J Med Dent Sci 2009； 56： 101－106

Original Article

Possible role of cyclooxygenase-2 in developing chronic subdural hematoma

Mutsuya Hara, Masashi Tamaki, Masaru Aoyagi and Kikuo Ohno

Department of Neurosurgery, School of Medicine, Tokyo Medical and Dental University

Inflammatory cytokines are reportedly involved in the pathogenesis of chronic subdural hemato-mas (CSH), and the angiogenesis of hematomas has particularly been in focus. Cyclooxygenase-2 (COX-2) is an essential enzyme for the synthesis of prostaglandin E2 (PGE2). The COX-2-PGE2 pathway has been shown to influence angiogenic factors such as vascular endothelial growth factor (VEGF). We investigated the association of COX-2 expres-sion in the dura mater and outer membrane with the pathogenesis of CSH, and suggested a treat-ment strategy on the basis of this association. Hematoma fluid and serum samples obtained from 37 patients, and samples of the dura mater and outer CSH membrane obtained from 13 patients during the operation were examined in this study. The concentrations of PGE2 in relation to COX-2 in the hematoma fluid were significantly higher than those in the serum. Immunohistochemical analyses revealed COX-2-positive cells in the outer mem-brane of CSHs. There was a linear and significant relationship between PGE2 concentration in hema-toma fluid and the interval from trauma to initial surgery. COX-2 may play a crucial role during the development of CSHs. Our study might lead to the development of anti-COX-2 treatment options that aim to minimize repeat surgery and choose medi-cal therapy by reducing CSH morbidity and recur-rence rate in patients with CSH.

Key words: chronic subdural hematoma, cyclooxyge-nase-2, prostaglandin E2,

Introduction

Chronic subdural hematoma (CSH) is one of the most common diseases affecting the aged, though it some-times involves younger individuals, including children. CSH develops as an encapsulated hematoma in the sub-dural space 2 weeks or more after minor head injuries. The clinical course of this disease is well known, but its causative mechanism remains unclear. Virchow de-scribed CSH as an inflammatory disease and named it pachymeningitis haemorrhagica interna. Later, several studies demonstrated that this process is a local inflam-matory reaction of the dura mater in response to ex-ternal stimuli such as trauma, blood, cerebrospinal fluid (CSF), fibrin, or fibrin degradation products1-5. In fact, as mesenchymal cells in the dural border proliferate and differentiate, they form a sort of inflammatory capsule or membrane around the blood clots or CSF, called the external or outer membrane of the CSH (Fig.1) 2,3,6. The outer membrane of a CSH is composed of granulation-like tissue, which also contains immature vessels and connective fibers, and on the whole, constitutes a source of inflammatory, angiogenic, fibrinolytic, and co-agulation factors7-14. The pathological vascularization of this outer membrane is thought to play a crucial role in the pathogenesis of CSH. Recently, evidence in favor of marked alterations in the local expression of important angiogenic factors has been presented 10,13,15. The ob-servations include (1) a suspiciously high concentration of VEGF in hematoma fluid10,13,15 and (2) stimulation of VEGF gene expression in cells freely floating in hemato-ma fluid15. Cianchi et al. identified a significant correla-tion between COX-2 and VEGF; the latter might be one of the most important mediators of the COX-2 angiogen-ic pathway16. PGE2, whose synthesis from arachidonic acid is catalyzed by COX-2, regulates VEGF expression, and COX-2 inhibitors directly affect angiogenesis17. If COX-2 is overexpressed in the outer membrane of CSH

Corresponding Author：Mutsuya Hara, M.D.Department of Neurosurgery, Musashino Red Cross Hospital, 1-26-1 Kyounancho, Musashino City, Tokyo 180-8610, JapanTel : 81-422-32-3111 Fax : 81-422-32-9551E-mail : mhara@musashino.jrc.or.jpReceived January 30；Accepted June 5, 2009

102 J Med Dent SciM.Hara et al.

and the hematoma fluid has a high concentration of VEGF and PGE2 in relation to COX-2, COX-2 inhibitors may be a new therapeutic modality for CSH. However, there has been no report on the role of COX-2 in CSH development. In the present study, we investigated the concentra-tions of interleukin (IL)-6, IL-8, VEGF, and PGE2 in the hematoma fl uid and serum from CSH patients.

