric ICH admitted within 12 hours from the onset of symptoms.14 Exclusion criteria were age younger than 18, surgical treatment on admission, coma with expected death within 48 hours, and hemorrhage secondary to brain tumor, trauma, drug abuse, coagulation disorders, anticoagulant therapy, or vascular malformation. For the purpose of this study we also excluded patients without stored frozen blood samples obtained on admission (n=74) and those in whom computed tomography (CT) scan was not available at 48 hours (n=4). The study was approved by the institutional review boards of all participating institutions and informed consent was obtained from the patients or their relatives.

    On arrival to the emergency department, blood pressure and body temperature were recorded and blood samples were taken. Each patient underwent a baseline head CT scan and a Canadian Stroke Scale (CSS)15 evaluation by an experienced neurologist. Patients were admitted to a neurological ward or an acute stroke unit and were treated by a specialized stroke team and nursing staff following established guidelines.16 Antihypertensive treatment with intravenous labetalol or captopril was administered in case of systolic blood pressure >185 mm Hg or diastolic blood pressure >105 mm Hg. Low-dose subcutaneous heparin was used for the prevention of deep vein thrombosis and pulmonary thromboembolism. None of the patients was part of a therapeutic clinical trial.

    A second CT scan was performed at 48±6 hours. The evaluation of all CT scans was performed at the coordinating center by a single investigator who was blinded to the clinical and biochemical data. Lesion volumes were calculated on the radiographic plate using the formula 0.5 x a x b x c (where a is the maximal longitudinal diameter, b is the maximal transverse diameter, and c is the number of 10-mm slices containing hemorrhage). The volume of the ICH plus that of the zone of peripheral hypodensity was determined using the same volumetric method described; the absolute volume of the hypodensity was calculated by subtracting the volume of the ICH from that of the total lesion (ICH plus peripheral hypodensity). According to a previous report, relevant EHG was defined as a volume increase >33% between the 2 CT for those ICHs with a baseline volume <20 mL, and a volume increase >10% for those hemorrhages with a baseline volume 20 mL.14 Secondary analyses were performed using the >33% growth definition for all patients in a way to be compared with another prospective study in which this definition was used.2 The ICH topography was classified as lobar when it affected predominantly the cortical or subcortical white matter of the cerebral lobes, or as deep when it was limited to the internal capsule, the basal ganglia, or the thalamus. The presence of intraventricular extension of the hematoma, leukoaraiosis, and mass effect was also recorded.

    Early neurological deterioration (END) was diagnosed when the CSS score decreased 1 or more points between admission and 48 hours after admission. This difference represents the change with the highest sensitivity, retaining good specificity.17 Patients who died within the first 48 hours were classified in the END group if they had progressed during the observations that followed after inclusion. Functional outcome was evaluated by the modified Rankin scale at 90 days. Patients with a modified Rankin scale score >2 were classified in the poor outcome category.

    Laboratory Determinations

    Blood samples were collected on admission in tubes with potassium edetate, centrifuged at 3000g for 5 minutes, and immediately frozen and stored at –80°. IL-6 and tumor necrosis factor- (TNF-) were measured with commercially available quantitative sandwich enzyme-linked immunosorbent assay (Quantikine) kits obtained from R&D Systems. MMP-9 was measured with commercially available quantitative sandwich enzyme-linked immunosorbent assay kits obtained from Biotrack Amersham Pharmacia, UK. c-Fn was measured with enzyme-linked immunosorbent assay kits obtained from Boehringer, Germany. Laboratory determinations were performed blinded to clinical and neuroimaging findings.

    Statistical Analysis

    Categorical variables are shown as percentages. Lesion volumes, CSS score, and the molecular markers are presented as median values and , and the rest of the continuous variables are presented as mean (SD). Tests performed were the 2 or 2-sided Fisher exact tests for categorical variables, and the Student t test or the Mann–Whitney test for continuous variables as appropriate (SPSS 10 software). Spearman correlation was used to correlate continuous variables.

