Keywords

Introduction

The unique nature of brain tissue poses considerable challenges for neurosurgery. Under the rigid and tough protection of the skull and meninges, the brain is extremely vulnerable to the mechanical insults produced by neurosurgical maneuvers, such as direct incisions, electrocauterization, and retraction. Healthy tissues at the margins of the operative target are inevitably subjected to this surgical brain injury (SBI). Brain edema and hemorrhage are major ­complications that develop following SBI in neurosurgeries [3]. The blood–brain barrier (BBB), comprising the vascular endothelium, pericyte, and astrocytic processes, prevents the leakage of plasma proteins from the vascular bed into the brain tissue [25]. Vasogenic edema results from the passage of water along with the plasma proteins into the brain tissue because of damage to the capillary endothelium and the interendothelial tight junctions of the BBB, whereas cellular swelling of the injured brain cells results in cytotoxic edema. Both result in increased intracranial pressure, which may lead to further brain injury from cell death or hypoperfusion [3]. Currently, SBI is clinically addressed by nonspecific postoperative care; however, it has become possible to study potential therapies in the preclinical laboratory setting with the recent development of a rodent model for SBI. This review discusses the SBI rodent model and various molecules implicated in the pathogenesis of brain edema as well as treatments that have been applied to reduce brain edema in this model.

Surgical Brain Injury Animal Model

An in vivo rodent model was recently developed to study SBI pathophysiology as well as potential therapeutic targets [14, 16, 23]. This model was designed to mimic the injuries sustained from neurosurgical manipulation of brain tissue. The rodent brain is exposed through a small cranial window through which tissue resection is performed. The margins of the resection are designated in relation to the bregma, as shown in Fig. 1. The model provides consistently measureable brain edema using the brain water content (Fig. 2) in the perisurgical site at 24 h. Studies have shown that brain edema in the perisurgical site peaked 24 h after inducing SBI, was significantly higher up to day 3 post-injury, and had subsided by day 7; SBI was associated with neurobehavior deficits that had dissipated by day 7 [16, 23, 34]. Brain edema was assessed using brain water content measurement, Evans blue dye, and IgG extravasation, and by measuring the apparent diffusion coefficient (ADC) values (Table 1).

Fig.  1
figure 00231

Partial right frontal lobectomy. Two incisions (dashed lines) are made leading away from the bregma (white X) along the sagittal and coronal planes, 2 mm lateral to sagittal and 1 mm rostral to coronal respectively

Full size image
Fig.  2
figure 00232

Twenty-four-hour brain water content. The figure shows marked edema in the right frontal region (bordering the surgical injury) of the brain compared with the sham control group 24 h after the surgical brain injury (SBI). The other brain regions did not differ statistically from each other after the injury. n  =  6. p  <  0.05 for * versus sham. Data are expressed as mean  ±  SEM. Statistical significance was verified using Student’s t test

Full size image
Table 1 Experimental rodent studies of therapeutic agents for surgical brain injury
Full size table

Vascular Endothelial Growth Factor

Vascular endothelial growth factors (VEGFs) are a large family of proteins designated VEGF-A through VEGF-E and are expressed in the choroid plexus and neurons in the normal brain [24]. A well-known growth factor with angiogenic, mitogenic, and permeability-inducing effects, VEGF has also been shown to contribute to SBI-induced brain edema. VEGF-A has a potent hyperpermeability-inducing effect on the microvascular endothelium that is mediated through its receptor VEGF receptor-2 (VEGF-2, kdr) [24]. VEGF-2, a transmembrane tyrosine kinase present in the endothelium of the brain vessels, activates MAPK signaling [19]. An upregulation in VEGF has also been reported in models of TBI predominantly due to the infiltrating neutrophils [5]. VEGF administration and overexpression with viral vectors have both been associated with compromised integrity of the BBB in the rodent brain [29]. A potential mechanism by which VEGF-A might increase BBB permeability involves downregulating the expression of occludin, a tight junction protein; this would disrupt the organization of occludin and ZO-1, another junctional protein, leading to tight junction disassembly [25, 28]. Although VEGF may have potentially reparative actions during later phases after an injury, early inhibition of VEGF or its upstream mediator Src tyrosine kinase have been shown to reduce brain edema in various stroke models [16, 19, 24].

The expression of VEGF was increased at 24 h after SBI in the ipsilateral frontal region surrounding the surgical resection site [16]. The Src family is implicated in VEGF-dependent hyperpermeability [25] and has been shown to be involved in SBI-induced BBB disruption and subsequent brain edema in rodents [16]. An increase in expression of VEGF and p-ERK 1/2 as well as a corresponding decrease in the tight junction protein ZO-1 were reversed when rats were pretreated with PP1, an Src tyrosine kinase inhibitor, prior to inducing SBI [16]. In addition, VEGF was shown to play a possible role in the erythropoietin-induced increase in brain edema during the early phase in a SBI rodent model. Erythropoietin administration led to significantly increased brain water content in the perisurgical region at 24 h and was associated with increased expression of VEGF [23]. Thus, VEGF-A and its upstream Src tyrosine kinase present potential therapeutic targets for preserving the BBB and reducing brain edema following neurosurgical procedures.

