Inflammation and edema is poorly tolerated by the central nervous system (CNS) because it can result in secondary injury through the release of local and systemic cytotoxic inflammatory mediators [1]. Secondary injury can be the sequela of traumatic, infectious, auto-immune, or neoplastic processes in the CNS. Steroids such as dexamethasone and methylprednisolone have anti-inflammatory properties that consequently may be beneficial adjuncts for patients with these diseases. A recent meta-analysis and Cochrane review concluded that both adults and children have reduced neurologic morbidity and mortality when steroids were used as adjuncts in bacterial meningitis [2, 3]. Steroids have been used in the treatment of bacterial meningitis to reduce secondary injury for over 60 years; therefore, many authors have also extrapolated this modality for treatment of traumatic brain and spinal cord injuries [4].

Traumatic brain injury (TBI) is a leading cause of premature death and disability in the world with the majority of TBIs resulting from road injuries [5, 6]. As road related TBIs decrease in industrialized nations, a much steeper rise has afflicted the developing world with the increasing popularity of motorized vehicles [7, 8]. In the United States the estimated incidence of TBI related disability is 33 cases per 100,000 persons per year [9]. In the developing world, this number is thought to be significantly higher with an estimated incidence of over 200 cases per 100,000 persons per year [10]. Even this number is thought to be underestimated as motorized vehicles are on the rise in Asia. TBI has become a critical public health and socioeconomic problem with the majority of TBI cases affecting younger persons resulting in early death or long-term disability [9]. Various attempts have been made to identify interventions that may provide even a moderate reduction in the morbidity and mortality associated with TBI.

Steroids were investigated in the 1960s as a therapeutic option for reducing cerebral edema after Galicich et al. reported significant improvements in patients suffering from increased intracranial pressure secondary to brain tumors or from post-operative swelling [11]. Additional experimental evidence has accrued since then confirming and elucidating the mechanisms through which steroids reduce vascular permeability and free radical production in cerebral edema [12–15]. Although the clinical benefit of steroid use in reducing vasogenic cerebral edema has long been established, it was not until years later when authors first reported the potential benefit of steroids for cytotoxic edema in the setting of TBI [14–18].

In 1976 Faupel et al. conducted a prospective double-blind review comparing dexamethasone to placebo in 95 patients suffering from severe TBI [19]. In their review, the authors found a significant decrease in mortality in the treatment group but noted no overall improvement in outcomes as the surviving patients in the control group were more likely to be severely disabled or in a persistent vegetative state [19]. A subsequent prospective double-blind randomized clinical trial comparing low dose and high-dose methylprednisolone to placebo was conducted in 88 patients with severe TBI by Giannotta et al. in 1984 [20]. While the authors found no significant difference in 6 month outcomes in the treatment groups, subgroup analysis showed improved survival and speech function in patients under the age of 40 with high-dose methylprednisolone [20].

While the early reports were promising, multiple subsequent studies were conducted and none showed substantial benefit [20–24]. Cooper et al. performed a prospective double-blind study involving 97 patients with severe TBI treated with high-dose dexamethasone or placebo. The authors found no significant difference in outcomes, serial neurological examinations, or intracranial pressure at 6 month follow-up [23]. Gaab et al. performed a similar study in 1994 with high-dose steroids given early after injury (within 3 h) and found no significant difference after 12 month follow up [25]. Low dose methylprednisolone was also evaluated in a similar fashion and similarly showed no significant difference in outcome after 6 months follow up [21].

Despite these findings, in 1995 a survey was conducted regarding the management of patients with TBI indicating that steroids were being used in 64 % of trauma centers in the United States [26]. This was likely because the results of the National Acute Spinal Cord Injury Study (NASCIS) were published in the early 1990s concluding that treatment with extremely high doses of methylprednisolone is indicated for acute spinal cord injury (this recommendation has since been discouraged by the CNS/AANS guidelines) [27, 28].

In 1997 Alderson et al. performed a systematic review of all randomized control trials evaluating steroids for the treatment of TBI. The result of their extensive review was a lack of benefit for the use of steroids in the treatment of TBI. Although the results were convincing, the authors maintained that the lack of benefit was uncertain and emphasized the need for a larger randomized double-blind clinical trial [19].

In 2004 the results of such a trial was published. The Corticosteroid Randomization After Significant Head Injury (CRASH) trial was an international randomized double-blind study evaluating the effects of methylprednisolone for the treatment of TBI [29]. This study involved the enrollment of over 10,000 patients from 239 hospitals in 49 countries. Each patient presenting to the hospital within 8 h of injury with a GCS of 14 or less was randomized to receive methylprednisolone for 48 h or placebo. The protocol of this study was similar to the NASCIS trial in order to capture any potential benefit from early intervention. The study was concluded early by the data monitoring committee after over 5 years of enrollment when an interim analysis showed not only no benefit but a higher 2-week mortality in the steroid group [29]. While the exact mechanism is uncertain, the increase in mortality seemed to be related to additional extra cranial injuries.

