Steroids Improve Outcome in Spinal Cord Injury


Synopsis: Administration of methylprednisolone for 24 or 48 hours, depending on how soon after the event it is started, improves functional status at six weeks and six months in patients with acute spinal cord injury.

Source: Bracken MB, et al. JAMA 1997;277: 1597-1604.

This randomized, controlled trial, carried out in 16 centers over a four-year period, included 499 patients with acute spinal cord injury. Patients received an intravenous dose of methylprednisolone (30 mg/kg) as soon after initial injury as possible, and were then randomized to receive either 24 or 48 hours of the same drug (5.4 mg/kg/h) or tirilazad mesylate (a potent inhibitor of central nervous system lipid peroxidation), 2.5 mg/kg intravenously every six hours for 48 hours. The study’s outcome measurements were assessments of motor function change between initial presentation and at six weeks and six months after injury, and change in Functional Independence Measure (FIM) after the same intervals.

The study design excluded patients with serious comorbidity, gunshot wounds, or "specific health conditions that might affect treatment assessment." As a result, the 499 patients were mainly previously healthy white males aged 14-34 years who had been injured in motor vehicle accidents or falls. Most were admitted directly to an acute spinal cord injury center. Recovery was complicated by pneumonia in approximately 15% of patients, and by sepsis and the acute respiratory distress syndrome in about 4% of the patients. Overall, six-month mortality was about 6%.

Patients starting treatment within three hours of injury showed essentially the same recovery pattern in all treatment regimens. For those in whom treatment commenced between three and eight hours after injury, receiving steroids for 48 hours was associated with greater improvement after six weeks and six months, although the differences were relatively small. Results following administration of tirilazad were indistinguishable from those in the 24-hour steroid group. Differences in complications among the three treatment groups were also relatively small; however, severe pneumonia occurred statistically less often in the tirilazad group than in the two steroid treatment groups (0.6% vs 2.6% [24 hr group] and 5.8% [48 hr group]; P = 0.02), and there was a similar but nonsignificant trend in the incidence of severe sepsis.

Bracken and colleagues conclude that patients with acute spinal cord injury who receive methylprednisolone within three hours of injury should be maintained on the treatment regimen for 24 hours, and when steroid treatment is begun three to eight hours after injury, patients should receive the drug for a total of 48 hours.


Patients with acute spinal cord injury are commonly cared for in intensive care units, particularly in institutions that serve as trauma centers. The most common etiologies in the United States are motor vehicle accidents, falls, and gunshots. Therapeutic options for most patients, particularly those who have suffered complete functional cord transection, are frustratingly limited, and even a modest return of neurological function with therapy is a most desirable goal.

Studies on experimental models of acute spinal cord injury have shown that a single bolus dose of corticosteroids does not ameliorate the neurologic damage. This study adds to the available evidence in humans that these drugs can have a favorable effect on outcomes in spinal cord injury when given promptly after the event and continued for at least 24 hours. When treatment was begun immediately (that is, within 3 hours of the injury), there was no difference in outcomes between the two steroid treatment regimens studied. However, continuing methylprednisolone therapy for a second 24 hours was associated with additional long-term improvement if the initiation of therapy occurred between three and eight hours. The theoretically promising drug tirilazad did not prove to be as effective as the 48-hour steroid regimen, although as expected (since it has no glucocorticoid immunosuppressive effects) it was associated with fewer serious infectious complications. Whether any regimen could produce improvement when therapy started more than eight hours after injury was not addressed in this study.

By design, this study included only patients with injury limited to the spinal cord lesion and excluded patients with complicating medical diseases. In addition, patients with gunshot wounds, an all-too-common cause for spinal cord injury in the United States, were excluded. The degree to which these selection factors limit the generalizability of the study’s findings to all ICU patients with acute spinal cord injury are unclear. However, the significantly higher rate of serious pneumonia and severe sepsis in the two steroid-treated groups in comparison with tirilazad-treated patients is a cause for concern in this respect. Critically ill patients with multiple trauma including spinal cord injury are already at greatly increased risk for these complications in comparison to less seriously ill individuals, and the risk of further increasing this risk with 48 hours of high-dose steroid therapy cannot be discounted.

This study was a huge project involving dozens of clinicians in 16 major medical centers who screened nearly 800 patients for inclusion. Conducting further studies with the power to answer the questions raised by these results would be an even more formidable task, and yet further clarifying the relative risks and benefits of steroid treatment in patients with acute spinal cord injury would require just such an undertaking.