This study was done in Istanbul University, Cerrahpasa Medical School, Department of Neurosurgery

Introduction

After acute head injury, neural degeneration occurs through a combination of primary and secondary mechanisms. While primary mechanical disruption of the central nervous system (CNS) parenchyma and blood vessels is obviously important, much of the neural injury is due to a cascade of secondary mechanisms by means of neurochemical and pathophysiological events set in motion by the primary mechanical insult. These secondary injury processes involve a complex interplay of multiple mechanisms. Available evidence suggests that one of the principle player in secondary injury process is prostaglandins (PG's) [₁₀]. Prostaglandins are one form of phospholipids and mainly synthesized from the arachidonic acids (AA). They are formed by the activation of a specific enzyme, namely phopholipase which splits AA from the phospholipids as a response to a variety of mechanical, chemical, and humoral factors. More than 80 % of AA are synthesized by this enzyme and one part of the remaining is metabolized by the cyclooxygenase pathway to form PG's, such as thromboxane A (TXA₂), prostocycline (PGI₂), prostaglandin E (PGE₂), and Prostaglandin F (PGF_2α) [₁,₁₅,₂₉].

Two of these, TXA₂ and PGI₂ are known to have a potent vasoactive effects on the cerebral circulation, allowing the prostanoids to play an important role in the maintenance of cerebral blood flow (CBF) [₁₃]. Furthermore, PG's are likely to contribute to the pathophysiologic consequences of head injury. Several previous studies have shown an increase in TXA₂ and PGI₂ levels in blood samples of patients with severe head injury (SHI) [₂]. Elevated free AA and various oxygenated metabolites were also demonstrated in the cerebrospinal fluid (CSF) of severe head-injured patients [²]. It is also known that TXA₂ and PGI₂ have antagonistic pharmacological activity. Previous studies have shown that, TXA₂ upon release cause vasoconstriction and platelet aggregation, whereas PGI₂ are potent inhibitors of platelet aggregation and cause relaxation of smooth muscles of vessels [₁₂]. It has been also demonstrated that the alteration in the ratio of TAX₂/PGI₂ has significant effects on the regulation of CBF [²]. Since, both TXA₂ and PGI₂ are unstable and have a short half life, the measurement of the stable metabolites, TXB₂ and 6-keto prostaglandin F_1α (6-keto PGF_1α) is required to reflect the levels of TXA₂ and PGI₂ respectively.

To our knowledge, increase in the levels of TXA₂ and PGI₂ have been demonstrated in specimens taken from the veins, arteries, and even from the jugular bulb of the patients with acute head injury [²,²⁹]. The increased concentrations of such metabolites in the ventricular CSF were first demonstrated by Westcott, et al. [₂₈]. But in this preliminary study, the ventricular CSF concentrations of prostaglandins in patients with other pathologic conditions such as meningitis, gunshot wounds etc., were also measured in addition to the patients with closed head injury and there is no available data related to posttraumatic periods.

In our study, we wished to clarify the alterations in CSF concentrations of TXA₂ and PGI₂ in SHI and tried to delineate the changes of these metabolites in the course of time after trauma. Finally, we evaluated the relationship between the results and the prognosis of the patients with SHI.

Patients and Methods

To establish the study group for this prospective study, we have evaluated each 15 consecutive severe head-injured patients who were admitted to Cerrahpasa Medical School Hospital at the University of Istanbul between 1999 and 2000. Table-1 lists the clinical and demographic data of the patients. Patients between 8 and 56 years of age who had Glosgow Coma Scale (GCS) of 3 to 8 on admission were included in this study. The two exclusion criteria were prolonged hypotension (systolic blood pressure of <90 mmHg for 30 minutes) and prolonged hypoxia (oxygen saturation of <94% for 30 minutes).