Materials and Methods

Patients We examined 37 patients with CSH, of whom 25 were male and 12 female, at the Tokyo Medical and Dental University Hospital between February 2003 and March 2004. The ages of the patients ranged from 26 to 94 years (mean, 71.4 ± 13.6 years), and CSH was diag-nosed on the basis of computed tomography (CT). Of the 37 patients, 16 had experienced apparent head trauma in the past. The patients had not received any previous treatment for CSH. We treated all the patients surgically by a small burr-hole craniotomy and irrigation under lo-cal anesthesia, followed by closed external drainage for 1 night. Hematoma fluid and serum samples were obtained from all the patients, and samples of the dura mater and outer membrane were obtained from 13 pa-tients during their operation (Fig.1). The hematoma fl uid and serum samples were centrifuged at 3000 rpm for 10 min immediately after the operation, and the super-natant was stored at ‒80℃ until assayed. Indomethacin was added to the samples of hematoma fl uid and serum

immediately after sample collection to block prostaglan-din synthesis. We obtained written informed consent for the analyses of the materials from individual patients or their families. The protocol of this study was approved by the institutional ethics committee of Tokyo Medical and Dental University.

Measurement of IL-6, IL-8, VEGF, and PGE2

We measured the concentrations of IL-6, IL-8, VEGF, and PGE2 in hematoma fluid and serum by using en-zyme-linked immunosorbent assay (ELISA) kits for hu-man IL-6, IL-8, VEGF (TECHNE Corporation, Minneapolis, MN), and PGE2 (Cayman Chemical Company, Ann Arbor, MI). Furthermore, we measured the concentrations of IL-6, IL-8, VEGF, and PGE2 in CSF obtained from 4 patients with hydrocephalus. The normal concentrations of IL-6, IL-8, VEGF, and PGE2 in serum were cited from previous report18,19.

Immunohistochemistry The overlying dura and outer membrane of CSH were obtained during primary burr hole surgery. Resected specimens were fi xed in 10% formalin solution, routine-ly processed, embedded in paraffi n and sectioned at a thickness of 7 μm. These sections were autoclaved at 121℃ for 20 min in 10 mM citrate buff er at pH 6.0. The sections were immersed in 3% H2O2, prepared in distilled water, for 20 min and then in rabbit serum for 30 min to block endogenous peroxidase activity and the nonspecific binding sites, respectively. Goat poly-clonal antibody for human COX-2 (Santa Cruz, CA, USA) diluted 1:80 was then added to the sections, and they were incubated overnight at 4℃. After washing with Tris-buffered saline, a secondary biotin-labeled rabbit anti-goat antibody (IgG) diluted 1:400 was added to the sections, and they were incubated for 30 min. The sec-tions were also incubated with a monoclonal antibody against human macrophage CD68 (DAKO, Glostrup, Denmark), and were then incubated with biotinylated secondary anti-mouse antibodies. Immunohistochemical detection was carried out using the labeled streptavi-din-biotin method (Dako). The sections were fi nally de-veloped with diaminobenzidine and counterstained with hematoxylin.

Analysis of PGE2 and the interval from trauma to initial surgery In the 16 patients who had previously experienced head trauma, we analyzed the correlation between the interval from trauma to initial surgery and the PGE2 concentration in the hematoma fl uid.

Fig.1 : Schema of the pathophysiology of CSH. External stimuli act on the dura matter and induce the formation of the outer membrane. The inside membrane extends onto the brain, forming the inner membrane of the CSH. When the inner membrane totally covers the brain surface, encapsulation of the CSH is complete. * indicates surgical specimen.

103Possible role of cyclooxygenase-2 in developing chronic subdural hematoma

Statistical Analysis All values are expressed as mean ± SD. Statisti-cal significance was analyzed using a nonparametric (Kruskal-Wallis) test. The statistical significance be-tween the trauma group and non-trauma group was an-alyzed using Studentʼs t-test. The correlation between the interval from trauma to initial surgery and PGE2 concentration in the hematoma fl uid was analyzed us-ing Pearsonʼs correlation test. P < 0.01 was considered to be statistically signifi cant.

