Two new randomised controlled trials were released online by the New England Journal of Medicine on 10th December 2014.
Both papers examined the role of progesterone on the impact of traumatic brain injury. Both found no benefit.
Traumatic brain injury accounts for millions of hospitalisations globally each year. Severe brain injury can result in death in significant numbers, and for those who survive, gross life-changing morbidity awaits. Health care costs associated with TBI have been estimated in the hundreds of billions of dollars.
So its not surprising that all manner of interventions have been tried in this group.
Interest in progesterone as a potential mitigator of brain injury centres on its role as a neurosteroid which may limit inflammation, reduce vasogenic oedema and control apoptosis. After apparent success in laboratory and animal studies, 2 small scale, single centre studies suggested a mortality benefit compared with placebo.
The PROTECT III trial was a phase 3, double blinded, placebo controlled RCT that focussed on the early (<4 hours) IV administration of progesterone to victims of moderate to severe blunt head trauma (GCS 3-12). The study was managed by the NINDs funded network of neurotrauma research centres, including 49 trauma centres in the US.
The dose was weight based, front-loaded and extended over 72 hours by continuous infusion.
The primary outcome measure of an extended Glasgow Outcome Score at 6 months, classified as either a good or bad outcome. The definition of a good outcome as measured by GOS changed depending on the initial severity of injury, as defined by initial GCS.
The study aimed to include over1100 patients in an attempt to detect a 10% difference in favourable outcomes. It was ceased at the second planned interim analysis based on a predefined futility stopping rule. Ultimately slightly more than 800 patients were included, about half of which had "moderate to severe brain injury" (GCS 6-8).
The primary endpoint was similar in both groups, with slightly more than 50% of patients having a good outcome. Mortality was also unchanged. No apparent increase in complications of progesterone therapy were noted.
The SYNAPSE trial was a multinational, placebo controlled, blinded, randomised trial involving trauma centres in over 20 countries. It included patients with moderate-severe brain injury, and treatment was initiated within 8 hours of injury.
Again, weight based treatment was given intravenously, with front loading and completed over 120 hours.
Similarly, the primary endpoint was a functional outcome, the Glasgow Outcome Score, measured at 6 months. Extended GOS was also calculated, and unlike PROTECT 3, more refined rehabilitation measures such as the SF-36 were reported.
Like PROTECT 3, the study aimed to recruit 1200 patients to detect a 10% difference in a "good" outcome. Drug levels were monitored, and were consistent with previous trials that demonstrated efficacy.
The results were almost identical to the PROTECT 3 trial - no difference in the primary endpoint; around 50% had a bad outcome, of which 20% were deaths or vegetative.
So what's the upshot?
At least for now, there is no reliable evidence for the use of progesterone in brain injury.
These trials join more than 30 other trials on the TBI scrap heap. Together, they (yet again) raise the question, how can so many confirmatory phase 3 trials be negative, despite such encouraging work in early phases? Can we continue to invest enormous quantities of research funding in this way? Or do we need to look closer at the criteria for commencing these trials?
It also calls into question the persistent hope that a single agent or therapy will influence a complex pathological process such as brain injury. Or sepsis. Or ARDS.