    Potential predictors of EHG in the bivariate analyses (P<0.05) were tabulated and were then analyzed by logistic regression (probability of entry P<0.05). In a further logistic model, we investigated whether predictors of EHG were also independently associated with END, poor functional outcome, and mortality at 3 months. We tested the linearity of the explanatory variables related to the risk of EHG before performing the logistic models. Variables that showed no linearity were categorized by means of the Robert method.18 Moreover, all possible plausible interactions among variables were tested. Results were expressed as adjusted odds ratio with corresponding 95% confidence intervals.

    Results

    the total series of 266 patients, 183 patients were included in the study. Early ICH growth occurred in 54 patients (29.5%), in 15 (22.3%) of 67 patients with baseline ICH <20 mL, and in 39 (33.6%) of 116 patients with baseline ICH 20 mL. The frequency of EHG in the total series was similar (26.9%), and there were no statistical differences in epidemiological, clinical, radiological, or analytical data between the general population and the target population (data not shown).

    Potential predictors of EHG in the bivariate analysis are shown in Table 1. Age, gender, frequency of risk factors, time from symptoms onset to admission, CSS score, body temperature, and blood pressure were similar in both groups. Patients with EHG had larger volume of peripheral hypodensity, higher leukocyte count and plasma fibrinogen levels at admission, and lower platelet count and intraventricular bleeding than did non-EHG patients.

    Plasma concentrations of IL-6, TNF-, MMP-9, and c-Fn were significantly higher in patients with subsequent EHG (Table 2). Similar results were found when EHG was defined according to the >33% growth definition for all patients. Concentrations of these molecules by the percentage of change in the ICH volume at 48 hours are shown in the Figure. A highly significant correlation was found between plasma c-Fn and MMP-9 levels on admission and the percentage of ICH growth (r=0.77 and r=0.64, respectively; both P<0.001). Also, a significant moderate correlation was found between baseline TNF- and IL-6 levels and the percentage of ICH growth (r=0.26 and r=0.32, respectively; both P<0.001).

    Boxplots showing median values (horizontal line inside the box), quartiles (box boundaries), and the largest and smallest observed values (lines drawn from the end of the box) of IL-6 (A), TNF- (B), MMP-9 (C), and c-Fn (D) by the percentage of ICH growth between admission and 48 hours.

    Because of lack of linearity, IL-6, TNF-, MMP-9, and c-Fn were classified in 2 categories. Of all these variables associated with EHG in bivariate analyses, plasma c-Fn levels >6 μm/mL and IL-6 levels >24 pg/m were associated with increased risk of EHG in the final logistic model, whereas intraventricular bleeding was associated with a decreased risk (Table 3). No interactions were found. c-Fn >6 μm/mL (OR, 297; 95% CI, 28 to 3128) was the only predictive factor of EHG according to the >33% growth definition for all patients.

    Follow-up at 90 days was completed in 53 patients with EHG and in 123 without EHG. EHG was significantly associated with an increased risk of END (OR, 3.7; 95% CI, 1.1 to 12.4), mortality (OR, 5.2; 95% CI, 1.9 to 14.2), and poor functional outcome at 3 months (OR, 3.7; 95% CI, 1.0 to 13.2) after adjustment for potential confounders in the bivariate analysis (data not shown).

    Discussion

    The knowledge of the underlying mechanisms and factors associated with EHG is crucial because they represent potential targets for therapeutic interventions. In the only previous prospective study, Brott et al failed to reveal any clinical, radiological, or analytic predictor of ICH growth.2 The present study has demonstrated that high plasma levels of c-Fn and IL-6 at baseline are independent predictors of ICH enlargement. These findings support the idea that some molecular signatures in blood of endothelial damage and inflammatory response may help to predict patients with a high risk of subsequent EHG. This fact is clinically relevant, because this study has confirmed that ICH growth is associated with a 3.7-fold increase in the odds of early neurological deterioration and poor functional outcome, and a 5.2-fold increase in the odds of mortality at 3 months.