Figure 1

TABLE.1:Clinical and demographic data of the fifteen severe head-injured patients*

ASDH: acute subdural hematoma; C: contusion; DAI: diffuse axonal injury; F: female; GCS: Glasgow Coma Scale; ISS: Injury Severity Score; GOS: Glasgow Outcome Scale; ICH: intracerebral hematoma; M: male; MD: mild disability; MVA: motor vehicle accident; SAH: subarachnoid hemorrhage; SD: severe disability; Skull fx: skull fracture; y=year;

Based on the above criteria, 15 consecutive patients with SHI who admitted to the intensive care unit were enrolled in the study. The ten male and five female patients ranged in age from 15 to 65 years ( mean 30.6 years). Five patients with normal pressure hydrocephalus were made up of the control group.

After being intubated in the emergency room, all the patients immediately underwent computerized tomography (CT) scans. Each patient was managed according to the severe head injury treatment guidelines notified by Bullock, et al.[₄]. All individuals underwent intracranial pressure (ICP) monitoring via right lateral ventriculostomy. ICP was measured hourly via an external ventricular drainage system, while mean arterial pressure (MAP) was measured by invasive methods. The aim was to maintain cerebral perfusion pressure (CPP) at approximately 70 mmHg, and this was accomplished by setting targets of ICP<20 mmHg and MAP>90 mmHg.

CSF samples were drawn from an intraventricular catheter and were centrifuged immediately to remove cellular material and the CSF samples which contained gross blood were not evaluated. The stable metabolites, TXB₂ and 6-keto PGF_1α concentrations were assayed in the ventricular CSF samples between 4 and 72 hours after the insult. The patients were studied during the following four time periods posttrauma as 6 to 10, 20 to 28, 40 to 54, and 64 to 74 hours. Hourly MAP and ICP findings were also recorded in all patients throughout the study. Control CSF samples were obtained from the NPH patients while their shunt procedures were performing.

The CSF samples were all stored at –70 °C and asseyed for TXB₂ and 6-keto PGF_1α determinations, purified (C-18 reverse phase cartridge, Sep-Pak) CSF specimens were used. TXB₂ and PGF_1α concentrations were measured by the enzyme-linked immunosorbent assay (ELISA) using a kit (Cayman Chemical Co, Ann Arbor). Glasgow Outcome Score (GOS) for patients with SHI was obtained at 6 months following trauma. Statistical results were found by using ANOVA test and considered significant when the probability level was less than 0.05.

Results

TXB₂ concentrations (pg/ml): 6-10 hours: 227±6.3; 20-28 hours: 221±5; 40-56 hours: 223±5; and 64-74 hours: 234±28 (Figure-1). 6-keto PGF_1α concentrations (pg/ml): 6-10 hours: 219±83; 20-28 hours: 479±156; 40-56 hours: 335±58; and 64-74 hours: 255±45 (Figure-2). Control group: TXB₂ concentrations; 169±26.5, and 6-keto PGF_1α concentrations: 376±3. When compared to the ventricular CSF control values, CSF concentrations of TXB₂ were significantly increased in SHI patients at all times post-trauma (p<0.0001). There was a decline at the levels of 6-keto-PGF_1α between 6 to 10 hours following trauma (p<0.05). TXA₂/PGI₂ is an important ratio in the evaluation of CBF regulation, since this displayed significant alterations during posttraumatic 6-10, 40-56, and 64-74 hours (Figure-3). In addition, TXA₂/PGI₂ ratio was found to be high in patients with poor GOS.

Figure 2

Figure-1: Bar graph showing the alterations on CSF TXA2 levels during posttraumatic time periods. Note that CSF concentrations of TXB2 is markedly increased at all times posttrauma when compared with the control group.

Figure 3

Figure-2: Bar graph showing the alterations on CSF PGI2 levels over posttraumatic 74 hours. Note the decline at the level of PGI2 between 6-10 hours following trauma.

Figure 4

Figure-3: Bar graph showing the alterations on CSF TXA2/PGI2 ratios during posttraumatic 74 hours. There is significant alteration in the ratio of TXA2/PGI2 during posttraumatik 6-10, 40-56, and 64-74 hours.