VEGF-B, another member of the VEGF family expressed in the CNS, mediates its effects through the VEGF receptor-1 (VEGF-1, flt-1); VEGF-B is likely responsible for maintaining the BBB in a steady state [25]. Furthermore, when bound to the soluble extracellular portion of VEGFR-1, VEGF is inactive, sequestered, and unable to bind to VEGF receptors [24]. The role of VEGF-B and its receptors in brain edema development following SBI remains to be elucidated.

Matrix Metalloproteinases

Matrix metalloproteinases (MMPs), zinc-dependent endopeptidases involved in tissue remodeling and repair, have been implicated in the SBI-induced destruction of the extracellular matrix proteins of the neurovascular unit. The target substrates of MMPs include collagen IV, fibronectin, and laminin, all of which are critical to maintaining the integrity of the BBB [28]. MMP-2 (gelatinase A) and MMP-9 (gelatinase B) degrade the basal lamina and tight junction proteins of the BBB and promote BBB injury that leads to vasogenic edema during the acute stage in experimental models of brain injury and stroke [9, 28]. Upregulation of MMPs has been demonstrated following subarachnoid hemorrhage, cerebral ischemia, traumatic brain injury (TBI), and intracerebral hemorrhage, and has been shown to contribute to BBB disruption during the early stage of injury [8, 30, 31, 33]. Additionally, MMP-9 knockout mice had improved functional outcomes, lowered BBB permeability, and reduced lesion volume after transient focal cerebral ischemia and TBI [1, 32].

Similarly, the role of MMPs in mediating BBB disruption and brain edema after SBI was demonstrated by Yamaguchi and colleagues [34]. An upregulation of MMP-2 and MMP-9 was temporally associated with BBB disruption after SBI in rats. A significant increase in MMP activity, particularly that of MMP-9, compared with presurgery levels was detected at days 1–14 after SBI, with highest MMP activities observed at days 1 and 3 coinciding with peak values in brain edema; MMP inhibitor, reversed these effects, preserving the BBB integrity and reducing SBI-induced brain edema as early as 3 h post-injury [34]. Evidence of the role of MMPs in the development of brain edema in the early stages after stroke and in SBI models supports the potential use of MMP inhibition to prevent brain edema following neurosurgical procedures.

Cyclooxygenase-2

Various inflammatory mediators have been implicated in BBB disruption and brain edema after stroke and brain injury [25, 27]. The role of cyclooxygenases, enzymes that catalyze the conversion of arachidonic acid to prostaglandins and thromboxanes [6], in mediating brain edema development has been extensively studied. The upregulation of cyclooxygenase-2 (COX-2), expressed in various cell types including neurons, astrocytes, endothelial cells, macrophages, and microglia in the CNS, has been demonstrated following focal and global cerebral ischemia, neonatal hypoxia–ischemia, and intracerebral hemorrhage [4, 6, 10, 11]. Selective inhibition of COX-2 provided protection after ICH by reducing prostaglandin E2 production, thereby decreasing inflammation, brain edema, and cell death, which translated into improved functional recovery [6]. Further, the role of COX-2 in mediating preconditioning-induced protection has been suggested. COX-2 has been shown to mediate ischemic preconditioning in vitro [18]. Studies from our laboratory have determined COX-2 to be a potential mediator of hyperbaric oxygen preconditioning (HBO-PC) in SBI and global ischemia rodent models; animals preconditioned for 1 h daily for 5 days with HBO prior to inducing injury had significantly improved neurological function and brain edema [4, 17].

Cyclooxygenase-2 has been shown to play a part in SBI pathophysiology as well. An increase in COX-2 expression was detected 24 h after SBI in mice, and HBO-PC significantly attenuated the increase in COX-2 possibly through suppression of HIF-1α, the upstream regulator of COX-2 [17, 26]. The study showed that HBO-PC increased COX-2 level twofold in comparison to the four-fold increase by the SBI, which suggested that protection conferred by HBO-PC might have involved brain preconditioning by increasing COX-2 to ­subinjurious levels. Furthermore, the protective effects of HBO-PC were reversed in the presence of selective COX-2 inhibitor. These studies demonstrate HBO-PC or COX-2 inhibition to be promising therapies in attenuating brain edema following neurosurgical procedures.

Conclusion

The SBI rodent model mimics injuries caused by neurosurgical procedures and produces postoperative brain edema. It allows for the study of the cellular signaling pathways and the identification of key molecular targets for neuroprotective pretreatment before neurosurgical intervention. To date, VEGF, MMPs, and COXs have been revealed to be potential targets for therapy. As clinically applicable therapeutic intervention for SBI is likely to result in significant benefits for patients and healthcare organizations, further preclinical and clinical studies are necessary to explore the applicability of these targets.