    The present results are consistent with the pathophysiology of brain edema and secondary neuronal injury in ICH. After the initial arterial rupture, the activation of the coagulation cascade produces a large quantity of thrombin that is implicated in several functions, including chemotaxis of leukocytes, expression of adhesion molecules, release of inflammatory cytokines, blood–brain barrier disruption, and local metalloproteinase generation.7,19 Furthermore, the release of iron after erythrocyte lysis may contribute to blood–brain barrier dysfunction, possibly through a free radical-mediated damage of endothelial wall.20 Although all these mechanisms seem to be involved in edema formation after ICH,9,10 their role in the EHG remains unclear. Taken together, our findings suggest a greater acute inflammatory response in patients with subsequent ICH enlargement.21,22 Higher number of leukocytes and levels of fibrinogen, IL-6, and TNF- in the peripheral blood were found in patients with EHG, in whom we also found a greater volume of peripheral edema at baseline. Notably, IL-6 levels >24 pg/mL increased 16-fold the risk of EHG after controlling for other markers of inflammation. Intraventricular bleeding appeared to be a protective factor for ICH growth but presumably was caused by the extravasation of blood into the ventricular system.

    The relationship between an increased inflammatory reaction and EHG might be caused by the disappearance of the basal lamina components, such as c-Fn, laminin, and collagen IV, and by the loss of microvascular integrity in the tissue around the hematoma caused by the activation of matrix metalloproteinases.6,23–26 In this context, both MMP-9 and c-Fn concentrations in blood were significantly higher in patients with EHG, and c-Fn was the most powerful predictor of ICH enlargement. Plasma c-Fn levels >6 μg/mL were associated with 92-fold increase in the risk of EHG, and c-Fn levels showed a high correlation with the percentage of the ICH growth. Because c-Fn is largely confined to the vascular endothelium,27 high plasma levels of this molecule might be indicative of endothelial damage. In fact, plasma c-Fn levels have been reported to be increased in patients with vascular injury secondary to vasculitis, sepsis, acute major trauma, and diabetes, and in patients with ischemic stroke.28,29 In addition, c-Fn plays an important role in blood clot formation by mediating the adhesion of platelets to fibrin,30 so the disappearance of the c-Fn of the vascular endothelium might damage this clotting mechanism, facilitating ICH enlargement. However, the synthesis of c-Fn may be triggered during inflammatory processes by agents such as transforming growth factors and leukocytes.31,32

    This study has a number of limitations. First, this is a secondary study in patients selected from a larger series of ICH studied with the aim to investigate factors associated with neurological deterioration. However, we can reasonably rule out a selection bias because baseline characteristics and the frequency of ICH growth were similar in the original and target populations. Second, because of exclusion criteria, our findings cannot be generalized to patients in coma at admission or who die within the first 48 hours after admission. Finally, an increase of the molecular markers as a result of an acute phase reaction or previous systemic diseases cannot be completely ruled out. However, a direct relationship is likely because increased c-Fn and IL-6 levels were detected at admission before EHG in patients in whom radiological findings, biochemical parameters, and vital signs evaluated at the moment the blood samples were drawn were not different from those in patients in whom a relevant ICH growth did not develop, so we cannot attribute the differences in these markers to a different acute-phase response or to a distinct previous comorbidity.

    In conclusion, the present study demonstrates that molecular signatures in blood of inflammatory response and endothelial basal lamina disruption within the first 12 hours of ICH are important predictors of subsequent EHG. These findings may open new therapeutic strategies for the treatment of ICH.

    Appendix

    Study coordinator: José Castillo, Hospital Clínico Universitario, Santiago de Compostela.

    Participating centers, investigators, and the number of patients studied: Hospital Clínico Universitario, Santiago de Compostela (62): José Castillo, Rogelio Leira; Hospital Universitari Doctor Josep Trueta, Girona (47): Antonio Dávalos, Yolanda Silvia; Hospital Virgen del Rocío, Sevilla (21): Alberto Gil Peralta, Enrique Montes; Hospital Virgen Blanca, León (21): Javier Tejada; Hospital Clinic, Barcelona (20): ángel Chamorro, Nicolás Vila; Hospital Arquitecto Marcide, Ferrol (19): Francisco López, José Aldrey; Hospital Clínico, Valencia (17): José Miguel Láinez, Raquel Chamarro; Hospital Provincial, Pontevedra (16): Manuel Seijo Martínez; Hospital de La Princesa, Madrid (13): José Vivancos, Raquel González; Hospital Clínico, Madrid (9): José Egido; Hospital Vall d’Hebrón, Barcelona (8): José álvarez-Sabín, Joan Montaner; Hospital Gregorio Maraón, Madrid (5): Antonio Gil, Fernando Díaz; Hospital Virgen de la Concha, Zamora (3): José Carlos Gómez; Hospital La Paz, Madrid (3): Exuperio Diez-Tejedor; Hospital Clínico, Zaragoza (2): Enrique Mostacero.

