Lessons from epidemiologic studies in clinical trials of traumatic brain injury.

Lessions from epidemiological studies. The Clinical Trial Group for Neurosurgery of the University of California San Diego (UCSD) is involved in epidemiological studies and trials of new pharmacological agents in traumatic brain injury. A great number (> 10,000) of patients has been prospectively analyzed forming an integrated database for further purposes. The development of these databases is based on earlier work by the European Neurosurgeons Jennett and Braakmann and the US-Traumatic Coma Data Bank Study. These studies allowed for the development of sophisticated data collection instruments which were used in the international Tirilizad Trials which enrolled over 1,100 patients. A major observation from that trial was that pretreatment hypotension or hypoxia could be unbalanced even in a large two arm blinded study. Another issue of the tirilazad trial was the influence of gender affecting not only outcome but also drug metabolism. Similar experiences were gathered with the phase-III trial on the competitive NMDA-receptor antagonist selfotel, which interferes with the excitotoxic amino acid glutamate as mediator of secondary brain damage, as ischemia-induced neuronal degeneration. Unfortunately, the trial, already underway, had to be prematurely aborted, since concurrent stroke studies with enrollment of nonintubated patients on low-dose selfotel revealed an increased number of deaths and other adverse events. A retrospective analysis did not confirm that Selfotel was associated with an increased mortality in TBI, but there was also no evidence that the drug was efficacious. A problem here was that a major portion of patients did not have intracranial mass lesions (contusion, subdural haematoma) on CT, questioning whether these had a treatment responsive brain injury. Both studies on tirilazad or selfotel underscore the significance of well designed and conducted phase-I and -II trials to characterize the pharmacokinetics of the agent, to confirm availability of drug in the brain, and to identify a sufficient number of patients with lesions responding to the drug. A major issue is the blood-brain barrier permeability of the agent under study. Further, the phenomenon of secondary deterioration - neurological worsening - turned out as a powerful predictor of poor outcome. The findings and conclusions of both clinical trials (tirilazad, selfotel) were utilized for a subsequent patient study on CP101-606 in consultation with the Pfizer company, the US Brain Injury Consortium, and the San Diego Clinical Trial Group. The patient population was a priori selected towards responsiveness of the brain lesions to the treatment. The major conclusions are: I Development of therapeutic regimens targeted towards the mechanisms of brain injury. II Availability of adequate preclinical data. III Directing treatment towards an appropriate patient population. IV Central gathering and interpretation of the neuroradiological findings. V Monitoring of trial center performance. VI Stratification and pre-trial prognostic analysis for identification of subgroups.