One such extra cranial injury associated with TBI is acute spinal cord injury (SCI) . Just as cerebral edema occurs after TBI, spinal cord edema occurs with both primary and secondary insults to the spinal cord in an acute spinal cord injury (SCI). In fact, much of the excitement regarding the use of steroids in TBI has stemmed from the data published from studies regarding the use of steroids in SCI. Despite more recent evidence suggesting the lack of efficacy of steroids in TBI, the data regarding the use of steroids in SCI has not been so clear-cut.

Spinal cord injuries may occur from a variety of high-energy accidents, predominantly affecting the young adult patient population. Traumatic SCI incidence is approximately 40 new cases per million population, with over half of these patients younger than 30 years of age [30]. The WHO reports as many as 40–80 new cases per million per year worldwide. The etiology of SCI in developed nations is predominantly seen after motor vehicle accidents. In countries with lower GDPs, falls are the predominant cause of SCI, with motor vehicle accidents less common [31]. The high level of morbidity leading to long-term disability, combined with the limitations of current medical and surgical treatment, makes prevention of SCI the most effective management. Despite having many recent advances in the prehospital transfer of patients at risk for SCI, little progress has been made in the treatment of SCI once a neurologic injury has occurred.

Neurological injury after acute SCI was found in animal studies to occur as a result of cell membrane breakdown at the site of trauma, peaking by the 8 h post-injury time period. Methylprednisolone has been shown in these same animal model studies to inhibit lipid peroxidation and hydrolysis, thus limiting the breakdown of the phospholipid bilayer after SCI. As lipid peroxidation is believed to cause further vasoactive processes within the metabolism of arachidonic acid, prevention of this pathway was theorized to improve vascularity of the injured spinal cord [32].

Standard practice in the mid-late 1900s included the administration of steroids to patients presenting with evidence of acute spinal cord injury, a practice based predominately on animal models. A national study (National Acute Spinal Cord Injury Study, NASCIS) was undertaken in the early 1980s to analyze outcomes in patients receiving steroids following acute SCI [33]. This was the first of several multicentered, double-blinded randomized controlled trials to review steroid use following SCI.

NASCIS I randomized 330 patients into 2 groups based on varying dosages of methylprednisolone. Patients were eligible for randomization if presentation was consistent with an acute traumatic spinal cord injury as assessed by an attending neurosurgeon within 48 h of injury. The high-dose group received 1000 mg methylprednisolone bolus on admission, followed by 250 mg every 6 h for 10 days; the standard dose group received 100 mg bolus on admission, followed by 25 mg every 6 h for 10 days. Detailed neurological exams were performed and documented on admission, at 6 weeks, 6 months, and 1 year. Examination was directed for outcome assessments of motor function, response to pinprick and light touch, with each graded in both an expanded score and a 5-point scale. Results were reported based on right side neurological exam only. Analysis of follow up scores at both 6 weeks and 6 months revealed improvements in all aspects of the neurological exam, without statistically significant difference between the two groups. In regard to morbidity and mortality, wound infection was found to occur 3.6-fold more often in patients in the high-dose group, with p = 0.01. Although not reaching statistical significance, mortality was three times more common at 14 days and twice as common within 28 days in the high-dose group. Given the increased risk and lack of benefit, the NASCIS I trial was closed early [33].

The second national trial regarding treatment of acute spinal cord injury patients investigated the dose of steroids, as well as the use of naloxone. Animal models had revealed that the dose of methylprednisolone used in the original NASCIS study may have been subtherapeutic, leading NASCIS II to utilize a much higher dose but only for a 24 h period. The addition of naloxone therapy was also included in this trial based on positive results in animal models suggesting neurological benefit to treatment with naloxone therapy. The study performed randomization on 487 patients by the 12 h post-injury mark, where patients received either methylprednisolone + placebo for naloxone, naloxone + placebo for methylprednisolone, or placebo for naloxone + placebo for methylprednisolone, in a double-blinded fashion. Patients were again assessed at admission, 6 weeks and 6 months for pinprick, light touch, and motor function [32].

Overall evaluation of patients receiving methylprednisolone revealed initial improvement in regard to sensory function over the placebo and naloxone groups. In addition, when the methylprednisolone group was stratified for patients receiving treatment within 8 h post-injury, there were statistically significant improvements seen in regard to both sensory and motor function at the 6 month assessment. By the one-year mark however, these sensory gains equilibrated and were no longer significant. Again, the methylprednisolone group was seen to suffer a higher rate of infection and wound healing complication, but this did not reach statistical significance [32].