    Acknowledgments

    We thank Dr Maria M. Garcia (Unit of Biostatistics. Hospital Universitari Doctor Josep Trueta, Girona, Spain) for her helpful advice and work on the statistical analysis of data. Dr Silva was granted by the Sociedad Espaola de Neurología, and by the Fundació Doctor Josep Trueta.

    Footnotes

    Partial results of this investigation were presented at the 28th International Stroke Conference in Phoenix, Ariz, February 2003.

    References

    Kazui S, Naritomi H, Yamamoto H, Sawada T, Yamaguchi T. Enlargement of spontaneous intracerebral hemorrhage. Incidence and time course. Stroke. 1996; 27: 1783–1787.

    Brott T, Broderick J, Kothari R, Barsan W, Tomsick T, Sauerbeck L, Spilker J, Duldner J, Khouri J. Early hemorrhage growth in patients with intracerebral hemorrhage. Stroke. 1997; 28: 1–5.

    Fujii Y, Tanaka R, Takeuchi S, Koike T, Minakawa T, Sasaki O. Hematoma enlargement in spontaneous intracerebral hemorrhage. J Neurosurg. 1994; 80: 51–57. [Order article via Infotrieve]

    Mayer Sa, Lignelli A, Fink ME, Kessler DB, Thomas CE, Swarup R, Van Heertum RL. Perilesional blood flow and edema formation in acute intracerebral hemorrhage: a SPECT study. Stroke. 1998; 29: 1791–1798.

    Mendelow AD. Mechanisms of ischemic brain damage with intracerebral hemorrgae. Stroke. 1993; 24 (suppl 1): 115–117.

    Lee KR, Kawai N, Seoung K, Sagher O, Hoff JT. Mechanisms of edema formation after intracerebral hemorrhage: effects of thrombin on cerebral blood flow, blood brain barrier permeability and cell survival in a rat model. J Neurosurg. 1997; 86: 272–278. [Order article via Infotrieve]

    Xi G, Wagner KR, Keep RF, Hua Y, de Courten-Mayers G, Broderick JP, Brott TG, Hoff JT. Role of blood clot formation on early edema development after experimental intracerebral hemorrhage. Stroke. 1998; 29: 2580–2586.

    Rosenberg GA, Navratil M. Metalloproteinase inhibition blocks edema in intracerebral hemorrhage in the rat. Neurology. 1997; 48: 921–926.

    Castillo J, Davalos A, Alvarez-Sabin J, Pumar JM, Leira R, Silva Y, Montaner J, Kase CS. Molecular signatures of brain injury after intracerebral hemorrhage. Neurology. 2002; 58: 624–629.

    Abilleira S, Montaner J, Molina C, Monasterio J, Castillo J, Alvarez-Sabín J. Matrix metalloproteinase-9 concentration alter spontaneous intracerebral hemorrhage. J Neurosurg. 2003; 99: 65–70. [Order article via Infotrieve]

    Makogonenko E, Tsurupa G, Ingham K, Medved L. Interaction of fibrin(ogen) with fibronectine: further characterization and localization of the fibronectin-binding site. Biochemistry. 2002; 41: 7907–7913. [Order article via Infotrieve]

    Castellanos M, Leira R, Serena J, Pumar JM, Lizasoain I, Castillo J, Davalos A. Plasma metalloproteinase-9 concentration predicts hemorrhagic transformation in acute ischemic stroke. Stroke. 2004; 35: 1671–1676.

    Castellanos M, Leira R, Serena J, Blanco M, Pedraza S, Castillo J, Dávalos A. Plasma cellular-fibronectin concentration predicts hemorrhagic transformation after thrombolytic therapy in acute ischemic stroke.

    Leira R, Dávalos A, Silva Y, Gil-Peralta A, Tejada J, Garcia M, Castillo J. Predictors and associated factors of early neurological deterioration in intracerebral hemorrhage. Neurology. 2004; 63: 461–467.

