HONOLULU—New approaches to improving outcomes of traumatic brain injury (TBI) have not shown much benefit when tested during randomized clinical trials—but at least intensivists have learned not to use magnesium sulfate or high-dose steroids for neuroprotection, or albumin for fluid resuscitation, said Jamie Cooper, MD, at the Society of Critical Care Medicine's 37th Critical Care Congress.
Dr. Cooper, who is Deputy Director of the Intensive Care Unit at Alfred Hospital in Melbourne, said that major, well-designed trials of neuroprotective agents that previously had performed well in preclinical studies have shown no benefits. He attributed the negative results in human trials of TBI in part to the difficulty of balancing the many variables that may affect study results, such as heterogeneity of patients and the different types and causes of head injuries (eg, penetrating vs blunt injuries).
Timing is also a factor, and many of the interventions tested may have been started too late, suggested Dr. Cooper. “The future must lie in starting neuroprotective or prophylactic treatments prior to reaching the hospital. However, performing such a trial would be quite difficult.”
One of the difficulties for researchers would be obtaining informed consent in the prehospital setting, particularly in the ambulance, where patients might be unable to give consent themselves and where the next of kin may not be available. “We have to start looking at settings where waived consent can be assumed,” Dr. Cooper projected.
Trial designers also are trying to determine whether to include older patients in trials, which outcome measurement to use, and when to measure outcomes since TBI improves slowly over time.
Dr. Cooper described three landmark randomized controlled trials that provide important guidance on what not to do for head injury patients. These were the SAFE (Saline vs Albumin Fluid Evaluation) study, which showed that albumin fluid resuscitation produced worse outcomes than saline resuscitation; the CRASH (Corticosteroid Randomization After Significant Head Injury) study, which concluded that corticosteroids should not be used routinely in the treatment of head injury; and a randomized controlled trial of magnesium sulfate conducted by Temkin et al, which showed that patients who received magnesium infusions had worse outcomes than did controls.
In the original SAFE study, published in 2004, 6,997 ICU patients were randomized to fluid resuscitation with albumin or saline. The researchers found that there were no significant differences between the groups in new single- or multiple-organ failure, or in mean days spent in the ICU, days spent in the hospital, days of mechanical ventilation, or days on renal-replacement therapy. These findings led the FDA in 2005 to call for more evaluation of albumin in patients with TBI.
In the post hoc evaluation, the SAFE investigators reevaluated data for 231 TBI patients from the original study randomized to fluid resuscitation with albumin and 229 TBI patients randomized to fluid resuscitation with saline. At 24 months, 33.2% of patients in the albumin group versus 20.4% in the saline group had died (relative risk [RR], 1.63). Albumin resuscitation nearly doubled mortality in the subgroup of patients with severe brain injury (mortality rate, 41.8% with albumin vs 22.2% with saline; RR, 1.88). There was no demonstrable benefit from albumin in any subgroup.
“The outcome curves move apart quickly. All the action seems to have happened in those early days, perhaps in the first two weeks after injury,” Dr. Cooper said. “Most of the deaths occurred early, within 28 days. Differences in outcomes were maintained up to 24 months.”
Dr. Cooper noted that the biological mechanism behind this “rather striking finding” is unknown. He speculated that in the injured brain, albumin might cross the blood-brain barrier and not exit, thus increasing brain edema.
“This also adds to the strength of our feeling that saline resuscitation alone might be worthwhile. These results are likely to change practice in many parts of the world,” Dr. Cooper remarked.
The CRASH trial examined the effect of IV corticosteroids in 10,000 adults with clinically significant head injuries. The trial had been designed to accrue 20,000 patients but was stopped early by the data-monitoring committee after interim analysis showed significantly worse outcomes in the patients randomized to a 48-hour infusion of methylprednisolone compared with those who received placebo.
Specifically, the CRASH investigators found that at six months after injury, mortality in placebo-treated patients was 22.3%, versus 25.7% in steroid-treated patients (RR, 1.15). “This was a highly significant difference in this very large trial,” Dr. Cooper said.
“It seems clear from the CRASH trial that an agent that has been very widely used internationally clearly increased mortality after TBI—by about 3%. We can therefore improve survival of TBI simply by avoiding use of this agent,” he maintained.