The use of anti-inflammatory drugs to treat severe head injuries is actually dangerous

The use of anti-inflammatory drugs to treat patients with severe head injuries—common practice worldwide for the past 30 years—is actually dangerous and associated with around a 20% increase in death within two weeks of hospital admission, conclude authors of an international study in this week’s issue of THE LANCET.

Around 3 million people worldwide die of trauma every year, many after arrival at hospital. Findings of a 1997 systematic review suggested that these drugs might reduce the risk of death by a few percent; previous individual trials had been too small to provide a clear-cut answer. The CRASH trial (corticosteroid randomisation after significant head injury)—a multicentre international collaboration—aimed to confirm or refute such an effect by recruiting 20,000 patients. In May, 2004, the data monitoring committee disclosed the unmasked results to the steering committee, which stopped recruitment to the trial.

10,008 adults (from 239 hospitals in 49 countries) with head injury were randomly allocated corticosteroids (methylprednisolone) or placebo for 48 hours after admission to a hospital emergency department. Those patients given corticosteroids had a higher death rate within 2 weeks (21% of patients) than those given placebo (18%). The 6-month follow-up data will be presented in a subsequent paper.

Clinical co-ordinator of the study Professor Ian Roberts (London school of Hygiene and Tropical Medicine, UK) comments: “Our early results show that corticosteroids should not be used routinely to treat head injury, whatever the severity. By clearly refuting a mortality benefit from corticosteroids in head injury, the CRASH trial results should protect many thousands of patients from any increased risk of death associated with these drugs... The effect of corticosteroid treatment on disability 6 months after head injury will be reported as soon as these data are available. Many other treatments of uncertain effectiveness for head injury are in widespread use, and further large-scale randomised trials are needed. The CRASH trial has shown that we can enrol many trauma patients into clinical trials in the emergency setting”.

In an accompanying commentary (p 1291), Stefan Sauerland (University of Cologne, Germany) concludes: “CRASH partly shakes our pathophysiological understanding of what is of primary importance after traumatic brain injury. The role of inflammatory mediators and the presumed effects of blocking their release must now be elucidated in more detail. Other mechanisms leading to increased intracranial pressure must also be targeted. The key message of CRASH, however, is that applying treatments with unproven effectiveness is like flying blindly. In future, we should avoid trusting in underpowered clinical trials with surrogate rather than clinical endpoints, and transferring evidence from one disease to another”.