    Cote R, Battista RN, Wolfson C, Boucher J, Adam J, Hachinski V. The Canadian Neurological Scale: validation and reliability assessment. Neurology. 1989; 39: 638–643.

    Láinez JM, Pareja A, Martí-Fábregas J, Leira R. Guía de actuación clínica en la hemorragia cerebral. Neurología. 2002; 17 (suppl 3): 76–86.[In Spanish].

    Dávalos A, Cendra E, Teruel J, Martínez M, Genís D. Deteriorating ischemic stroke: risk factors and prognosis. Neurology. 1990; 40: 1865–1869.

    Roberts C, Vermont J, Bosson JL. Formulas for thershold computations. Comput Biomed Res. 1991; 24: 514–529. [Order article via Infotrieve]

    Lee KR, Colon GP, Betz AL, Keep RF, Kim S, Hoff JT. Edema from intracerebral hemorrhage: the role of thrombin. J Neurosurg. 1996; 84: 91–96. [Order article via Infotrieve]

    Xi G, Hua Y, Bhasin RR, Ennis SR, Keep RF, Hoff JT. Mechanisms of edema formation after intracerebral hemorrhage. Effects of extravasated red blood cells on blood flow and blood-brain barrier integrity. Stroke. 2001; 32: 2932–2938.

    Suzuki S, Kelley RE, Dandapani BK, Reyes-Iglesias Y, Dietrich WD, Duncan RC. Acute leukocyte and temperature response in hypertensive intracerebral hemorrhage. Stroke. 1995; 26: 1020–1023.

    Margaglione M, Grandone E, Mancini FP, Di Minno G. Genetic modulation of plasma fibrinogen concentrations: possible importance of interleukine-6. J Thromb Thrombolysis. 1996; 3: 51–56. [Order article via Infotrieve]

    Hamman GF, Okada Y, Fitridge R, Del Zoppo GJ. Microvascular basal lamina antigens disappear during cerebral ischemia and reperfusion. Stroke. 1995; 26: 2120–2126.

    Horstmann S, Kalb P, Koziol J, Gardner H, Wagner S. Profiles of matrix metalloproteinases, their inhibitors, and laminin in stroke patients. Influence of different therapies. Stroke. 2003; 34: 2165–2172.

    Aoki T, Sumii T, Mori T, Wang X, Lo EH. Blood-brain barrier disruption and matrix metalloproteinase-9 expression during reperfusion injury. Mechanical versus embolic focal ischemia in spontaneously hypertensive rats. Stroke. 2002; 33: 2711–2717.

    Forsyth KD, Levinsky RJ. Fibronectin degradation: an in-vitro model of neutrophil mediated endothelial cell damage. J Pathol. 1990; 161: 313–319. [Order article via Infotrieve]

    Vartio T, Laitinen L, Narvanen O, Cutolo M, Thornell LE, Zardi L, Virtanen I. Differential expression of the ED sequence-containing form of cellular fibronectin in embryonic and adult human tissues. J Cell Sci. 1987; 88: 419–430.

    Peters JH, Maunder RJ, Woolf AD, Cochrane GH, Ginsberg MH. Elevated plasma levels of ED1+ ("cellular") fibronectin in patients with vascular injury. J Lab Clin Med. 1989; 113: 586–597. [Order article via Infotrieve]

    Kanters SD, Banga JD, Algra A, Frijns RC, Beutler JJ, Fijnheer R. Plasma levels of cellular fibronectin in diabetes. Diabetes Care. 2000; 24: 323–327.

    Hynes RO. Fibronectins. Sci Am. 1986; 254: 42–51. [Order article via Infotrieve]

    La Fleur M, Beaulieu AD, Kreis C, Poubelle P. Fibronectin gene expression in polymorphonuclear leukocytes. Accumulation of mRNA in inflammatory cells. J Biol Chem. 1987; 262: 2111–2115.

    Roberts CJ, Birkenmeier TM, McQuillan JJ et al. Transforming growth factor  stimulates the expression of fibronectin and of both subunits of the human fibronectin receptor by cultured human lung fibroblasts. J Biol Chem. 1988; 263: 4586–4592.