Results

The concentrations of IL-6, IL-8, VEGF, and PGE2 in the hematoma fl uid were markedly elevated. The con-centrations of IL-6 and VEGF in the serum samples were almost within the normal range. The concentra-tions of IL-8 in serum were slightly higher than those in normal subjects. The concentrations of IL-6, IL-8, VEGF, and PGE2 in the hematoma fl uid were signifi cantly high-er than those in serum (Fig.2). Hematoxylin-eosin (HE) staining of the outer mem-brane showed vascularized and fi brocollagenous tissue infi ltrated by infl ammatory cells, such as neutrophils, eo-sinophils, lymphocytes, macrophages, and plasma cells (Fig.3A). The endothelial cells of sinusoids and capillar-ies and the inflammatory cells in the outer membrane were observed to be positive for COX-2 (Fig.3B). The dura mater was not immunoreactive for COX-2. CD68-positive cells were present in the outer membrane, sug-

gesting that many of the infl ammatory cells are of mac-rophage origin (Fig.3C). A correlation was observed between the PGE2 con-centration in the hematoma fl uid and the interval from trauma to initial surgery. The PGE2 level was signifi-cantly higher in the hematoma fl uid than in the CSF and serum (Fig.2), whereas there was no signifi cant diff er-ence between the PGE2 concentration in the samples of hematoma fl uid obtained from the trauma group and those obtained from the non-trauma group. In the 16 patients who had previously experienced head trauma, a strong positive correlation was observed between the PGE2 concentration in the hematoma fl uid and the inter-val from trauma to initial surgery (Fig.4). On the other hand, no such relationship was observed between the VEGF concentration and the interval from trauma to ini-tial surgery.

Discussion

Currently, it is considered that a CSH is a chronic self-perpetuating inflammatory process that involves the dura mater, and develops in response to injury or ex-ternal stimuli such as trauma, blood, CSF, fi brin, or fi brin degradation products1,2,4,5. Several indications of infl am-mation, such as proliferation of fibroblasts, immature capillaries, and collagen fi brils, and infi ltration by infl am-matory cells have been described in the outer mem-brane of a CSH. Research interest has been focused on the local activity in the outer membrane of a CSH14,20.

Fig.2 : Box plot demonstrating the concentrations of IL-6, IL-8, VEGF, and PGE2 in hematoma fl uid, CSF, and serum. The data are presented as box-and-whisker plots. Boxes represent the 25th and 75th percentiles, and the heavy bar represents the median. Whiskers represent the 10th and 90th percentiles. * indicates signifi cant diff erence (P < 0.01).

104 J Med Dent SciM.Hara et al.

This outer membrane is a source of tissue plasminogen activator and infl ammatory cytokines11,18,21-23. Therefore, CSH can be studied as a type of infl ammatory phenom-enon. The pathological analyses of the outer membrane of a CSH strongly suggest ongoing angiogenesis. Su-zuki et al. first described the higher VEGF concentra-tions in hematomas than in the serum10. Shono et al. also reported the higher concentrations of VEGF in the hematoma fl uid than in serum, and an increased expres-

sion of VEGF and macrophages in the outer membrane of CSHs22. By using immunohistochemical staining for VEGF, Vaquero et al. found the source of angiogenic factors to be the granulation tissues (outer membrane) of the CSH12. Neovascularization with vascular hyper-permeability in the outer membrane has been identifi ed in surgical specimens, and VEGF is deemed pathogno-monic for this structural integrity, although there is still some discordance on this issue24,25. In this study, we found numerous CD68-positive mac-rophages in the outer membrane. A variety of cytokines are secreted from these immune cells and macrophag-es. Suzuki et al. reported the local elevation of infl am-matory cytokines, such as IL-6 and IL-8, in the subdural space of CSH and subdural eff usion11. These cytokines may be responsible for initiating COX-2 expression. In the present study, we found the overexpression of COX-2 in the outer membrane of CSH, and elevated levels of PGE2 in the hematoma fl uid. The COX-2-PGE2 pathway has been shown to infl uence angiogenic fac-tors such as VEGF. An increased expression level of COX-2 in the outer membrane of CSH would lead to an increase in prostaglandin production. It is thought that the major role of COX-2 in angiogenesis is the trigger-ing of the synthesis of prostanoids such as PGE2, which

Fig.3 : Photomicrographs of the outer membrane of the CSH. (A) Hematoxylin-eosin (HE) staining showed many infl ammatory cells and sinusoidal vessels in the fi brous tissue. (B) On immunohistochemical staining for COX-2, infl ammatory and endothelial cells showed immunoreactivity. (C) On immunohistochemical staining for CD68, prominent macrophage infi ltration was observed in the outer membrane of CSH (original magnifi cation, ×200). * indicates hematoma cavity.

