| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
,
Departments of *Anesthesiology,
Neurobiology, and
Surgery, Duke University Medical Center, Durham, North Carolina
Address correspondence and reprint requests to David S. Warner, MD, Department of Anesthesiology, Box 3094, Duke University Medical Center, Durham, NC 27710.Address e-mail to warne002{at}mc.duke.edu
Over the past several decades there has been little change in the probability that humans suffering from traumatic brain injury (TBI) will have a good outcome. Numerous randomized trials, aimed to identify effective pharmacologic interventions, have failed to offer meaningful advances in the care of these patients. Perhaps the most valuable studies were those that demonstrated adverse effects or lack of efficacy of commonly used physiologic interventions such as hyperventilation and moderate hypothermia (1,2). Other routine interventions, including the use of mannitol or ventricular drainage, remain unsupported by valid clinical trials (3,4).
This frustrating lack of progress can be attributed to numerous causes. Experimental therapies evaluated in laboratory TBI models seem to promise efficacy but when translated to clinical scenarios they repeatedly fail. Perhaps this is due to lack of validity of the animal studies or, instead, the irrational exuberance of clinical investigators who prematurely extrapolate such data to clinical scenarios quite different from those studied in the laboratory. We are also learning that acute brain injury is a far more complex disease than originally envisioned. It has become clear that our understanding of the pathobiology of TBI (and in particular, the brain’s response to it) is inadequate to allow meaningful advances. This does not mean that the situation is hopeless, but indicates the need for breakthrough strategies if TBI outcome is to improve as a result of clinical therapy.
One such breakthrough is the emerging recognition that not all patients with TBI are alike. TBI categorically represents a constellation of mechanical, excitotoxic, hemodynamic, apoptotic, inflammatory, and oxidative stresses that exist in a unique state at any one time for any one patient. Thus, a standardized intervention may at one time be beneficial, while at another time the same intervention may be useless or even harmful. The article published in this issue of Anesthesia & Analgesia by Cremer et al. (5) exemplifies that point. By studying severe TBI patients, they showed that pharmacological augmentation of cerebral perfusion pressure (CPP) was beneficial only when intracranial pressure (ICP) was >20 mm Hg. In patients with ICP values <20 mm Hg, pharmacologically augmented CPP provided no benefit but potentially exposed patients to increased risk of pulmonary and cardiovascular complications. Importantly, Cremer et al. (5) also showed that efficacy of CPP augmentation on cerebral physiologic well-being can be repeatedly defined for an individual patient. This can allow CPP management decisions to be made on an individualized basis as the patient recovers from the initial traumatic insult.
Because the brain swells within the rigid cranial vault, intracranial hypertension causes secondary ischemic insults. Thus, it has been a central tenet of TBI management that therapy be directed toward reduction of brain bulk and support of arterial blood pressure, yielding improved cerebral perfusion. In many intensive care units, only crude monitoring is available (e.g., ICP and arterial blood pressure). These measurements allow calculation of CPP. The dilemma is that no one knows what the optimal CPP should be, although consensus guidelines now recommend that CPP be maintained above 60 mm Hg (6). This is an important issue, not only because deliberate hypertension carries risk in TBI patients (7), but also because there is little evidence that management of large groups of patients at strictly controlled CPP values has any effect on overall outcome (8). This lack of a homogeneous response of a TBI population to a standardized intervention is consistent with the finding that interventions proven to reduce ICP are not necessarily associated with improved outcome when examined in large groups of patients (2,9–12).
When one examines the human TBI literature, a common thread emerges. A broad population of patients is entered into a clinical trial. Inclusion criteria typically require an admission Glasgow Coma Scale (GCS) score 
8. This means that all study patients will have an injury sufficiently severe to justify endotracheal intubation and mechanical ventilation. Despite this uniformity, this still represents a very broad spectrum of injury severity. Thus, these studies examine a population with heterogeneous types and severities of injury with the presumption that the tested intervention will be sufficiently potent to benefit most, if not all, patients.
The most common rationale for studying a heterogeneous population is that it allows more rapid recruitment of study patients so that trial conclusions can be drawn with expediency and reduced cost. While convenient, studies such as the Intraoperative Hypothermia for Aneurysm Surgery Trial (13) have shown that it is possible to efficiently collect large numbers of patients with relatively homogeneous acute brain insults through multinational collaboration. Such an approach is more likely to allow valid conclusions to be drawn.
On a physiologic basis, it simply is not justified to assume a commonality of disease across a broad range of TBI severity. Intracerebral microdialysis and tissue oxygen/pH sensors show marked flux of metabolic state within patients over time and in different brain regions (14–16). Positron emission tomography clearly demonstrates marked differences among TBI patients with respect to volume of ischemic tissue and flow metabolism mismatching despite similar CPP and jugular venous oxygen saturation (SjO2) values during either steady state conditions or in response to changes in arterial carbon dioxide partial pressure (17,18). Thus, fixation on the numeric value of an isolated global monitoring parameter, such as CPP, would not be expected to offer specificity in management of individual patients or provide meaningful effects on population outcome.
This is why the work of Cremer et al. (5) is important. Like most other investigators, they enrolled patients with GCS scores 8. CPP was routinely maintained at or above 70 mm Hg. ICP and brain tissue oxygen partial pressure (PtiO2) were continuously measured while middle cerebral artery blood flow velocity was used to define cerebral autoregulatory status. Under otherwise stable physiologic conditions, CPP was briefly increased to 105–110 mm Hg or reduced to 50–55 mm Hg and then returned to baseline. During either the mean arterial blood pressure increase to deliberate hypertension, or recovery from induced hypotension, the static rate of cerebral autoregulation was calculated and PtiO2 changes were recorded. The study was repeated daily during the first 3 days post-TBI, thus allowing assessment across a spectrum of conditions within individual patients. When intracranial hypertension was present at the time of testing, increases in CPP improved PtiO2 and autoregulatory capacity. This was associated with a decrease in ICP, presumably due to autoregulatory reduction of arterial cross-sectional diameter. In those patients with ICP < 20 mm Hg at the time of testing, increases in CPP caused little or no change in ICP, PtiO2, or autoregulatory capacity. Thus, increased CPP had a favorable impact on brain oxygenation and vasoreactivity on some occasions but not others. The lesson to be learned from this study is that the potential value of CPP enhancement can be physiologically assessed on a repeated basis in individual patients using simple maneuvers available in most neurointensive care units.
The study by Cremer et al. (5) did not use this method of defining CPP management in a sufficient number of patients to allow assessment of effects on outcome or rate of complications. Thus, it remains speculative whether or not their strategy is effective. However, their study demonstrates that a commonly used intervention has different effects on cerebral physiology across a population of severe head injury patients at different times during their convalescence. It is clear that TBI management is moving in the direction of continuous individualized assessment of neurophysiologic responses to intervention strategies so that optimal therapy can be offered with reduced risk of iatrogenic injury. Such individualized management may not only allow direct effects on outcome from TBI, but also may be important in defining subsets of the TBI population for whom experimental therapies may be effective.
 |
References |
|---|
 Top References |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