Discussion

Eicosanoids and their precursor AA's are released in the brain parenchyma and in part by the vessel wall itself under pathologic conditions [₂₇]. Although the mechanisms of prostanoid production in the brain have not yet been fully understood, the possible causes of the elevation of prostaglandin level in patients following acute brain injury are the mechanical damage to the cerebral tissue and blood vessels, necrosis, the release of lysosomes and seroronin, and the accumulation of calcium ions into cells, ischemia/anoxia in the cerebral tissue, and a high concentration of catecholamine in plasma. All of these pathological sequences may stimulate phospholipase to hydrolize membrane phospholipid that releases large amount of AA and the PG's are biosynthesized from it [₅,₁₆].

It seems that primary origin of TXA₂ appears to be glial cells such as astrocytes and microglia [₁₉] but PGI₂ has been shown to be released mainly from the cerebral vascular endothelial cells [¹]. In addition, it has been shown that PGI₂ in the human central nervous system is also produced from the choroid plexus [₂₁].

In the last two decades, studies have shown that TXA₂ and PGI₂ have opposite pharmacological effects. TXA₂ constricts cerebral vasculature, whereas PGI₂ relaxes vascular smooth muscle both in vitro and in vivo [₉,¹⁰,₁₁,₁₄]. PGF_2α and PGE₂ may also affect the cerebral vasculature. PGF_2α is a potent vasoconstrictor but less powerfull than that of TXA₂, and PGE₂ constricts intracranial arteries while relaxes the external carotid artery [⁹,¹⁰,¹¹,¹⁴]. Nearly all of the previos studies indicated that the precise balance between dilating (PGE₂ and PGI₂) and constricting (TXA₂) prostanoids play an important role in the regulation of vascular resistance under pathological conditions.

There have been limited number of clinical studies that demonstrated TXA₂ and PGI₂ levels in patients with severe head injury [²,²⁸,²⁹].

In their clinical study by Yang et al. [²⁹], the prostanoid levels in the plasma of 65 patients with variable degree of head injury were evaluated. In this study they demonstrated that, TXB₂ levels in the severe injury group were significantly greater than the values for control and other mild/moderate groups on the first day of injury and 6-keto-PGF_2α also had similar results but to a lesser extend. However, beyond the third day of injury, there was no increase in the 6-keto-PGF_2α levels, but TXB₂ levels remained high. They showed the ratio of T/K remained high throughout 14 days after trauma. They also made comparison between plasma levels of TXB₂ and 6-keto-PGF_2α and the T/K ratio in outcome groups. TXB₂ was high in the death and poor outcome groups over 14 days, and all were higher than levels in the good otcome and control groups. Levels of 6-keto-PGF_2α were elevated in a similar way in both death and poor otcome groups on the first and third days. They concluded that, the more severe the injury, the greater was the elevation and the longer it lasted.

Aibiki et al. [²] measured TXB₂ and 6-keto-PGF_1α levels in the arterial and internal jugular bulb sera of 26 consecutive patients with severe traumatic brain injury. They showed that TXB₂ levels from the arterial plasma on admission were significantly high and remained so untill the 24 ^th hour of trauma. By then, the levels decreased transiently but increased again 3 days after the injury. On the other hand, there was no such elevation in the level of 6-keto-PGF_1α. The same prostaniod levels in the internal jugular bulb were also measured. Consequently, they showed arterial-jugular bulb differences were similar to the values obtained from the arterial samples.

The main question that can be drawn from these studies is that, “To what extends are the changes in plasma levels reflected in the CSF?” We think that, measurements of plasma prostanoid levels may not reflect the changes in the levels of the brain accurately. Thus, as a complamentary to all these studies, we measured the ventricular CSF levels of TXB₂ and 6-keto-PGF_1α in patients with SHI.