Fig.4 : Signifi cant correlation between PGE2 in hematoma fl uid and interval from trauma to initial surgery, as observed in 16 patients. The regression line was obtained using the method of least squares. y = 3088.7 + 34.6x, correlation coeffi cient (r) = 0.66, P < 0.01

105Possible role of cyclooxygenase-2 in developing chronic subdural hematoma

then stimulate the expression of VEGF26. The overexpression of COX-2 and the accompanying increase in prostaglandin production likely create an abnormal state in the outer membrane. The production of prostaglandins plays an important role in regulating the production of angiogenic factors 26 and increasing vascular permeability. The overexpression of COX-2 in human colon carcinoma cells results in angiogenesis by factors such as VEGF16,27. Matsumori et al. demonstrat-ed that prostaglandin participates in both the formation and healing of CSHs19. On the other hand, Katano et al. showed a hypothetical time course for development of CSH with a gradual surge in VEGF levels 28. Our study showed that there is a positive correlation between the concentration of PGE2 in the hematoma fluid and the in-terval from trauma to initial surgery, whereas this corre-lation was not observed in case of VEGF. The changes of PGE2 concentrations in the hematoma fluid are prob-ably associated with the development of CSH. The high levels of PGE2 and VEGF or vascular permeability fac-tor in the hematoma fluid might influence the functional status of vessels inside the outer membrane, causing leakage and extravasation of proteins, and this is an im-portant driving force for hematoma enlargement. VEGF concentrations in the hematoma fluid vary widely, and there was no positive correlation with the interval from trauma to initial surgery because VEGF is produced via several pathways. However, PGE2 is one of promoters of VEGF and produced via the COX-2-PGE2 pathway. The widespread availability of cross-sectional imaging has markedly increased the number of asymptomatic CSH patients or patients in whom the CSH is still devel-oping; such patients do not need immediate evacuation of the hematoma, and can be placed under observation. Our data suggest that drug-based treatment strategies such as the administration of COX-2 inhibitors might provide new and promising pharmacological alterna-tives to inhibit the growth of CSH, presumably by inter-fering with the COX-2 angiogenic pathway. Such treat-ment might even supersede neurosurgical intervention in CSH patients, and potentially decrease the rate of morbidity and mortality of this disease. In the present study, we showed that there was a linear and significant correlation between PGE2 con-centration in hematoma fluid and the interval from trauma to initial surgery. High PGE2 concentrations are thought to reflect an increased activity of COX-2. Thus, in conclusion, COX-2 may play a crucial role in the development of CSH.

References1. Friede RL, Schachenmayr W. The origin ofsubdural

neomembranes. II. Fine structural of neomembranes. Am J Pathol 1978;92: 69-84.

2. Markwalder TM. Chronic subdural hematomas: a review. J Neurosurg 1981;54: 637-45.

3. Markwalder TM, Steinsiepe KF, Rohner M, et al . The course of chronic subdural hematomas after burr-hole craniostomy and closed-system drainage. J Neurosurg 1981;55: 390-6.

4. Nakaguchi H, Tanishima T, Yoshimasu N. Factors in the natural history of chronic subdural hematomas that influence their postoperative recurrence. J Neurosurg 2001;95: 256-62.

5. Yamashima T, Yamamoto S. The origin of inner mem-branes in chronic subdural hematomas. Acta Neuropathol 1985;67: 219-25.

6. Kwon TH, Park YK, Lim DJ, et al. Chronic subdural hema-toma: evaluation of the clinical significance of postopera-tive drainage volume. J Neurosurg 2000;93: 796-9.

7. Hirasima Y, Endo S, Kato R, et al. Platelet-activating fac-tor (PAF) and the development of chronic subdural hae-matoma. Acta Neurochir (Wien) 1994;129: 20-5.

8. Ito H, Saito K, Yamamoto S, et al. Tissue-type plasmino-gen activator in the chronic subdural hematoma. Surg Neurol 1988;30: 175-9.

