21 

Are Steroids Indicated in the Treatment of Head Injury?

John L.D. Atkinson

BRIEF ANSWER

Background

The chemical synthesis of cortisone by Nobel laureate Dr. Edward Kendall¹ at the Mayo Clinic led to its rapid dissemination into multiple and diverse medical fields. The results were often astounding. Initially used to treat rheumatoid arthritis, cortisone was subsequently adopted by neurologic surgery for perioperative management during pituitary surgery in the 1950s. It markedly altered the morbidity and mortality associated with hypopituitary states. Its use also made adenohypophysectomy a viable neurosurgical treatment option for endocrinologically driven malignancies that produced severe pain from widespread metastatic disease. In the early 1960s, cortisone proved remarkably successful as a treatment for peritumoral brain edema. For several decades, glucocorticoids remained the single most utilized pharmacologic agent in the medical management of neurosurgical patients. The often remarkable perceived or actual patient improvement that steroids effected in a variety of neurosurgical diseases resulted in a deep entrenchment of their universal use by a whole generation of neurosurgeons.

Chapter 12 of the original Guidelines for the Management of Severe Head Injury summarized the literature of steroid use in traumatic brain injury (TBI) up to 1995.² Based on the strength of the data, the authors concluded that a standard of care exists: the use of corticosteroids is not recommended for improving outcome or reducing intracranial pressure in patients with severe head injury.

However, entrenched medical practice is not so easily defeated, and underlying questions remain. Specifically, if steroids are reported to be efficacious for spinal cord injury, then shouldn’t they also be efficacious for head injury?

Literature Review

The following is a laboratory and clinical review of the literature since the publication of the Guidelines for the Management of Severe Head Injury.

Administration of high-dose steroids after severe head injury in rats reduces lipid peroxidation, which is also reported to be a major mechanism underlying the efficacy of high-dose steroids in spinal cord injury.^3,4 However, glucocorticoids do not appear to be beneficial in the treatment of cerebral insults involving glutamate toxicity.⁵

Assumptions about clinical benefit of steroids in TBI are all based on laboratory and clinical studies [conducted after the first National Acute Spinal Cord Injury Study (NASCIS 1)] that suggested the existence of a therapeutic window for high-dose steroid therapy. To review, the NASCIS 1 trial suggested no clear benefit of higher-dose over lower-dose peri-injury glucocorticoids (level I), and early fatality and wound infection were more prevalent in the higher-dose arm of the trial.6 However, subsequent laboratory evidence strongly suggested that a pharmacologic window exists for remarkably high doses of methylprednisolone.⁴ The subsequent NASCIS 2 trial became a benchmark in the medical management of acute spinal cord injury, with purported neuroprotective effects occurring by a variety of possible mechanisms (class II data).⁷ This led to the NASCIS 3 trial, which suggested mild additional benefit in select time frames (class II data).⁸ It is of note, however, that a retrospective review of glucocorticoid use in the setting of gunshot wounds to the spine revealed no benefit and a higher complication rate, similar to the NASCIS 1 results (class III data).⁹

One might think that the encouraging results of the NASCIS 2 trial could logically be applied to the treatment of severe head injury. However, there are differences in pathophysiology between these two polar regions of the central nervous system. Severe closed head injury is frequently associated with a variable degree of brainstem-derived apnea and catecholamine surge, affecting oxygen and carbon dioxide levels and pH, with potentially significant change in the cellular milieu over and above the contusions, lacerations, and hematomas that both spinal cord and brain may acquire by primary injury mechanisms.^10,11 Despite these potential differences in secondary injury mechanisms, a prospective, controlled, multicenter trial suggested clinical benefit from peri-injury use of glucocorticoids after head injury based on a retrospective subgroup analysis (class III data).¹² However, a subsequent systematic review of randomized controlled trials reveals considerable uncertainty as to whether peri-injury steroid use has any benefit, not to mention harmful effects.¹³ The most informative multicenter trial utilized tirilazad mesylate, a 21-aminosteroid derivative specifically designed to be a powerful antioxidant, which is reportedly the most significant mechanism by which glucocorticoids improve posttraumatic neuronal salvage. Overall efficacy in patients with moderate and severe head injury could not be demonstrated, but there were some problems inherent in the trial (class II data).¹⁴

In conjunction with this negative trial, serious questions regarding the validity of the NASCIS 2 results have been raised. An in-depth review of the trial suggests that methylprednisolone use in acute spinal cord injury is not a proven standard of care; in fact, it should not even be considered a recommended treatment.¹⁵ The uncertainty of the literature as to any benefit, a recent designer drug trial proving no benefit, and questions regarding the validity of the NASCIS 2 spinal cord injury results support not using glucocorticosteroids in the management of severe head injury. Despite these reasons, another randomized trial of glucocorticoid use in head injury is currently under way.¹⁶