In the majority of studies, both experimental and clinical, the concentrations of prostaglandins, particularly TXB₂ and 6-ketoPGF_1α, have been measured in the samples taken from the peripheral vessels. To our knowledge, since there is little in the literature about PG levels in the CSF of the patients with TBI, in this study we measured the CSF levels of TXB₂ and PGF_1α in 15 patients with severe head injury at the 6-10, 20-28, 40-56, and 64-74 hours after injury. The study showed that TXB₂ levels were increased and remained high throughout 3 days after trauma in all patients with injury compared with control group. On the other hand, the levels of 6-keto-PGF_1α showed diversity in values. It tended to decrease within the first hours after trauma and at the 24 ^th hour was significantly increased compared with control group, by then tended to decrease again significantly. The T/K ratio was significantly elevated compared with control within the first hours and 24 ^th hour after trauma.

In another clinical study proposed by Westcott et al. [²⁸] prostaglandin concentrations were measured in the ventricular CSF of pediatric patients with head injury. They also found large increase in arachidonate cyclooxygenase products in CSF, with 6-keto-PGF_1α and TXB₂ typically reaching the highest level. However, this study did not show posttraumatic time interval in which CSF samples were taken.

For TXA₂, our results are consistent with the results of clinical studies mentioned but not for PGI₂, because the decrease of PGI₂ levels was noted during the first hours after trauma. All these studies including the present one show that, in human CSF TXA₂ is increased after traumatic brain injury, but PGI₂ levels does not always increase more than TXA₂ even after the injury, resulting in differences between the two prostanoids. Such differences may play an important role in hypoperfusion in the damaged area, possibly mediated through vasoconstriction caused by excessive TXA₂ production.

Both TXA₂ and PGI₂ are originated from the same precursors. But, it is probably that damaged epithelial cells of the cerebral vessels loss its ability to produce PGs after trauma. Consequently, as in our study, PGI₂ is produced less than normal [²,²⁸,²⁹]. Therefore, this imbalance may be the primary reason of a marked disturbance of the ratio of TXA₂/PGI₂ which appears to be an important indicator of secondary brain damage.

Recent experimental [₇,₈,₂₄] and clinical studies have shown that the precise balance between prostanoids with vasoconstricting and with vasodilating features on the cerebral vasculature, as reflected by the ratio of TXA₂/PGI₂, becomes more important in the regulation of vascular resistance in case of pathological conditions [²,¹³]. It has been discussed that the development of vasospasm after subarachnoid hemorrhage is at least in part mediated by the reduced release of PGI₂ and by an increased formation of leukotrienes and TXA₂ [₂₅,₂₆]. In the literature, TXB₂, PGF_2α, and PGE₂ were shown to be a strongly vasoconstrictive and can cause arterial diameter to reduce, and finally spasm which in turn result in the reduction of cerebral blood flow and oxygen availability [₂₀].

Therefore, imbalance of TXA₂/PGI₂ ratio reduces local cerebral blood flow and promotes the formation of minute thrombi, while TXA₂ is produced from such thrombi, at the same time damaged epithelial cells produce PGI₂ less than normal amount which results in loss of vasodilation. All these pathological events may aggravate the ischemia and hypoxia of injured tissue and worsen the patient's condition [₃,₆,₁₇,₁₈, ₂₂, ²⁶]. Either the present clinical study or the clinical study of Yang et al. [²⁹]. showed that, in case of high TXA₂/PGI₂ ratio, there is almost always worse glasgow outcome score.

In conclusion, the results from a limited number of the patients show that concentrations of TXA₂ in the CSF in severe head injury rise following trauma, however PGI₂ levels demonstrates a tendency for a decline. Furthermore, as an important ratio in the evaluation of CBF regulation, TXA₂/PGI₂ showed significant alterations during 6-10, 40-56, and 64-74 hours. We also demonstrated that patients with poor GOS scores were found to have higher TXA₂/PGI₂ ratio. Therefore, all these changes may be an important cause of poor outcome in patient with SHI.

Acknowledgments

Correspondence to

Mustafa Uzan, MD P.O. Box: 5, Cerrahpasa 34301, Istanbul/TURKEY Tel: +90 212 632 00 26 Fax: +90 212 632 00 26 e-mail: tanerato2000@yahoo.com