• A I R Maas, A Marmarou, G D Murray, and E W Steyerberg.
• Department of Neurosurgery, Erasmus University Medical Center Rotterdam, Rotterdam, The Netherlands. airmaas@erasmusmc.nl
• Acta Neurochir. Suppl. 2004 Jan 1;89:113-8.

AbstractOver the past decade many neuroprotective agents have been developed with the hope of being able to improve outcome in patients with traumatic brain injury. Unfortunately, none of the phase III trials performed have convincingly demonstrated efficacy in the overall population. A common misconception is that consequently these agents are ineffective. Such has not been proven and some trials show evidence of efficacy in subgroups of the population studied. The negative results, as reported in the overall population, may in part be caused by specific aspects of the TBI population, as well as by aspects of clinical trial design and analysis. Clinical trials in TBI pose several complicated design issues. Methodological challenges relate particularly to heterogeneity of the population and to outcome assessment. Heterogeneity pertains both to the range of pathologies included in TBI, and to prognostic factors, each causing specific problems. Mechanistic and/or prognostic targeting, as well as possibilities for covariate adjustment, are suggested as possible solutions to deal with the problems of heterogeneity. The aim in most trials was to demonstrate a 10% absolute improvement in favorable outcome in patients with head injury. This may be considered overoptimistic and unrealistic in relation to the heterogeneous patient population. Specific problems are further incurred by the use of the dichotomized Glasgow Outcome Scale as primary outcome measure. Optimal statistical power may expected to be present when the point of dichotomization results in a 50:50 distribution of outcome categories. It is proposed to differentiate the point of dichotomization according to prognostic risk profile, in order to maintain statistical power. Solutions described may be expected to enhance chances of demonstrating benefit of potentially effective neuroprotective agents in future studies. The complexity of problems occurring in clinical trial design and analysis in TBI is such that a strong and sustained multidisciplinary input and effort is required from all experts involved in the field of neurotrauma.

15 keywords...

hide…