9. Sarkar C, Lakhtakia R, Gill SS, et al. Chronic subdural haematoma and the enigmatic eosinophil. Acta Neurochir (Wien) 2002;144: 983-8; discussion 8.

10. Suzuki K, Takano S, Nose T, et al. Increased concen-tration of vascular endothelial growth factor (VEGF) in chronic subdural hematoma. J Trauma 1999;46: 532-3.

11. Suzuki M, Endo S, Inada K, et al. Inflammatory cytokines locally elevated in chronic subdural haematoma. Acta Neurochir (Wien) 1998;140: 51-5.

12. Vaquero J, Zurita M, Cincu R. Vascular endothelial growth-permeability factor in granulation tissue of chronic subdural haematomas. Acta Neurochir (Wien) 2002;144: 343-6; discussion 7.

13. Weigel R, Schilling L, Schmiedek P. Specific pattern of growth factor distribution in chronic subdural hematoma (CSH): evidence for an angiogenic disease. Acta Neuro-chir (Wien) 2001;143: 811-8; discussion 9.

14. Yamashima T, Kubota T, Yamamoto S. Eosinophil degran-ulation in the capsule of chronic subdural hematomas. J Neurosurg 1985;62: 257-60.

15. Hohenstein A, Erber R, Schilling L, et al. Increased mRNA expression of VEGF within the hematoma and imbalance of angiopoietin-1 and -2 mRNA within the neomembranes of chronic subdural hematoma. J Neurotrauma 2005;22: 518-28.

16. Cianchi F, Cortesini C, Bechi P, et al. Up-regulation of cy-clooxygenase 2 gene expression correlates with tumor angiogenesis in human colorectal cancer. Gastroenterol-ogy 2001;121: 1339-47.

17. Jones MK, Wang H, Peskar BM, et al. Inhibition of angio-genesis by nonsteroidal anti-inflammatory drugs: insight into mechanisms and implications for cancer growth and

106 J Med Dent SciM.Hara et al.

ulcer healing. Nat Med 1999;5: 1418-23.18. Frati A, Salvati M, Mainiero F, et al. Inflammation mark-

ers and risk factors for recurrence in 35 patients with a posttraumatic chronic subdural hematoma: a prospective study. J Neurosurg 2004;100: 24-32.

19. Matsumori K, Yoshioka M. [Kinetics of prostaglandin and its significance in chronic subdural hematoma]. Neurol Med Chir (Tokyo) 1987;27: 498-504.

20. Sato S, Suzuki J. Ultrastructural observations of the capsule of chronic subdural hematoma in various clinical stages. J Neurosurg 1975;43: 569-78.

21. Fujisawa H, Ito H, Saito K, et al. Immunohistochemical localization of tissue-type plasminogen activator in the lining wall of chronic subdural hematoma. Surg Neurol 1991;35: 441-5.

22. Shono T, Inamura T, Morioka T, et al. Vascular endothe-lial growth factor in chronic subdural haematomas. J Clin Neurosci 2001;8: 411-5.

23. Weigel R, Hohenstein A, Schlickum L, et al. Angiotensin converting enzyme inhibition for arterial hypertension

reduces the risk of recurrence in patients with chronic subdural hematoma possibly by an antiangiogenic mecha-nism. Neurosurgery 2007;61: 788-92; discussion 92-3.

24. Ferrara N. Vascular endothelial growth factor. Eur J Can-cer 1996;32A: 2413-22.

25. Fukuhara T, Gotoh M, Asari S, et al. The relationship be-tween brain surface elastance and brain reexpansion after evacuation of chronic subdural hematoma. Surg Neurol 1996;45: 570-4.

26. Tsujii M, Kawano S, Tsuji S, et al. Cyclooxygenase regu-lates angiogenesis induced by colon cancer cells. Cell 1998;93: 705-16.

27. Gallo O, Franchi A, Magnelli L, et al. Cyclooxygenase-2 pathway correlates with VEGF expression in head and neck cancer. Implications for tumor angiogenesis and metastasis. Neoplasia 2001;3: 53-61.

28. Katano H, Kamiya K, Mase M, et al. Tissue plasminogen activator in chronic subdural hematomas as a predictor of recurrence. J Neurosurg 2006;104: 79-84.